GemPages Shopify Page Builder App
내용
[subtitle-meta] lang=en auto=False
So that brings us
to our next session,
AI and Funnel Marketing,
an in-depth overview
of AI era CRO trends
with a real world data and case
study across the funnel.
And to guide us
through this session,
please welcome back Mr.
Arvind, our moderator.
To come back to the
stage, Mr. Arvind.
Yeah, once again, welcome you.
So joining us, Mr. Arvind,
we would like to
invite to the stage,
Mr. Adam Pearce,
of Blend Commerce.
Yes, Mr. Adam.
Please, we're breaking down how AI
is revolutionizing
this whole funnel CRO,
and what a winning quarter four
setup really looks like.
And joining with both of you is
Mr. Alex Fedotoff from Gethookd.
Back again, yes.
Bringing over a decade of
e-commerce experience
to reveal how AI is transforming
modern ad
strategies and sales.
And also we would like
to invite you to stage,
Mr. Stefan from Ecommerce Today.
You will join from
hundreds of Shopify Plus brands
who shows how AI drives retention
and streamlining operations.
And lastly, I would like
to invite to the stage,
Mr. Louis Teo.
From Chronos Agency, sharing real
world case studies,
on leveraging AI for high-converting
email marketing.
So Mr. Arvind, so everyone please
have a seat first.
And Mr. Arvind, I
can't wait for more,
no, I pass the mic for you.
And from now, this stage is yours.
Thank you so much.
Folks, give a huge round of
applause for this panel again.
We're gonna talk about AI and
funnel marketing.
Huge round of applause, please.
I know it's panel
three for the day.
Hopefully I'm gonna collectively
take sad and say that.
We're gonna bring
that energy back.
It's gonna be
another heavyweight panel.
We have folks, experts flown in
from different parts of the world.
Thank you so much,
gents, for doing it.
Really excited to get started.
Before we do,
lovely, handsome pictures,
and none of those are AI.
They're all actual faces that
we've got over here.
We'll do a quick
round of introductions,
starting all the way over there.
Louis, thank you
so much for coming.
Do you wanna quickly
say one line about you
and what you do?
Hello, hello.
Louis, I run Chronos Agency.
We are a retention
marketing agency.
Email and SMS is kind of our main
choice of weapon
to help our clients
achieve better retention.
Outside of agency, I
also run Ecom North.
I believe that's how I
met Chris and Arvind.
We run an e-commerce event.
We have four locations right now.
Toronto, Singapore,
where's the other one?
Sydney.
Melbourne's coming up.
Melbourne.
And you know what,
I kind of forgot.
LA, LA.
LA.
LA is happening right
now, so I kind of forgot.
So that's kind of my second thing.
Yeah, that's pretty
much how to run an agency
and an e-commerce events company.
Excited to have you back again.
I think last year we had Josh,
who was your co-founder.
He joined us in the panel.
He's sitting in LA
as we speak today.
They're doing an event in LA.
So thank you so much for that.
Man with a man bun.
I would love to hear.
Stefan, thank you so much for
joining us from Romania.
I would love for you to tell a
little bit about yourself.
Hello everyone.
It's great to be here, Arvind.
It's great to be here, GemPages.
Chris, thank you for inviting me.
My name is Stefan.
I've been in
e-commerce for 16 years now,
my entire adult life.
I've been an
e-commerce business owner
in both Europe and in the US.
And I run e-commerce
today for the last six years.
32 people agency,
full service, Shopify,
Klaviyo, retention
marketing, you name it.
Thank you.
Thank you so much
Stefan for joining in.
And now, man in the middle
requires no introduction.
I'm gonna ask you something is,
introduce yourself
a little different
from what you did
in the first panel.
Yeah.
You're a fighter?
Yeah.
How many years you have?
We do e-commerce,
that's our main business.
So Gethookd is a
software that we built
for helping, helping
out more profitably.
And yeah, I've
generated over $200 million
in e-commerce sales.
I stopped counting because yeah,
it's like the first year
in e-commerce, I was just like,
we were just
churning and burning stores.
And yeah, now we run
brands with more longevity.
Don't worry, I think
we have, we can count.
If you have counting is a problem,
we can take over.
You can send you a bank account,
you can just,
where it jerks apart.
Here we have a handsome man who's
come all the way
from London, Adam
Pearce, gentlemen.
Adam, do you wanna tell a little
bit about yourself?
Good to be here, thank you, Arvind.
And yeah, guys, I'm Adam Pearce.
I am one of the
co-founders of Blend Commerce,
which is a Shopify CR agency.
I've been doing it
for about 11 years now.
I also mainly work
with US-based brands
and some UK brands as well,
mainly sort of eight and nine
figure beauty supplements,
some apparel.
And I also have
another business that I have,
which is called Ecom Collab Club,
which Arvind has
supported us up for a long time.
We run events in London once a
month, about 180 people.
And we've run those events all
across the world,
including New York, Sweden,
and a few other countries as well.
And yeah, great to be here back in
Vietnam after 15 years
and be with you guys.
Thank you so much,
coming back along.
Like we said, we
genuinely have a panel,
which is coming
from across the globe.
So thank you so much, Gens.
The purpose of this
panel here is to try and see.
We spoke about BFCM Playbook.
We started the note on that.
Then we went to Agente Commerce,
what's happening.
We got a platform point of view.
We got a frontier
model point of view.
Here we have
experts sitting over here.
The idea is to see that if all the
models are same,
if everyone types the same prompt,
if everyone tries to
sell similar products,
how do you differentiate?
That's something that
we want to touch upon.
We have the full
funnel and try to see
how do you kind of stand out in
each part of the funnel.
And I want to get
started with you, Adam.
I think, how is AI
changing CRO today?
Would love for you
to kickstart the ball.
Sure.
I think in some
respects, what I would say here
is that it is speeding things up,
but at the same time having to
slow things down.
And what I mean by
that is that, yeah,
from a CRO point of view, CRO,
like you can see here,
works across the
whole of the funnel.
And what it does with
AI is that it speeds up
on the development
front because it means
that you can build quicker,
like probably everyone
in the room is doing.
But at the same
time, what you have to do
is make sure that
from the quality,
assurance quality
control point of view
is that that's where the time
actually is spent.
So it is increasing the speed of
conversion re-optimization,
but maybe not as fast as we might
want to anticipate it to be.
When you say one connected funnel,
every stage can improve.
The question is, like
you said, 15 years back,
you were in Vietnam, and I could
say a little longer
than that, you've been doing CRO.
What's changed in today's D&H?
Well, we had a bit of
a wake up call really
about 18 months ago where we said,
look, we need to seriously start
thinking about AI
because we were doing
what I think a lot of us did
is go, la, la, la, la, la, la, and
just kind of ignore it.
And what we did is we sat down
effectively as a team
and said, right, what are all of
the really boring,
really mundane,
really admin heavy things
that our clients wouldn't really
want to pay for?
And that's kind of
where we started.
Now, we've seen some
great examples today
of some really cool
looking, sexy looking stuff
on the screen that is
very much front end.
We didn't start with that.
We started with a really
boring, really mundane,
but that's kind of helped us and
get from a process
that was very, very manual to now
being driven by AI
so that actually we sped up.
And equally is, I think from a
quality point of view,
it takes out that
the human mistakes
that obviously are gonna be there
when you are not doing
it in a kind of process
or AI driven way.
That's a pretty
personal wake up call
and kind of makes
sense to see that,
how your customer reacted
and how you proactively corrected,
cost corrected that.
I have the same question to you,
Louis and Stefan as well.
Both of you have
been running your agency
for close to a decade.
The way in which we operate,
the way in which we run the show
has kind of changed.
What was that wake up call
and when did that wake
up call happen to you
starting with Louis?
I think Adam is very right.
The landscape has
changed a lot for us
in the last one year.
What clients are,
what they kind of
value has changed a lot.
Creative production,
which is kind of our bread and
butter as agencies.
They kind of just want
an extra pair of hands,
they need an extra pair of brains.
The pair of hands
doesn't really matter anymore.
We internally, this is what we've
done for our clients.
We do email and SMS,
mainly as our business.
Internally our
designers and copywriters,
some of you I told them,
all of you got a three X output
and I cannot pay you more.
That's the reality
because clients are expecting
exactly the same.
Expecting three times faster,
while we charge our clients,
I used to be able to charge like,
"Hey, 15 emails
for $5,000 a month.
Today it's like 50
emails for $5,000."
That's kind of the reality.
And I think Namara mentioned
earlier on as well,
which is they're kind
of paying for results.
They don't really care about
the quantity, it just doesn't
matter that much anymore.
So internally, that's how we had
to align our team,
which is, "Hey guys,
this is kind of the reality.
If we don't reinvent
ourselves, that's it."
Fall behind, if not.
No, that's a great, great thing.
Stefan, how are things back in EU
and across the globe?
A, I want to ask, how
did you internalize it?
And B, generally,
how are you seeing this
as a pattern across?
Because you give a
lot of conferences,
you go, you're a
keynote speaker as well.
Want to take your perspective.
Yeah, so your question was,
what was the Eureka moment for us
as an organization
in order to actually use AI?
Differently than everybody else.
And it doesn't have
anything to do with like
email marketing or
CRO or SEO or GEO.
For us and for
everybody here in the room,
whether you're an app
developer, SaaS business,
merchant, brand, Shopify even,
for us it was organizational DNA.
The context that goes into the
actual AI assistants
that power each and
every one of our humans,
our human members of the teams.
The context about
our organization,
when we adjusted that
and when we actually made
our assistance aware about our
company's quarterly goals,
our company's yearly goals,
the actual content that came out
of the assistant changed.
Because if our goal was to
increase our profitability,
then an email prompt would have
given us an email
which takes 30 minutes to make.
But if our goal for
the next quarter was
to increase performance, increase
client retention,
get better
results for our clients,
then the prompt gives us a
different response
which is, "A, B,
split test this campaign,"
or "Do this, do that."
So, organizational DNA is
something that everyone
here in the room can
do, can define better,
and it is going to impact and it's
gonna change the way
your AIs respond to your teams,
and it's actually
gonna make you different
from everybody else.
Thanks for the sneak peek.
I think later in the day,
Stefan has built it on Showcase,
where he's gonna
expand a little bit more
on that organizational brain,
or the second brain that we
commonly refer to as,
more on that later.
But coming back to the CRO part of
the funnel, right?
I'm sure there's a
systemic process to it
on how you're doing.
For all the fellow agency folks,
sellers sitting on there,
what is your process to ensure
you kind of
personalize the full funnel?
Do you wanna walk us
through your steps?
Yeah, sure.
I mean, look, what
we're looking at here is,
I guess this is sort of an extent
the pre-AIA version,
in that when a Shopify merchant
will come to us,
the first thing that we would do
is what we call a heuristic audit.
So, i.e. looking through the site
from top to bottom,
and coming up with maybe 100 or
150 ideas or hypotheses.
Now, once that's done,
what we're then doing
is we're mining the
data from Shopify,
from GA4, from heat maps,
and what we're
trying to do is to say,
well, some of those
ideas that we had in step one
are stronger because we have data
associated with it.
Some of them fall away.
And I'm sure we've
all been that meeting
where we think, I think that we
should do this thing,
or I have a gut feeling that we
should do this thing.
That gets rid of
some of those ideas.
And then the final step, which is
the most important one,
is Pecte.
Now, I am sure
everyone sitting in this room
has a spreadsheet
with lots of ideas on.
You've maybe got emails you've
written to yourself,
little voice notes on your phone.
What Pecte does is allows you to
prioritize each of those.
And that's what we do
with all of these ideas.
So we have ideas, we have the data
that backs up those ideas,
and then we're prioritizing them
with the scoring system.
So proof is cost,
time, and impact.
And that's a proprietary framework
that you have over there.
It's not proprietary, no.
It is something that is used by a
number of different people,
but we have
personalized it specifically
for, I guess, the
e-commerce world.
And this just kind
of breaks it down.
I think, look, I would recommend,
it doesn't matter whether you're
doing CRO or anything,
realistically, in e-commerce.
There are always tons of ideas.
And breaking this
down affects you.
What you do is that
every idea that you have,
you score it out of five in terms
of each of these,
and then you roll that up into an
index out of 100.
Now, the beauty of
doing this is that,
for example, if you did this today
for maybe a brand
that is in the room,
that would be the situation today.
But as more data comes in,
maybe over the Black Friday
sub-minute period,
the data will shift,
and therefore the prioritization
of these will change.
And this is the
difficult thing, I think,
within e-commerce, isn't it?
Is that we always
take a snapshot of now,
but as we've heard
from this morning,
it is always something that's
dynamic and always moving.
So if you use a process like this,
whichever version that you use,
it just means that
you are maintaining,
then, changes to the site that are
right here, right now,
based on the data of that
particular moment.
That makes a lot of
sense to have a framework,
because right before this,
Alex and I, we were
having a conversation on
how is the output in
Dubai, SF, and India,
we were just talking about.
A good output
would actually stem out
from a really good prioritization.
Makes a lot of sense.
What's this fancy
dashboard that you have,
where you use it for a customer?
So this is our Pecti board.
So if you imagine, we've gone
through that process,
we've gone from
100 different ideas,
down to maybe 40 or 50,
either A-B tests or
implementations onto the site.
What this does is
that we have the scores
in terms of the stars,
in terms of proof fees, cost,
time, and impact,
and then the score out of 100.
So as you'd imagine,
you start in terms of
implementing the top
with items that score 100,
and then you go down that
level of prioritization.
So this just means it keeps
everyone focused,
and it also stops
those conversations
where maybe the CMO, head of
brand, the founder,
are saying, "No, I
think we need bundles."
No, I think we need landing pages.
This does, it kind of acts as,
I always say acts as a little bit
like a UMPs keeper,
and it comes and says,
"Here is the data to
say what the situation is,
"and here is the
roadmap that we then have
"to basically put into play."
And obviously with AI now,
this whole process I've described
has become a lot
faster, a lot more efficient.
By the way, that's
an actual screenshot
from one of his brands that he's
actually testing out.
Everybody can actually
connect, get an output,
get the framework,
but you built a
proprietary dashboard
to kind of further streamline it
to try and see how it works.
Can you walk us through on,
you have your idea,
you have your framework,
how do you distill it?
How do you give it to customers?
Can you tell us a little bit?
Sure, so we built
something in the 10,
they call Nugget Lab,
and what Nugget Lab
does is that it plugs in
all of the different data sources
that we talked about there,
and this screenshot
here is just basically
the dashboard that happens.
Now, I'm gonna say
something controversial.
I don't like conversion rate.
And the reason I
say that is because
I feel that
conversion rate is a metric
that you can hide behind.
The metric I prefer
is revenue possession,
because for me, what revenue
possession says is
how efficient is this website
at bringing cash into the bank
account of that business?
And what this does is that it's
basically tracking,
number one, the metrics,
but importantly
what it does is that
if we go onto that next slide,
we'll see here is
that once obviously
each of those
connectors are brought in
to the different data sources,
it is then making the
recommendations.
Once the
recommendations are then done,
the prioritization of Pecte is
also then done too.
Now, this is fed in from about
three and a half thousand
different A-B tests
that we've done. Indeed.
And off the back of that then,
it creates the prioritization.
So you can see
probably very clearly here,
the work that my team used to do,
which is very manual,
or kind of being
done with data analysis,
is now being done, you know,
relatively in a
short period of time
compared to what it used to take,
which was, you know,
weeks and weeks of work.
Is it okay for me to say that
while you're sitting in Saigon,
there is an adum which people are
able to chat in London
through the dashboard,
is it okay to say that?
Yeah, absolutely.
No, I think that's
pretty interesting.
Come to think of it, right?
I'm sure I'm gonna preempt a
question from the audience.
I could do this in Claude,
or I could do this
probably with a new model
with OpenAI is gonna
release, by the way.
They're gonna announce that in the
second half of the day.
How do you, let's say, blend in
the human and the AI to it?
How do you ensure
that, you know what,
there's a objective thing,
and there's a subjective thing.
How do you blend in to ensure
that your relationship is made in?
I think that the
important thing for us is,
like, when we did
this, is that we used
our sort of 11 years of experience
and put that into the model.
So, yes, absolutely.
And I think, look, it depends on
what kind of size brand
you are, because when
you're starting out,
you're not gonna be able to afford
to work with an agency.
And also, I don't
think you should,
realistically for me, unless a
brand is maybe doing
at least 25,
30,000 visits a month,
I don't think you should be
working with anyone like me.
So, you can use
that, your claw models,
you can use,
obviously, any of those models.
But as you don't
get to a bigger stage,
it's then good to kind of get some
real world examples
of A, B sets that
have actually been run
with similar brands to yourself.
Speaking of real world examples,
I have spoken to a
couple of your customers
while I was there at your event,
when I was touring and stuff.
I wanna understand how did this
help move the needle?
Do you wanna talk
about Stone Creek Coffee?
Yeah, sure.
So, nine figure coffee brand, two
brands across the US,
as well as a string of
coffee shops as well.
The beauty, obviously, now with
the kind of being able
to use AI is that the
kind of test that we can run
and discovering those is a lot
quicker than it was before.
It's not quicker in
the sense, obviously,
we still have to
wait for the traffic
because we need to get
statistical significance.
But I think what it
shows on the next slide here
is that basically, when we
inherited the site,
there was an issue
was picked up by AI.
And that was, is that on the
subscription page on the PDP,
the subscription
choices were hidden.
So what you saw is a dropdown.
Now, what we were able to say is
all very quickly
is identify that this is an issue.
We can see friction, not only in
terms of the GA4 data,
but also the heat map data.
And then when we go on
to the next version of it,
the A-B test, we simplified that
to be different
subscription options.
So much more visual, the
build of this particular
front-end part as well
was again done with AI.
And then as a result of
this, what we get is that
from this very specific change,
the attribution to that change
then actually leverages
about another $44,000 a year.
So just actually to that
particular change itself
on top of what was
there previously.
That's insane.
It's a probably classic case of
that Pareto principle, right?
20% gives you that outsized thing
and having those two changes.
If it gives that big a delta,
then we start a brand,
they're coming to
you for the playbook.
Well, how do we know that this is
not a one-hit wonder
kind of a model?
Sure, like it works for one.
How do we know that it doesn't,
or it does work for the other?
What are your thoughts
with the second case?
Yeah, so this is a UK-based brand,
eight figure business selling
barbecue equipment
and such like.
Again, with these guys, the
situation was similar
in that they needed them to have
first of all the analysis
and check in what was going on.
And what we saw is that very
different behaviors
by forms of payment.
And payment
methods obviously in Asia
is a very important
thing for us to consider.
And wherever you are selling
across the world,
something important.
Now what these guys
did is that they led
with the buy with shop.
And what we saw is
that the lifetime value,
and we see this a lot,
the lifetime value of
different payment methods
is always very, very different.
Now what we did with these guys is
that we actually then
said actually we need to expand
those payment options
that are shown.
We made this simple
change on the site,
which again was assisted by the
whole AI process,
making it quicker.
And as a result of this one,
but another $400,000 of revenue
attributed to that store.
Now, the one thing I
wanna pick up on here,
I have in this important is that
people in the audience
might say, well, this guy is
saying he does CRO.
The conversion rate
there has gone down 1.4%.
And that's my point.
Is that conversion rate alone is a
very poor measure
of how successful that
everyone or unsuccessful
the particular test is.
But the revenue per visitor here
gives us a clear review
of what we're doing.
So that would be my
message to anyone is that
consider revenue per visitor or
revenue per session
as a much more truer version of
how the site is performing.
That's the 11 years of
experience saying that
don't single out a metric,
look them through pulse together.
But how do I know if I'm a brand
or if I'm a seller
sitting on the audience, right?
How do I know that, okay,
what percentile am I standing
compared to the different brands
that's out there in e-commerce?
How do I kind of benchmark myself?
Sure.
The important thing
is that the way we all,
regardless of
whatever business you run,
we're all here to do three things.
We were here to get
customers to buy now.
We're gonna get
customers to buy more.
I'm gonna get
customers to buy again.
So buy now, buy more, buy again.
And that's what we
call the buy trifecta.
Now we built this tool that allows
any Shopify brand
to then go and pull
industry benchmarks
for your specific niche.
What it will do there is tell you
your starting point.
So each of these, buy now, buy
more, buy again,
we call as levers.
So your conversion rate, your AOV,
and your repeat purchase rate.
First of all, if you
use this to work out
which lever it is
that you need to pull,
you can then use that as the
guidance for the next steps.
So the steps I
described previously,
this happens before that.
So regardless of
whatever stage you're at,
use the trifecta tool on here to
get the starting point.
And then you can go forward
and look into CRO a bit more deep.
Now I wanna buy more
time to listen to this,
but we are capped onto this one.
What's this interesting slide?
Tell us more about it.
Yeah, absolutely.
So as I mentioned at
the start that I also run
an event called Econ collab club.
I've been running
about four years.
We have a community of about 7,000
e-commerce providers.
We have merchants and agencies.
We run events once
a month in London.
So if anybody
would like to join us
and obviously with our
collaboration now,
I have in with obviously ShopX.
This is something that we're gonna
be doing together more of.
And I'd say we have about 180
people gathering each month.
So if you are gonna be in London
or plan to be in London
over the course of
the next few months,
do come and say hi to us.
Lots of fun.
8 a.m. movie theater, e-commerce.
And it's nothing short of a movie.
I have been there.
Thank you so much,
Adam, for doing it.
A lot of this event
we drew inspiration
from Econ collab club.
Thanks for having us right there.
Hopefully we keep
expanding the ShopX network
to newer geographies.
That's about it.
And quickly going on
to the man to your left.
We spoke about CRO
in the full funnel.
Alex, your product stemmed out of
you building your brands.
That's what you said earlier.
Ads is a very
interesting part of it.
I think Bonnie covered a pretty
interesting thing
in the previous slide.
Making ads, I think the benchmark
has come down lower.
I think the entry to barrier is
pretty low right now.
Creatives, check.
P-Max, check.
Can I get enough in more
performance marketing agency?
Check.
Do we get credit cards?
We have air wallets in the group.
They're ready to
give a line of credit.
How do you still stand out?
Over to you.
That's the topic we
wanna start off with.
Yeah, for sure.
It's so easy to create ads now
with different tools.
I mean, ChargeGPT has tremendous
image generation.
Higgsfield has basically all of
the LLMs integrated into it.
And it can generate beautiful,
amazing creative.
Very fast and inexpensively.
So creating ads is not a mode.
So anyone can create lots of ads.
But the mode is, on the
advertising side specifically,
is in understanding your consumers
on a deeper level
and understanding what kind of
ads, what kind of formats,
what kind of messaging would
resonate with them
on a deeper level to get them to
buy your product.
And so that's kind of like how it
shifted over time.
Like, when I was starting out,
like e-commerce, like 2016,
I was using Canva
to make banner ads
and hiring freelancers to make
pages and stuff like that.
Now it's so much easier.
Many of those
things are just like,
even one person can go
through the whole cycle
from the initial idea to
actually creating the ad
and launching the ad
without any problems.
And it's like the landing page,
the conversion site,
also got simplified from that
standpoint as well.
So now it's really understanding,
like when you are
advertising, right?
I think this is a very
big point for everyone
who's spending money on ads.
You're not the only
person advertising.
And a lot of people,
like you look at their ads
and they're like,
their ads look like,
they think they're like alone.
There's no competition, like no
one selling similar products.
Hey, buy this
magnesium supplement.
It's like, yeah, like
there's like 300 other
magnesium supplements
that are being sold.
How is your magnesium
supplement different?
And so really like
what we focus on is
consistently analyzing that
competitive landscape.
Like who's advertising, what
angles that they're advertising
with and also micro
angles, which I'll explain.
Let's do that fun exercise.
I wanna do it.
I'm gonna role play
right now with you.
Let's say that I'm a supplement
brand owner right now.
Clearly you have a playbook,
that's why you're minting.
You stopped counting
beyond 200 million.
So something
that's working for you.
Let's say I am a
supplement brand owner.
That's why I'm making about a
million dollars already.
I need to make ads for this BFCM
or for a sale season.
Walk me through the
steps in which you suggest
that I kind of go ahead because
the time is huge.
I'm gonna sell it in all the
English markets.
How do I go about it?
What are my steps?
Yeah, for sure.
So number one,
you have to analyze,
let's say you're
already advertising?
Yeah, I'm already advertising.
Okay, so step number
one is really analyzing
your current ads and seeing which
parts of the market
you're covering and not covering.
So on every market, there is
different stages of awareness.
Raise your hand if you understand
stages of awareness.
Okay, yeah, it's
very important concept.
It's from Eugene Schwartz.
It's kind of very old, very
foundational marketing.
What that means is
in different market,
there are people that understand,
like there are
people that don't know
that they have a problem.
Okay, so first one, know the
problem awareness.
Yeah, yeah, so
they're like unaware.
They don't even know
that they have a problem.
So you have to
basically educate them
that they have a problem.
So then you can
sell them the solution.
Then there is problem aware.
So okay, oh, I
know I have a problem.
I'm actively
looking for a solution.
Then there is a solution aware.
So people already know like, oh, I
need a supplement.
And so they're
actively looking and choosing
which one
specifically they want to buy.
And then there's a product aware.
So this is where
Black Friday Cybermonet,
it's like 40% off, special offers.
People already know they want to
buy your product specifically.
You just give them
the best offer possible
and they make a purchase.
And so a lot of the advertising,
if you look at
advertising landscape,
it's like people like,
hey, buy this product,
buy this product,
buy this product.
But then there's no,
like the whole kind of like funnel
infrastructure, right?
Like where you took that
consumer from initially
not even knowing
they have the problem
to leading them to that solution.
They're just hammering those ads.
And those ads are
not very scalable.
You cannot spend a lot on these
like offer-based
very product aware ads,
because those are
only for people that know
about their product and
already want to buy it.
And probably drops
credibility also, right?
When you're a new brand out there,
you keep like slamming and buy
more, buy more, buy more.
That kind of drops on the
credibility also, right?
Yeah, yeah, yeah.
So, and so that's
one aspect, right?
Like of the whole kind of
advertising landscape.
So let's say you can analyze your
whole like ad library.
So I have an example here.
This is from like
one of the brands.
You can see like how
many ads they launched
in 12 months, like
63,000 ads, right?
And so basically this
is kind of the collective
holistic data.
They're also advertising for
multiple Facebook pages.
Who's advertising for more than
one Facebook page?
Okay.
That's quite a
few in the audience.
Yeah, that's good.
So yeah, that's one of the ways to
also kind of like,
because advertising and marketing
gets touch points, right?
So when you're advertising,
that's why you want to basically
hit people multiple times,
ideal at different stages of
awareness with your ads.
And it doesn't like,
you don't need to do
any targeting anymore.
Like targeting is outdated.
It's that you don't
need any targeting,
just like age,
gender, country, that's it.
That's all you need.
And Facebook will figure it out
because the algorithm is so good.
Andromeda is that new update,
which kind of that's that
targeting, I presume so.
Yeah, yeah.
So Facebook now at this stage,
I mean, they're collecting data
points from your behavior,
from every user.
So they know at which stage of
awareness your ads.
So they'll show you the right app,
even if you don't have any like
targeting settings.
And so yeah, let's say we, if we
took this brand, right?
And if we took it, we send it
through like Gethookd MCP,
which basically pulls all of the
ads from their library
and then analyze it based on the
stages of the awareness,
analyzes it also
on the micro angles.
Micro angle is like, it's
basically the domain,
kind of like reason why people buy
products, right?
So for example, if, you know, who
here has an iPhone?
Okay, so if I'll ask like each
person here separately,
like why did you buy iPhone?
You know, some people
say, oh, I bought it
because it has the best camera.
Some people would say, oh, I
bought it because,
I don't know, everyone has iPhone.
Everyone who's cool has an iPhone.
And it's like, and
different people,
oh, it's like, because
I could talk with my,
on the FaceTime, right?
With my parents, with
my kids, with my family.
And so different
people react to different,
kind of like emotional.
Messaging, positioning.
And so that's
what we need to have.
You need to have diversified,
diversified kind of
system in your marketing,
both on the awareness level.
So you're hitting all of these
layers from the unaware.
If you can track
unaware, that's like,
that's where you like
unlock the real scale.
That's like 50,
100K a day ad spend.
That's when you can push that, you
know, good volume.
And then on the
angle side is like,
you can analyze your whole
competitive landscape
in your niche specifically.
Then you can
cross-reference it with yours
and see exactly which angles,
which stages of
awareness you're missing.
So then you see the missing links
and specifically what
apps you need to launch
in order to have
this complete, you know,
final infrastructure
from start to finish.
And so that has been the biggest
like unlock for us,
previously it would
be just like manually
and like we would
have to look at every ad
and like think about it.
So AI simplified that
process of analysis.
But the good thing is, is that
very few people,
I mean, you guys
now know about this,
but like very few
people know about this.
Very few people have this like
holistic understanding
of advertising landscape.
So if you actually take
this to your advantage,
I think the, you know,
you unlock real skill.
I mean, that's a lot of things
that you just
said for that, right?
But I see on the
screen over here, you say,
what it took back in the day
versus what it's taking now.
How true is it?
Because let's say a
brand at your scale,
you're running a $200 million
brand out there.
Is it applicable
even at your level,
where you replace a team of a
large team to a smaller team?
Yeah, I mean, this is real.
Like, so for example, creative
strategies on our team,
like our creative strategies can
go from the ideation,
can analyze the
competitive landscape.
Okay, we're missing
these stages of the funnel.
We don't have underwear ads.
Okay, he can Gethookd MCP to
find the underwear ads.
Boom, he has a concept.
Then from there, you know, take,
extract the script,
rewrite it using, you
know, Clod or OpenAI,
and then create like a storyboard
of his video ad,
if it's image ad generated using
ChudgPT and like,
you know, launch it basically
within like a few hours.
So that whole
cycle that previously,
oh, you know, you
send it to designer,
designer's like, yeah, and you
know, creates a mock-up,
sends it back to you, oh, let's do
these revisions.
Let's adjust it back
and forth, back and forth.
Now it's like, bam, bam, bam.
But the important thing is like,
you need to know
exactly what to focus on.
And that's kind of,
it's not, you know,
the tools and everything makes it
easier to create stuff,
but you need to know
exactly what to create.
I'm pretty hooked on that line.
Thank you so much, Alex.
So how we make
VLSs with AI, right?
Like the research and stuff.
Do you want to quickly tell us
what is your checklist?
Yeah, this is just one of the
formats that we run.
I mean, there is a, I
mean, last time I checked,
there's like 30 plus like
different ad formats
that we run.
This VSL is just one of them.
VSL is like the video that's
typically longer
than like three minutes.
So they're the most scalable.
Out of all of the ad formats,
we found them to
be the most scalable
because if you crack a VSL,
we've had VSLs spend like three,
more than $3 million.
Like one VSL, one ad, profitably.
And so this is just,
you know, the process,
all of this process can be done.
I mean, for this one, because VSL
is a longer format,
so you need a
video editor for this,
but let's say creative strategies
can do the research,
write a script, because
copier, in my experience,
copywriters and
creative strategies
are kind of like merged together.
Like you don't need, I don't
believe you need,
you don't need like a separate
like copywriter.
Creative strategies, the core
skillset has to be the copy.
And then they can
write the script,
they can create the whole like
story line, storyboard,
then create the
footages, be it all,
A roll as well using Higgs filter,
if you have like a real
footage, real creator.
And then also there's like
separate AI workflow
that we use based on
all of our previous data
to rank the ad, the copy, all of
the factors against,
you know, all of the previous
winners we've had
to see how likely is
that one to succeed?
If it requires some revisions,
then we send it for
revisions and then launch.
Yeah, like I was
catching up with you,
clearly this format has been
working pretty well for you.
That's why I wanted you
to kind of break it down.
Makes sense.
I don't know if you
guys caught up on the tab,
hardly from Shopify recently made
a tweet last night.
They have announced an
integration with OpenAI
where Shopify ads can be run
within chat GPT,
like being a
generative engine optimized,
being visible in a search engine,
it's official right now and they
have some credits unlocked.
That's a partnership they kind of
unlocked last night.
I think that's a
segue to go to Stefan.
Two places of money
already flows in, right?
When we talk about a funnel,
I think we spoke about the full
funnel optimization,
we spoke about ads.
What are brands or people who buy
are looking up on
the search engine?
I mean, gone are the days that I
scroll down the page,
I go to the bunch of blue links
that's out there,
click on a link and
then get redirected.
Answers I look up over there.
What do you think are the places
where people search
and how do they
kind of translate on?
Yeah, I just want to
take a step back first
before I go into this and I want
to say one thing.
There's a lot of uncertainty.
There's a lot of noise.
There's a lot of fear about what's
happening with AI
and how things are
changing just yesterday.
In like six hours,
four different people,
while we were
networking, asked me,
"Do agencies still
have a future next year?
"How are things
changing for apps?"
And I was like,
"You're the fourth person
"that asked me this today."
And I'll go and I'll explain.
I'm sure there's
gonna be many more today.
(laughing)
Exactly.
So the thing is
with all of this noise,
we tend to get
distracted very easily.
And for my entire
professional life,
I've always been
successful doing one thing,
which is finding
the low-hanging fruit.
Because there's a lot
of shiny toys around.
There's Higgs Field.
There's 64,000 ads in one year.
It's amazing, it's
unhuman basically.
It would not be
possible two years ago.
So there's a lot of
shiny toys out there
and we tend to forget the
low-hanging fruit.
And there's like two things that
we do for all of our clients
and everyone here can do.
And it's really not
just for e-commerce.
SaaS businesses can do this for
their own businesses,
for their own sites.
And those two things are focusing
on traffic and acquisition
and focusing on personalizing for
the intelligence.
So making the
store legible for LLMs.
On this one, I actually have a
very simple example
from one of our clients.
We particularly chose a
mid-sized Shopify brand
from the US, which
is seven figures.
And we chose them because if you
go for a big brand,
then the numbers
are gonna be skewed.
If you go for a smaller brand,
the numbers are gonna be skewed.
Mid-sized is actually gonna show
what doing the work,
optimizing your PDPs,
personalizing the customer
experience in ways
in which I'm gonna show very soon,
actually helps you
to increase traffic.
And if we look at the
Google Search Console,
everybody knows in August,
there was like this
update from Google
and a lot of websites
traffic went actually down.
For this specific
brand that we did,
just what I'm about
to show you right now,
things right,
traffic continued to go up
because it's low-hanging fruit.
You need to focus on
improving the actual content
in a way that's
legible for humans,
but most importantly
for agents as well.
I like how you put it as a
low-hanging fruit
because we are still very early in
the learning curve
for us, we are still
catching up and stuff.
That's a pretty impressive number
that you see over there, right?
A certain amount of citations,
you've seen the visibility.
Clearly the numbers
speak for itself, right?
Definitely want to
uncover and try to see
what was your prep that led you to
the graph over there
and how can we translate the same
for the upcoming
windows that we have?
Well, the prep is
always going to be
understanding your
business very well
and making sure that
you're not just prompting
into an LLM of a choice
and just asking, help me improve
this product page
because if the LLM doesn't have
your business context,
if the LLM doesn't have your
organizational DNA,
which is, I'm gonna
keep talking about
this organizational DNA,
the LLM is gonna give you the
exact same answers
and the exact same information
is giving to everybody else.
So the prep, I think the
foundational prep
is making sure
the context is okay.
In our case, being an agency,
we have the client context,
which is brand voice,
products, top sellers,
history, et cetera.
It has another layer of context
which is our
relationship to the client.
So who we are, how
we started working,
meeting briefs
and everything else.
And a third layer of
context, which is us,
our organization,
what makes us different
from a different agency, what
gives us a different voice,
what we're focusing on
the low hanging fruit.
So it has to know these things.
No, speaking of this one,
one interesting thing that I
noticed is Itaos,
what is that, do you wanna quickly
tell us about it?
Yeah, Itaos is our
operating system.
We just built it
this year for ourselves
and whenever I
showed it to clients,
they were like, oh
my God, I want this.
So it actually spin off into a
separate business
and we signed a contract with a
huge civics organization
and we're building their own
operating system.
I just wanna show
you guys one thing.
So this is basically
asking for a Q4 planner
for one of the
clients and this is a demo.
So all the data you
see here is non PII.
I'm from European Union, so GDPR
is very important.
So this data is anonymized.
Just look at the
recommended Q4 approach.
Keep the planner standard and then
when it goes here,
it says that we need to discuss
with this client's request
for a performance linked bonus.
So the assistant
knows our contract,
it knows our
relationship with the client itself
and you can do this with
your own AI assistance.
Make sure it
understands your business goals,
whether those goals
are like conversion rate
or revenue per
session or ads spend.
If it has that context, it's gonna
give you better data.
What's pretty interesting is that
every single one
of them over here,
I think they had a
problem statement.
They went into builder mode,
him building a framework for CRO,
you building Gethookd
and you building it has
and Louis has something
interesting as well.
I think that's the best way in me,
which I think to stand out in this
commoditized AI world over there.
Speaking of which, now that you
have an assistant,
this cloud, this
open AI and stuff,
what questions we
used to ask before,
what questions do
we ask now so that
there's a personalized way in
which we get an output
so that it's not that
same cookie cutter response
that we all get, right?
End of the day, we wanna have that
amazing citation
that we wanna have.
How do you differentiate on that?
Yeah, so this is a
very simple experiment
that anyone in the room can make.
Whether you're a SaaS business
or you're an e-commerce business,
just go to any AI out
there that you wanna use
and have it do some deep research.
What are people
asking about my brand?
What are people asking about my
top selling product X?
And you're gonna get those 30
minutes long exercises
with questions.
It's gotten longer, right?
I think with every new model,
do you feel that Alex,
yeah, it's getting longer.
Now I'm like TLDR of
that also, I just go to it.
And you pick the top
three, four, five questions
over there and then you use a
prompt in which you're like,
hey, have a look at
this product, URL only.
Do not deep research, do not go
through the internet,
just based on the data
that exists on my PDP.
Am I answering any of
these four questions?
And sadly, most of the times the
answers are gonna be
the page does not state this, the
page does not state that
over and over again.
Experiment, if you
can turn the next slide.
Okay, so this is
basically the page
that does not state this before.
You're gonna have the image
information, standard image,
no title, no alt text, standard
product description.
Everybody has the same,
nobody differentiates.
And standard
frequently asked questions,
actually these are not standard.
We still see a lot of
sites with FAQs like,
how long will it take until my
order gets here,
which is completely useless from a
GEO perspective.
So this is the page
that got these prompts back
and you can see, is
this good for you?
Question one, the
page does not state this.
It never mentions,
polls, transition,
vert and everything else.
And for every single question,
I'm gonna speed up
because I saw that sign.
For every single question is the
page does not state this.
So that was the problem.
What did we do?
We asked AI based
on the questions,
based on our organizational DNA
and the client information
to optimize this product page
and based on the
research that we did.
Optimize the image, optimize
description, optimize the FAQs.
And then when we went
and reran the experiment,
if you can go to the next slide,
it does state that the page states
this information
about the product.
The product is a good fit for
specific individual.
The product does
need this and that.
So it does answer
to these questions.
It is visible now to
the robot, to the agent.
And it was
basically zero out of zero
and now it was four out of four,
zero out of four
and then it's four out of four,
which is something that everyone
here in the room can do.
Everyone here in the room
understands they need to do
and it's there for the taking.
Make yourself
legible for the agents,
if I may have to summarize that
and it kind of makes sense.
Yeah, I think this is
how the rework happens
and this is the answer, I presume.
Yeah, exactly.
So this is basically just as same
page in your test.
We got yes on every single one of
the most important
questions that people are asking
about that product online.
And again, this
works for SaaS as well
for service companies.
Agency also, I presume.
Agency can also make them visible.
We did that for our own website
and started adding that type of
information like a year ago
and it definitely worked.
Wanna jump
straight into it, right?
I think gonna take a cue on this
one because GEO,
we didn't cover
much in the BFCM panel.
I think we are eight
weeks away to the calendar.
If I am running a business, I'm
not gonna say a seller,
I'm not gonna say an
agency or a SaaS owner,
I'm gonna say I'm
running a business.
I wanna be
visible in the ZI world.
What are my eight weeks for me to,
what are my checklists for me to
be ready in eight weeks?
It's always gonna
be the 80-20 rule.
So focus on, if
you're in e-commerce,
20% of your products,
20% of your collections
slash category pages,
20% of your landing pages
and you want to make
sure those are optimized
as soon as possible with content
that's at everyone's disposal
as long as you personalize it for
your own identity.
So we wanna make sure we do that.
It's data work but it's AI
assisted nowadays.
It no longer takes months to do,
it takes days to do.
People are scared about data work.
They're scared about,
oh, we have to change this
for 20,000
products on our website.
But when you do the actual math,
it's not like it's
2010 or 2015 anymore.
Operating on 20,000 products is
much, much, much quicker.
Then you wanna make sure you get
your measurement in place
and there was one thing
earlier, when I saw Sam,
he had like zero
point, under 1% of traffic
came from a genetic search.
And that may be correct
for an entire industry
but what we've seen for our
clients is that this year
we are gonna most likely go over
10% from Agenic.
So I think that threshold of
Agenic referral traffic
versus SEO referral traffic is
gonna happen much, much
sooner than in the next 12 months.
That's super awesome.
I know that we have
little spilled over time.
Chris, next time we
should have workshops
instead of panels.
Clearly I could spend a whole day
listening to all of this.
Well, going on, last
but not least, right?
Going on to the
retention part of the funnel.
We've done our conversions, we
have run our ads well,
we've made ourselves visible.
The traffic has
started flowing in.
But end of the day, if
I have a leaky bucket,
I don't know how that's gonna go,
Xavier actually had
an interesting quote.
For every $3, $2 actually comes
from your email revenue.
That's the latest quote that he
got from K-Boss.
You work very closely with Klaviyo.
You're one of the
leading agencies out there.
Wanna try and understand with you
about the retention part
of the funnel, right?
How do you personalize it?
If you're going to do that same
volume of emails, SMS,
push notifications, or direct
mails, the same way
in a cookie cutter
answer, how do we step out?
Cool.
That's a tricky question.
I'm the guy who, every
time Klaviyo asks me, hey,
ask me a question,
it always depends.
It kind of depends on your goal.
And so that's kind of a tricky
question to answer.
I have a few, I think what I can
share is that there are
a few developments in the
retention marketing space
that has changed a
lot because of AI.
But I'm gonna say, I probably have
very little to share
today because all of them have
shared everything
I'm gonna share.
You need to have a brain, the DNA,
to basically run your company.
Now that AI has been
able to help us that much,
you kind of still need a really
smart people on your team
to make that human
judgment, you know?
And you kind of still
need a matrix system
to kind of decide, okay, now that
AI is telling me
to do these 50 things,
what should I be doing?
Right?
But let me show you one video.
If I cannot convince you today,
let the CEO of 8Sleep
convince you that you
probably have to rethink
how you are using AI
to run your company.
The tech gods are
not in our favor.
I'm not able to play
the video at the moment,
but jumping straight
into the case study, right?
Like if you wanna actually walk us
through on the format.
Sure, so I'll show you
guys really quickly here.
This is a case study
for one of the clients
called Mobile Tire Shop.
They're the largest mobile tire
fitting company in Australia.
And the four things they're kind
of seeing on the screen
is exactly what the rest of the
panel have mentioned.
Inputs, playbooks, which are kind
of the best practices.
We agencies love to say that
because we have a
lot of clients, right?
So we kind of ingest all
of these best practices
into our company brain.
I think Stefan uses
a custom software.
We use Obsidian and Clod,
whatever that kind of
floats your boat, right?
And human discretion, I think
still think that's important.
We're hiring a lot of
strategies right now,
a really good strategist.
Like Alex mentioned,
last time we need like three
or four or five people.
Now we just need one
really good strategist
to kind of make that happen.
And the last part, it's AI and
human execution.
So the earlier video that I wanted
to show you guys
from It's Sleep CEO,
you guys know It's Sleep.
I think they are reasonably valued
at about one and a
half billion dollars.
The video was saying, oh, we do a
hundred million dollars
from email marketing.
Last time I have two people, but
we're gonna quit.
So we're like, what do we do?
My co-founder went to, he was
like, give me three days.
He managed to automate their
entire email marketing system
with bots, just bots only.
And I believe them because I look
at their emails,
It's Sleep is different.
They already have
the brand equity.
Email is something that they can
probably automate
a lot with bots.
It serves that when you are a
million dollars in revenue
and you're one and a
half billion dollars,
email marketing means very
different, okay?
So that's for us as well.
We were able to generate 50 emails
for mobile tire shop.
This used to take three
months, two years ago.
We could charge a lot more.
It's a three
months project, right?
But what do we do to
kind of speed that up?
Agency, we have
sales calls, right?
We do like one or two months of
sales call to close this.
These companies are VC bags,
so they gotta go
through a lot of layers.
But every single call recorded.
Every internal
call recorded, right?
We just throw in one prompt
into our weapon of choices cloud,
hey, based on all the
calls that recorded,
all of the audits
that we have done,
the Klaviyo MCP
that's being connected
and also connected
to our company brain
and whatever
software that we are using,
generate 50 emails from me
according to the pitch, right?
This sucks because this is making
my job way too easy
and everyone can do
it as well, you know?
But obviously, why people are
still coming to us,
first, not the
best use of at a time.
Second, like all of the panels
here have mentioned,
that company brain, that DNA is
really important,
your best practices, that's what
people are coming to you for.
They're like, "Chronos Agency, we
have over 400 brands,
they've been doing
this for eight years."
They probably know
that their company brain
is probably better, right?
LLMs are great, but
they're too general.
If it makes your jobs easy
and still pays you that figure
that you quoted,
I think that's job
well done over there.
Lastly on this one, right?
Rich personalization
at the speed of light.
Okay, the Terrebonne,
because we still have
the last panel of this
morning's chat center.
So, yeah, it's the
time that I do believe
that we share a
lot of thought food
and insightful information for
this session's end panel.
Once again, thank you, you all.
Oh my God, so productive.
Can we have a big
applause for them?
원문 자막 펼치기
WEBVTT Kind: captions Language: en 00:00:02.333 --> 00:00:03.291 So that brings us 00:00:03.291 --> 00:00:04.750 to our next session, 00:00:05.083 --> 00:00:06.916 AI and Funnel Marketing, 00:00:07.208 --> 00:00:08.833 an in-depth overview 00:00:09.208 --> 00:00:11.708 of AI era CRO trends 00:00:11.750 --> 00:00:13.375 with a real world data and case 00:00:13.375 --> 00:00:14.791 study across the funnel. 00:00:15.166 --> 00:00:15.958 And to guide us 00:00:16.000 --> 00:00:16.916 through this session, 00:00:17.125 --> 00:00:18.791 please welcome back Mr. 00:00:18.958 --> 00:00:21.250 Arvind, our moderator. 00:00:21.791 --> 00:00:22.791 To come back to the 00:00:22.791 --> 00:00:24.125 stage, Mr. Arvind. 00:00:24.791 --> 00:00:26.750 Yeah, once again, welcome you. 00:00:28.500 --> 00:00:30.541 So joining us, Mr. Arvind, 00:00:30.916 --> 00:00:31.500 we would like to 00:00:31.500 --> 00:00:32.708 invite to the stage, 00:00:32.958 --> 00:00:34.458 Mr. Adam Pearce, 00:00:34.916 --> 00:00:36.708 of Blend Commerce. 00:00:37.791 --> 00:00:39.541 Yes, Mr. Adam. 00:00:40.708 --> 00:00:42.333 Please, we're breaking down how AI 00:00:42.375 --> 00:00:43.791 is revolutionizing 00:00:43.791 --> 00:00:45.458 this whole funnel CRO, 00:00:45.875 --> 00:00:47.750 and what a winning quarter four 00:00:47.750 --> 00:00:49.208 setup really looks like. 00:00:50.208 --> 00:00:52.583 And joining with both of you is 00:00:52.583 --> 00:00:55.958 Mr. Alex Fedotoff from Gethookd. 00:00:56.666 --> 00:00:58.208 Back again, yes. 00:00:58.500 --> 00:00:59.916 Bringing over a decade of 00:01:00.000 --> 00:01:01.125 e-commerce experience 00:01:01.125 --> 00:01:03.291 to reveal how AI is transforming 00:01:03.750 --> 00:01:04.375 modern ad 00:01:04.375 --> 00:01:06.291 strategies and sales. 00:01:07.041 --> 00:01:08.208 And also we would like 00:01:08.208 --> 00:01:09.041 to invite you to stage, 00:01:09.541 --> 00:01:14.875 Mr. Stefan from Ecommerce Today. 00:01:17.000 --> 00:01:18.000 You will join from 00:01:18.000 --> 00:01:19.750 hundreds of Shopify Plus brands 00:01:19.750 --> 00:01:22.708 who shows how AI drives retention 00:01:22.708 --> 00:01:24.750 and streamlining operations. 00:01:24.750 --> 00:01:27.500 And lastly, I would like 00:01:27.541 --> 00:01:29.416 to invite to the stage, 00:01:29.500 --> 00:01:32.458 Mr. Louis Teo. 00:01:33.541 --> 00:01:35.958 From Chronos Agency, sharing real 00:01:35.958 --> 00:01:36.958 world case studies, 00:01:37.250 --> 00:01:39.333 on leveraging AI for high-converting 00:01:39.541 --> 00:01:40.708 email marketing. 00:01:41.541 --> 00:01:45.000 So Mr. Arvind, so everyone please 00:01:45.750 --> 00:01:46.833 have a seat first. 00:01:47.875 --> 00:01:49.041 And Mr. Arvind, I 00:01:49.041 --> 00:01:50.125 can't wait for more, 00:01:50.125 --> 00:01:51.708 no, I pass the mic for you. 00:01:51.708 --> 00:01:55.500 And from now, this stage is yours. 00:01:56.000 --> 00:01:57.250 Thank you so much. 00:01:58.333 --> 00:01:59.625 Folks, give a huge round of 00:01:59.625 --> 00:02:01.250 applause for this panel again. 00:02:01.291 --> 00:02:02.541 We're gonna talk about AI and 00:02:02.583 --> 00:02:03.333 funnel marketing. 00:02:03.375 --> 00:02:04.833 Huge round of applause, please. 00:02:07.458 --> 00:02:08.125 I know it's panel 00:02:08.166 --> 00:02:08.958 three for the day. 00:02:09.166 --> 00:02:10.500 Hopefully I'm gonna collectively 00:02:10.541 --> 00:02:11.916 take sad and say that. 00:02:12.083 --> 00:02:12.541 We're gonna bring 00:02:12.541 --> 00:02:13.500 that energy back. 00:02:14.125 --> 00:02:14.541 It's gonna be 00:02:14.541 --> 00:02:15.750 another heavyweight panel. 00:02:16.291 --> 00:02:18.708 We have folks, experts flown in 00:02:18.708 --> 00:02:19.750 from different parts of the world. 00:02:19.791 --> 00:02:20.458 Thank you so much, 00:02:20.458 --> 00:02:21.541 gents, for doing it. 00:02:22.125 --> 00:02:23.458 Really excited to get started. 00:02:23.458 --> 00:02:24.166 Before we do, 00:02:24.833 --> 00:02:25.958 lovely, handsome pictures, 00:02:25.958 --> 00:02:27.083 and none of those are AI. 00:02:27.166 --> 00:02:28.416 They're all actual faces that 00:02:28.416 --> 00:02:29.208 we've got over here. 00:02:29.666 --> 00:02:30.208 We'll do a quick 00:02:30.208 --> 00:02:31.083 round of introductions, 00:02:31.083 --> 00:02:32.333 starting all the way over there. 00:02:32.708 --> 00:02:33.583 Louis, thank you 00:02:33.583 --> 00:02:34.583 so much for coming. 00:02:34.583 --> 00:02:35.208 Do you wanna quickly 00:02:35.208 --> 00:02:36.291 say one line about you 00:02:36.291 --> 00:02:36.958 and what you do? 00:02:40.041 --> 00:02:40.625 Hello, hello. 00:02:41.250 --> 00:02:43.291 Louis, I run Chronos Agency. 00:02:43.375 --> 00:02:44.583 We are a retention 00:02:44.583 --> 00:02:45.541 marketing agency. 00:02:45.875 --> 00:02:47.666 Email and SMS is kind of our main 00:02:47.666 --> 00:02:48.416 choice of weapon 00:02:48.416 --> 00:02:49.375 to help our clients 00:02:49.375 --> 00:02:51.333 achieve better retention. 00:02:51.958 --> 00:02:53.250 Outside of agency, I 00:02:53.250 --> 00:02:54.291 also run Ecom North. 00:02:54.333 --> 00:02:55.166 I believe that's how I 00:02:55.166 --> 00:02:56.916 met Chris and Arvind. 00:02:57.333 --> 00:02:58.958 We run an e-commerce event. 00:02:59.750 --> 00:03:02.041 We have four locations right now. 00:03:02.458 --> 00:03:04.083 Toronto, Singapore, 00:03:05.125 --> 00:03:05.791 where's the other one? 00:03:05.791 --> 00:03:06.291 Sydney. 00:03:07.041 --> 00:03:07.916 Melbourne's coming up. 00:03:07.916 --> 00:03:08.250 Melbourne. 00:03:09.333 --> 00:03:10.333 And you know what, 00:03:10.333 --> 00:03:10.875 I kind of forgot. 00:03:10.875 --> 00:03:11.291 LA, LA. 00:03:11.833 --> 00:03:12.166 LA. 00:03:12.291 --> 00:03:13.041 LA is happening right 00:03:13.041 --> 00:03:14.000 now, so I kind of forgot. 00:03:14.375 --> 00:03:16.750 So that's kind of my second thing. 00:03:17.458 --> 00:03:18.333 Yeah, that's pretty 00:03:18.333 --> 00:03:19.541 much how to run an agency 00:03:19.541 --> 00:03:21.166 and an e-commerce events company. 00:03:22.000 --> 00:03:23.541 Excited to have you back again. 00:03:23.666 --> 00:03:26.041 I think last year we had Josh, 00:03:26.041 --> 00:03:27.291 who was your co-founder. 00:03:27.291 --> 00:03:28.375 He joined us in the panel. 00:03:28.541 --> 00:03:29.541 He's sitting in LA 00:03:29.541 --> 00:03:30.708 as we speak today. 00:03:30.750 --> 00:03:32.291 They're doing an event in LA. 00:03:32.500 --> 00:03:33.541 So thank you so much for that. 00:03:34.000 --> 00:03:35.083 Man with a man bun. 00:03:35.291 --> 00:03:36.291 I would love to hear. 00:03:36.291 --> 00:03:37.500 Stefan, thank you so much for 00:03:37.500 --> 00:03:38.541 joining us from Romania. 00:03:38.833 --> 00:03:39.666 I would love for you to tell a 00:03:39.666 --> 00:03:40.500 little bit about yourself. 00:03:41.291 --> 00:03:42.000 Hello everyone. 00:03:42.416 --> 00:03:43.541 It's great to be here, Arvind. 00:03:43.583 --> 00:03:44.625 It's great to be here, GemPages. 00:03:44.625 --> 00:03:45.958 Chris, thank you for inviting me. 00:03:46.083 --> 00:03:47.000 My name is Stefan. 00:03:47.166 --> 00:03:47.958 I've been in 00:03:47.958 --> 00:03:50.125 e-commerce for 16 years now, 00:03:50.125 --> 00:03:51.541 my entire adult life. 00:03:52.125 --> 00:03:52.500 I've been an 00:03:52.500 --> 00:03:53.666 e-commerce business owner 00:03:53.791 --> 00:03:55.500 in both Europe and in the US. 00:03:55.500 --> 00:03:57.000 And I run e-commerce 00:03:57.000 --> 00:03:58.958 today for the last six years. 00:03:59.041 --> 00:04:00.291 32 people agency, 00:04:00.291 --> 00:04:01.500 full service, Shopify, 00:04:01.541 --> 00:04:02.541 Klaviyo, retention 00:04:02.541 --> 00:04:03.291 marketing, you name it. 00:04:03.958 --> 00:04:04.333 Thank you. 00:04:04.750 --> 00:04:05.333 Thank you so much 00:04:05.333 --> 00:04:06.416 Stefan for joining in. 00:04:06.458 --> 00:04:07.666 And now, man in the middle 00:04:07.666 --> 00:04:08.875 requires no introduction. 00:04:09.208 --> 00:04:10.375 I'm gonna ask you something is, 00:04:10.625 --> 00:04:11.666 introduce yourself 00:04:11.666 --> 00:04:12.500 a little different 00:04:12.500 --> 00:04:13.166 from what you did 00:04:13.166 --> 00:04:13.833 in the first panel. 00:04:14.166 --> 00:04:14.541 Yeah. 00:04:15.666 --> 00:04:16.375 You're a fighter? 00:04:16.458 --> 00:04:17.375 Yeah. 00:04:17.416 --> 00:04:18.250 How many years you have? 00:04:18.291 --> 00:04:20.166 We do e-commerce, 00:04:20.375 --> 00:04:21.416 that's our main business. 00:04:21.416 --> 00:04:22.458 So Gethookd is a 00:04:22.500 --> 00:04:23.583 software that we built 00:04:24.458 --> 00:04:27.208 for helping, helping 00:04:27.208 --> 00:04:30.541 out more profitably. 00:04:31.625 --> 00:04:32.875 And yeah, I've 00:04:32.875 --> 00:04:34.208 generated over $200 million 00:04:34.583 --> 00:04:35.583 in e-commerce sales. 00:04:36.000 --> 00:04:38.250 I stopped counting because yeah, 00:04:38.416 --> 00:04:40.208 it's like the first year 00:04:40.291 --> 00:04:41.375 in e-commerce, I was just like, 00:04:41.416 --> 00:04:41.875 we were just 00:04:41.875 --> 00:04:43.166 churning and burning stores. 00:04:44.125 --> 00:04:46.833 And yeah, now we run 00:04:46.833 --> 00:04:48.833 brands with more longevity. 00:04:49.291 --> 00:04:49.958 Don't worry, I think 00:04:49.958 --> 00:04:50.958 we have, we can count. 00:04:50.958 --> 00:04:52.000 If you have counting is a problem, 00:04:52.000 --> 00:04:52.791 we can take over. 00:04:52.791 --> 00:04:54.125 You can send you a bank account, 00:04:54.250 --> 00:04:54.625 you can just, 00:04:54.958 --> 00:04:55.833 where it jerks apart. 00:04:56.041 --> 00:04:57.500 Here we have a handsome man who's 00:04:57.500 --> 00:04:58.041 come all the way 00:04:58.041 --> 00:04:59.250 from London, Adam 00:04:59.250 --> 00:05:00.083 Pearce, gentlemen. 00:05:00.708 --> 00:05:01.833 Adam, do you wanna tell a little 00:05:01.833 --> 00:05:02.625 bit about yourself? 00:05:03.083 --> 00:05:04.208 Good to be here, thank you, Arvind. 00:05:04.250 --> 00:05:05.791 And yeah, guys, I'm Adam Pearce. 00:05:05.833 --> 00:05:06.500 I am one of the 00:05:06.500 --> 00:05:08.500 co-founders of Blend Commerce, 00:05:08.791 --> 00:05:11.000 which is a Shopify CR agency. 00:05:11.000 --> 00:05:11.333 I've been doing it 00:05:11.333 --> 00:05:12.750 for about 11 years now. 00:05:13.166 --> 00:05:15.750 I also mainly work 00:05:15.750 --> 00:05:17.750 with US-based brands 00:05:18.000 --> 00:05:19.333 and some UK brands as well, 00:05:19.333 --> 00:05:20.458 mainly sort of eight and nine 00:05:20.458 --> 00:05:22.375 figure beauty supplements, 00:05:22.541 --> 00:05:23.250 some apparel. 00:05:23.583 --> 00:05:25.208 And I also have 00:05:25.583 --> 00:05:27.208 another business that I have, 00:05:27.208 --> 00:05:28.833 which is called Ecom Collab Club, 00:05:29.125 --> 00:05:29.791 which Arvind has 00:05:29.791 --> 00:05:31.375 supported us up for a long time. 00:05:31.875 --> 00:05:33.416 We run events in London once a 00:05:33.416 --> 00:05:34.958 month, about 180 people. 00:05:35.250 --> 00:05:36.625 And we've run those events all 00:05:36.625 --> 00:05:37.333 across the world, 00:05:37.333 --> 00:05:39.791 including New York, Sweden, 00:05:39.875 --> 00:05:41.625 and a few other countries as well. 00:05:41.625 --> 00:05:43.875 And yeah, great to be here back in 00:05:43.875 --> 00:05:45.416 Vietnam after 15 years 00:05:45.416 --> 00:05:46.541 and be with you guys. 00:05:46.875 --> 00:05:47.875 Thank you so much, 00:05:47.916 --> 00:05:48.833 coming back along. 00:05:48.833 --> 00:05:50.458 Like we said, we 00:05:50.458 --> 00:05:51.666 genuinely have a panel, 00:05:51.666 --> 00:05:52.291 which is coming 00:05:52.291 --> 00:05:53.333 from across the globe. 00:05:53.333 --> 00:05:54.500 So thank you so much, Gens. 00:05:55.000 --> 00:05:55.833 The purpose of this 00:05:55.875 --> 00:05:57.583 panel here is to try and see. 00:05:57.833 --> 00:05:59.166 We spoke about BFCM Playbook. 00:05:59.208 --> 00:06:00.291 We started the note on that. 00:06:00.333 --> 00:06:02.125 Then we went to Agente Commerce, 00:06:02.166 --> 00:06:02.916 what's happening. 00:06:03.208 --> 00:06:04.583 We got a platform point of view. 00:06:04.583 --> 00:06:05.291 We got a frontier 00:06:05.291 --> 00:06:06.083 model point of view. 00:06:06.291 --> 00:06:07.375 Here we have 00:06:07.541 --> 00:06:08.666 experts sitting over here. 00:06:08.708 --> 00:06:12.125 The idea is to see that if all the 00:06:12.125 --> 00:06:12.833 models are same, 00:06:13.291 --> 00:06:15.083 if everyone types the same prompt, 00:06:15.208 --> 00:06:16.083 if everyone tries to 00:06:16.083 --> 00:06:17.458 sell similar products, 00:06:17.791 --> 00:06:18.791 how do you differentiate? 00:06:18.833 --> 00:06:19.416 That's something that 00:06:19.416 --> 00:06:20.291 we want to touch upon. 00:06:20.541 --> 00:06:21.583 We have the full 00:06:21.583 --> 00:06:22.708 funnel and try to see 00:06:22.708 --> 00:06:24.041 how do you kind of stand out in 00:06:24.083 --> 00:06:25.083 each part of the funnel. 00:06:25.333 --> 00:06:26.541 And I want to get 00:06:26.541 --> 00:06:27.458 started with you, Adam. 00:06:27.583 --> 00:06:29.458 I think, how is AI 00:06:29.458 --> 00:06:30.958 changing CRO today? 00:06:31.375 --> 00:06:31.916 Would love for you 00:06:31.916 --> 00:06:32.875 to kickstart the ball. 00:06:33.208 --> 00:06:33.708 Sure. 00:06:34.041 --> 00:06:35.500 I think in some 00:06:35.500 --> 00:06:37.333 respects, what I would say here 00:06:37.333 --> 00:06:39.458 is that it is speeding things up, 00:06:39.750 --> 00:06:41.041 but at the same time having to 00:06:41.041 --> 00:06:41.958 slow things down. 00:06:42.458 --> 00:06:43.541 And what I mean by 00:06:43.541 --> 00:06:44.583 that is that, yeah, 00:06:44.625 --> 00:06:47.666 from a CRO point of view, CRO, 00:06:47.666 --> 00:06:48.708 like you can see here, 00:06:48.750 --> 00:06:49.833 works across the 00:06:49.833 --> 00:06:50.666 whole of the funnel. 00:06:51.125 --> 00:06:52.291 And what it does with 00:06:52.291 --> 00:06:53.875 AI is that it speeds up 00:06:53.875 --> 00:06:54.583 on the development 00:06:54.583 --> 00:06:56.291 front because it means 00:06:56.291 --> 00:06:57.250 that you can build quicker, 00:06:57.250 --> 00:06:58.000 like probably everyone 00:06:58.000 --> 00:06:59.041 in the room is doing. 00:06:59.541 --> 00:07:00.333 But at the same 00:07:00.333 --> 00:07:01.791 time, what you have to do 00:07:01.875 --> 00:07:02.875 is make sure that 00:07:02.875 --> 00:07:03.916 from the quality, 00:07:03.916 --> 00:07:04.625 assurance quality 00:07:04.666 --> 00:07:05.791 control point of view 00:07:06.208 --> 00:07:07.291 is that that's where the time 00:07:07.291 --> 00:07:08.416 actually is spent. 00:07:08.708 --> 00:07:12.666 So it is increasing the speed of 00:07:13.250 --> 00:07:14.583 conversion re-optimization, 00:07:15.083 --> 00:07:17.250 but maybe not as fast as we might 00:07:17.250 --> 00:07:18.750 want to anticipate it to be. 00:07:19.791 --> 00:07:21.750 When you say one connected funnel, 00:07:21.791 --> 00:07:23.000 every stage can improve. 00:07:23.458 --> 00:07:24.541 The question is, like 00:07:24.541 --> 00:07:25.791 you said, 15 years back, 00:07:25.791 --> 00:07:27.250 you were in Vietnam, and I could 00:07:27.250 --> 00:07:28.166 say a little longer 00:07:28.166 --> 00:07:29.583 than that, you've been doing CRO. 00:07:29.875 --> 00:07:32.000 What's changed in today's D&H? 00:07:32.500 --> 00:07:33.416 Well, we had a bit of 00:07:33.416 --> 00:07:34.500 a wake up call really 00:07:34.500 --> 00:07:36.625 about 18 months ago where we said, 00:07:36.625 --> 00:07:38.250 look, we need to seriously start 00:07:38.250 --> 00:07:39.208 thinking about AI 00:07:39.625 --> 00:07:40.791 because we were doing 00:07:40.791 --> 00:07:41.791 what I think a lot of us did 00:07:41.791 --> 00:07:43.416 is go, la, la, la, la, la, la, and 00:07:43.416 --> 00:07:44.500 just kind of ignore it. 00:07:44.916 --> 00:07:47.208 And what we did is we sat down 00:07:47.291 --> 00:07:48.250 effectively as a team 00:07:48.250 --> 00:07:50.375 and said, right, what are all of 00:07:50.375 --> 00:07:51.416 the really boring, 00:07:51.541 --> 00:07:52.583 really mundane, 00:07:52.750 --> 00:07:54.708 really admin heavy things 00:07:54.958 --> 00:07:56.875 that our clients wouldn't really 00:07:56.875 --> 00:07:57.583 want to pay for? 00:07:58.125 --> 00:07:58.916 And that's kind of 00:07:58.916 --> 00:07:59.541 where we started. 00:07:59.583 --> 00:08:00.541 Now, we've seen some 00:08:00.541 --> 00:08:01.791 great examples today 00:08:01.791 --> 00:08:02.791 of some really cool 00:08:02.791 --> 00:08:04.500 looking, sexy looking stuff 00:08:04.500 --> 00:08:05.333 on the screen that is 00:08:05.333 --> 00:08:06.375 very much front end. 00:08:06.625 --> 00:08:07.541 We didn't start with that. 00:08:07.583 --> 00:08:08.666 We started with a really 00:08:08.666 --> 00:08:10.500 boring, really mundane, 00:08:10.791 --> 00:08:12.541 but that's kind of helped us and 00:08:12.541 --> 00:08:13.291 get from a process 00:08:13.291 --> 00:08:15.625 that was very, very manual to now 00:08:15.625 --> 00:08:17.250 being driven by AI 00:08:17.500 --> 00:08:19.291 so that actually we sped up. 00:08:19.708 --> 00:08:21.375 And equally is, I think from a 00:08:21.375 --> 00:08:22.541 quality point of view, 00:08:22.791 --> 00:08:24.083 it takes out that 00:08:24.958 --> 00:08:25.833 the human mistakes 00:08:25.833 --> 00:08:27.125 that obviously are gonna be there 00:08:28.000 --> 00:08:29.083 when you are not doing 00:08:29.125 --> 00:08:30.291 it in a kind of process 00:08:30.333 --> 00:08:31.666 or AI driven way. 00:08:32.041 --> 00:08:32.666 That's a pretty 00:08:32.708 --> 00:08:33.958 personal wake up call 00:08:33.958 --> 00:08:34.541 and kind of makes 00:08:34.541 --> 00:08:35.666 sense to see that, 00:08:36.125 --> 00:08:37.500 how your customer reacted 00:08:37.583 --> 00:08:39.375 and how you proactively corrected, 00:08:39.375 --> 00:08:40.166 cost corrected that. 00:08:40.833 --> 00:08:41.916 I have the same question to you, 00:08:41.916 --> 00:08:44.000 Louis and Stefan as well. 00:08:44.166 --> 00:08:44.541 Both of you have 00:08:44.541 --> 00:08:45.416 been running your agency 00:08:45.416 --> 00:08:46.333 for close to a decade. 00:08:47.000 --> 00:08:48.750 The way in which we operate, 00:08:48.791 --> 00:08:50.333 the way in which we run the show 00:08:50.458 --> 00:08:51.208 has kind of changed. 00:08:51.416 --> 00:08:52.666 What was that wake up call 00:08:52.708 --> 00:08:53.541 and when did that wake 00:08:53.541 --> 00:08:54.541 up call happen to you 00:08:54.625 --> 00:08:55.458 starting with Louis? 00:08:56.541 --> 00:08:58.458 I think Adam is very right. 00:08:58.666 --> 00:09:00.958 The landscape has 00:09:00.958 --> 00:09:02.208 changed a lot for us 00:09:02.250 --> 00:09:04.000 in the last one year. 00:09:04.500 --> 00:09:06.291 What clients are, 00:09:06.750 --> 00:09:07.166 what they kind of 00:09:07.166 --> 00:09:08.416 value has changed a lot. 00:09:09.291 --> 00:09:10.250 Creative production, 00:09:10.291 --> 00:09:11.250 which is kind of our bread and 00:09:11.250 --> 00:09:12.333 butter as agencies. 00:09:13.333 --> 00:09:14.000 They kind of just want 00:09:14.000 --> 00:09:14.958 an extra pair of hands, 00:09:14.958 --> 00:09:16.458 they need an extra pair of brains. 00:09:16.708 --> 00:09:17.250 The pair of hands 00:09:17.250 --> 00:09:18.375 doesn't really matter anymore. 00:09:20.416 --> 00:09:22.583 We internally, this is what we've 00:09:22.583 --> 00:09:23.458 done for our clients. 00:09:23.791 --> 00:09:24.916 We do email and SMS, 00:09:24.958 --> 00:09:27.750 mainly as our business. 00:09:28.083 --> 00:09:28.791 Internally our 00:09:28.791 --> 00:09:30.458 designers and copywriters, 00:09:31.083 --> 00:09:32.250 some of you I told them, 00:09:33.208 --> 00:09:34.875 all of you got a three X output 00:09:34.916 --> 00:09:35.833 and I cannot pay you more. 00:09:36.500 --> 00:09:37.291 That's the reality 00:09:37.333 --> 00:09:38.583 because clients are expecting 00:09:38.666 --> 00:09:39.416 exactly the same. 00:09:40.250 --> 00:09:41.791 Expecting three times faster, 00:09:42.375 --> 00:09:44.250 while we charge our clients, 00:09:44.583 --> 00:09:45.625 I used to be able to charge like, 00:09:45.625 --> 00:09:46.583 "Hey, 15 emails 00:09:46.583 --> 00:09:47.708 for $5,000 a month. 00:09:48.083 --> 00:09:49.083 Today it's like 50 00:09:49.083 --> 00:09:50.208 emails for $5,000." 00:09:50.541 --> 00:09:51.625 That's kind of the reality. 00:09:52.083 --> 00:09:53.250 And I think Namara mentioned 00:09:53.250 --> 00:09:53.958 earlier on as well, 00:09:53.958 --> 00:09:54.625 which is they're kind 00:09:54.625 --> 00:09:55.583 of paying for results. 00:09:55.583 --> 00:09:57.083 They don't really care about 00:09:57.500 --> 00:09:58.625 the quantity, it just doesn't 00:09:58.833 --> 00:10:00.000 matter that much anymore. 00:10:00.000 --> 00:10:03.000 So internally, that's how we had 00:10:03.000 --> 00:10:04.000 to align our team, 00:10:04.041 --> 00:10:05.250 which is, "Hey guys, 00:10:05.333 --> 00:10:06.458 this is kind of the reality. 00:10:06.541 --> 00:10:07.375 If we don't reinvent 00:10:07.375 --> 00:10:09.166 ourselves, that's it." 00:10:09.458 --> 00:10:10.708 Fall behind, if not. 00:10:10.750 --> 00:10:12.208 No, that's a great, great thing. 00:10:12.250 --> 00:10:14.875 Stefan, how are things back in EU 00:10:14.875 --> 00:10:15.750 and across the globe? 00:10:16.125 --> 00:10:17.333 A, I want to ask, how 00:10:17.333 --> 00:10:18.583 did you internalize it? 00:10:18.583 --> 00:10:19.333 And B, generally, 00:10:19.333 --> 00:10:20.041 how are you seeing this 00:10:20.041 --> 00:10:21.083 as a pattern across? 00:10:21.083 --> 00:10:21.541 Because you give a 00:10:21.541 --> 00:10:22.416 lot of conferences, 00:10:22.625 --> 00:10:23.291 you go, you're a 00:10:23.291 --> 00:10:24.416 keynote speaker as well. 00:10:24.416 --> 00:10:25.541 Want to take your perspective. 00:10:26.625 --> 00:10:29.458 Yeah, so your question was, 00:10:29.875 --> 00:10:32.916 what was the Eureka moment for us 00:10:32.916 --> 00:10:34.041 as an organization 00:10:34.458 --> 00:10:36.041 in order to actually use AI? 00:10:37.208 --> 00:10:39.041 Differently than everybody else. 00:10:39.083 --> 00:10:40.500 And it doesn't have 00:10:40.500 --> 00:10:41.416 anything to do with like 00:10:41.458 --> 00:10:42.875 email marketing or 00:10:43.250 --> 00:10:45.333 CRO or SEO or GEO. 00:10:45.708 --> 00:10:47.000 For us and for 00:10:47.041 --> 00:10:48.125 everybody here in the room, 00:10:48.416 --> 00:10:49.250 whether you're an app 00:10:49.250 --> 00:10:51.083 developer, SaaS business, 00:10:51.083 --> 00:10:53.208 merchant, brand, Shopify even, 00:10:54.208 --> 00:10:56.208 for us it was organizational DNA. 00:10:56.958 --> 00:10:59.666 The context that goes into the 00:10:59.666 --> 00:11:01.416 actual AI assistants 00:11:01.458 --> 00:11:02.333 that power each and 00:11:02.333 --> 00:11:03.708 every one of our humans, 00:11:03.750 --> 00:11:05.458 our human members of the teams. 00:11:06.041 --> 00:11:07.000 The context about 00:11:07.041 --> 00:11:08.291 our organization, 00:11:08.625 --> 00:11:10.416 when we adjusted that 00:11:10.416 --> 00:11:12.625 and when we actually made 00:11:12.625 --> 00:11:15.000 our assistance aware about our 00:11:15.000 --> 00:11:16.708 company's quarterly goals, 00:11:16.708 --> 00:11:18.416 our company's yearly goals, 00:11:18.583 --> 00:11:21.000 the actual content that came out 00:11:21.000 --> 00:11:22.291 of the assistant changed. 00:11:22.416 --> 00:11:23.833 Because if our goal was to 00:11:23.833 --> 00:11:25.166 increase our profitability, 00:11:25.958 --> 00:11:28.166 then an email prompt would have 00:11:28.166 --> 00:11:29.208 given us an email 00:11:29.250 --> 00:11:30.750 which takes 30 minutes to make. 00:11:31.041 --> 00:11:32.166 But if our goal for 00:11:32.166 --> 00:11:33.416 the next quarter was 00:11:33.416 --> 00:11:36.250 to increase performance, increase 00:11:36.250 --> 00:11:37.333 client retention, 00:11:37.458 --> 00:11:38.041 get better 00:11:38.041 --> 00:11:39.208 results for our clients, 00:11:39.250 --> 00:11:41.500 then the prompt gives us a 00:11:41.541 --> 00:11:43.500 different response 00:11:43.500 --> 00:11:44.500 which is, "A, B, 00:11:44.500 --> 00:11:45.958 split test this campaign," 00:11:45.958 --> 00:11:47.125 or "Do this, do that." 00:11:47.416 --> 00:11:50.416 So, organizational DNA is 00:11:50.416 --> 00:11:51.291 something that everyone 00:11:51.291 --> 00:11:52.250 here in the room can 00:11:52.291 --> 00:11:53.750 do, can define better, 00:11:54.166 --> 00:11:56.041 and it is going to impact and it's 00:11:56.041 --> 00:11:57.208 gonna change the way 00:11:57.666 --> 00:12:00.125 your AIs respond to your teams, 00:12:00.750 --> 00:12:01.541 and it's actually 00:12:01.541 --> 00:12:02.875 gonna make you different 00:12:02.958 --> 00:12:04.500 from everybody else. 00:12:05.125 --> 00:12:06.000 Thanks for the sneak peek. 00:12:06.125 --> 00:12:07.791 I think later in the day, 00:12:07.791 --> 00:12:10.041 Stefan has built it on Showcase, 00:12:10.291 --> 00:12:10.666 where he's gonna 00:12:10.666 --> 00:12:11.666 expand a little bit more 00:12:11.666 --> 00:12:13.625 on that organizational brain, 00:12:13.708 --> 00:12:14.875 or the second brain that we 00:12:15.416 --> 00:12:16.583 commonly refer to as, 00:12:16.625 --> 00:12:17.625 more on that later. 00:12:17.875 --> 00:12:21.000 But coming back to the CRO part of 00:12:21.041 --> 00:12:22.000 the funnel, right? 00:12:22.500 --> 00:12:23.125 I'm sure there's a 00:12:23.125 --> 00:12:24.416 systemic process to it 00:12:24.416 --> 00:12:25.166 on how you're doing. 00:12:25.458 --> 00:12:27.416 For all the fellow agency folks, 00:12:27.500 --> 00:12:28.791 sellers sitting on there, 00:12:29.041 --> 00:12:31.041 what is your process to ensure 00:12:31.041 --> 00:12:31.500 you kind of 00:12:31.541 --> 00:12:32.833 personalize the full funnel? 00:12:33.041 --> 00:12:33.583 Do you wanna walk us 00:12:33.583 --> 00:12:34.500 through your steps? 00:12:34.750 --> 00:12:35.375 Yeah, sure. 00:12:35.375 --> 00:12:36.083 I mean, look, what 00:12:36.083 --> 00:12:37.166 we're looking at here is, 00:12:37.208 --> 00:12:39.000 I guess this is sort of an extent 00:12:39.333 --> 00:12:40.958 the pre-AIA version, 00:12:41.375 --> 00:12:43.375 in that when a Shopify merchant 00:12:43.375 --> 00:12:43.958 will come to us, 00:12:43.958 --> 00:12:45.083 the first thing that we would do 00:12:45.333 --> 00:12:47.000 is what we call a heuristic audit. 00:12:47.041 --> 00:12:49.250 So, i.e. looking through the site 00:12:49.250 --> 00:12:50.333 from top to bottom, 00:12:50.750 --> 00:12:53.041 and coming up with maybe 100 or 00:12:53.041 --> 00:12:55.208 150 ideas or hypotheses. 00:12:55.625 --> 00:12:56.583 Now, once that's done, 00:12:56.625 --> 00:12:57.416 what we're then doing 00:12:57.416 --> 00:12:58.166 is we're mining the 00:12:58.166 --> 00:12:59.541 data from Shopify, 00:13:00.291 --> 00:13:02.500 from GA4, from heat maps, 00:13:02.791 --> 00:13:03.250 and what we're 00:13:03.250 --> 00:13:04.375 trying to do is to say, 00:13:04.375 --> 00:13:05.541 well, some of those 00:13:05.541 --> 00:13:07.291 ideas that we had in step one 00:13:07.583 --> 00:13:09.291 are stronger because we have data 00:13:09.291 --> 00:13:10.208 associated with it. 00:13:10.416 --> 00:13:11.708 Some of them fall away. 00:13:12.125 --> 00:13:12.916 And I'm sure we've 00:13:12.916 --> 00:13:13.666 all been that meeting 00:13:13.666 --> 00:13:15.416 where we think, I think that we 00:13:15.416 --> 00:13:16.375 should do this thing, 00:13:16.416 --> 00:13:18.083 or I have a gut feeling that we 00:13:18.083 --> 00:13:18.750 should do this thing. 00:13:19.083 --> 00:13:20.041 That gets rid of 00:13:20.083 --> 00:13:21.166 some of those ideas. 00:13:21.208 --> 00:13:22.583 And then the final step, which is 00:13:22.583 --> 00:13:23.458 the most important one, 00:13:23.875 --> 00:13:24.708 is Pecte. 00:13:24.708 --> 00:13:26.041 Now, I am sure 00:13:26.041 --> 00:13:27.541 everyone sitting in this room 00:13:27.750 --> 00:13:28.791 has a spreadsheet 00:13:28.916 --> 00:13:30.041 with lots of ideas on. 00:13:30.083 --> 00:13:31.291 You've maybe got emails you've 00:13:31.291 --> 00:13:31.833 written to yourself, 00:13:31.833 --> 00:13:33.291 little voice notes on your phone. 00:13:33.833 --> 00:13:35.583 What Pecte does is allows you to 00:13:35.583 --> 00:13:37.041 prioritize each of those. 00:13:37.041 --> 00:13:37.875 And that's what we do 00:13:37.875 --> 00:13:38.875 with all of these ideas. 00:13:38.875 --> 00:13:41.916 So we have ideas, we have the data 00:13:41.958 --> 00:13:43.458 that backs up those ideas, 00:13:43.708 --> 00:13:45.208 and then we're prioritizing them 00:13:45.208 --> 00:13:46.166 with the scoring system. 00:13:46.791 --> 00:13:48.625 So proof is cost, 00:13:48.875 --> 00:13:49.458 time, and impact. 00:13:49.875 --> 00:13:51.291 And that's a proprietary framework 00:13:51.291 --> 00:13:52.208 that you have over there. 00:13:52.250 --> 00:13:53.541 It's not proprietary, no. 00:13:53.958 --> 00:13:55.666 It is something that is used by a 00:13:55.666 --> 00:13:56.458 number of different people, 00:13:56.458 --> 00:13:56.916 but we have 00:13:56.916 --> 00:13:58.291 personalized it specifically 00:13:58.458 --> 00:13:59.500 for, I guess, the 00:13:59.500 --> 00:14:00.166 e-commerce world. 00:14:00.625 --> 00:14:01.166 And this just kind 00:14:01.166 --> 00:14:01.916 of breaks it down. 00:14:01.916 --> 00:14:02.916 I think, look, I would recommend, 00:14:03.416 --> 00:14:04.333 it doesn't matter whether you're 00:14:04.333 --> 00:14:05.458 doing CRO or anything, 00:14:05.458 --> 00:14:06.541 realistically, in e-commerce. 00:14:07.250 --> 00:14:09.166 There are always tons of ideas. 00:14:10.041 --> 00:14:11.166 And breaking this 00:14:11.166 --> 00:14:11.833 down affects you. 00:14:11.833 --> 00:14:12.458 What you do is that 00:14:12.458 --> 00:14:13.500 every idea that you have, 00:14:13.666 --> 00:14:15.916 you score it out of five in terms 00:14:15.916 --> 00:14:16.708 of each of these, 00:14:16.916 --> 00:14:18.333 and then you roll that up into an 00:14:18.333 --> 00:14:19.291 index out of 100. 00:14:19.833 --> 00:14:20.666 Now, the beauty of 00:14:20.666 --> 00:14:21.875 doing this is that, 00:14:22.291 --> 00:14:23.791 for example, if you did this today 00:14:23.833 --> 00:14:24.625 for maybe a brand 00:14:24.625 --> 00:14:25.375 that is in the room, 00:14:26.208 --> 00:14:27.458 that would be the situation today. 00:14:27.500 --> 00:14:28.916 But as more data comes in, 00:14:28.916 --> 00:14:30.083 maybe over the Black Friday 00:14:30.083 --> 00:14:31.541 sub-minute period, 00:14:31.833 --> 00:14:32.958 the data will shift, 00:14:33.208 --> 00:14:34.750 and therefore the prioritization 00:14:34.750 --> 00:14:35.541 of these will change. 00:14:35.833 --> 00:14:36.416 And this is the 00:14:36.458 --> 00:14:37.500 difficult thing, I think, 00:14:37.583 --> 00:14:38.708 within e-commerce, isn't it? 00:14:38.750 --> 00:14:40.041 Is that we always 00:14:40.041 --> 00:14:41.583 take a snapshot of now, 00:14:41.875 --> 00:14:42.708 but as we've heard 00:14:42.708 --> 00:14:43.458 from this morning, 00:14:43.458 --> 00:14:45.166 it is always something that's 00:14:45.166 --> 00:14:46.250 dynamic and always moving. 00:14:46.541 --> 00:14:48.416 So if you use a process like this, 00:14:48.416 --> 00:14:50.041 whichever version that you use, 00:14:50.625 --> 00:14:51.458 it just means that 00:14:51.500 --> 00:14:53.541 you are maintaining, 00:14:53.541 --> 00:14:55.166 then, changes to the site that are 00:14:55.583 --> 00:14:56.791 right here, right now, 00:14:56.833 --> 00:14:58.041 based on the data of that 00:14:58.041 --> 00:14:58.750 particular moment. 00:14:58.916 --> 00:14:59.666 That makes a lot of 00:14:59.666 --> 00:15:00.916 sense to have a framework, 00:15:00.916 --> 00:15:02.125 because right before this, 00:15:02.333 --> 00:15:03.000 Alex and I, we were 00:15:03.000 --> 00:15:04.333 having a conversation on 00:15:04.375 --> 00:15:05.583 how is the output in 00:15:05.583 --> 00:15:07.166 Dubai, SF, and India, 00:15:07.166 --> 00:15:08.250 we were just talking about. 00:15:09.083 --> 00:15:10.000 A good output 00:15:10.041 --> 00:15:10.916 would actually stem out 00:15:10.916 --> 00:15:12.166 from a really good prioritization. 00:15:12.416 --> 00:15:13.458 Makes a lot of sense. 00:15:13.666 --> 00:15:14.875 What's this fancy 00:15:14.875 --> 00:15:15.833 dashboard that you have, 00:15:15.916 --> 00:15:17.500 where you use it for a customer? 00:15:17.583 --> 00:15:19.500 So this is our Pecti board. 00:15:19.583 --> 00:15:20.833 So if you imagine, we've gone 00:15:20.833 --> 00:15:21.750 through that process, 00:15:21.750 --> 00:15:22.291 we've gone from 00:15:22.291 --> 00:15:23.625 100 different ideas, 00:15:23.875 --> 00:15:25.833 down to maybe 40 or 50, 00:15:26.250 --> 00:15:27.541 either A-B tests or 00:15:27.541 --> 00:15:29.083 implementations onto the site. 00:15:29.833 --> 00:15:30.958 What this does is 00:15:30.958 --> 00:15:32.125 that we have the scores 00:15:32.125 --> 00:15:33.083 in terms of the stars, 00:15:33.083 --> 00:15:34.291 in terms of proof fees, cost, 00:15:34.291 --> 00:15:35.000 time, and impact, 00:15:35.250 --> 00:15:36.708 and then the score out of 100. 00:15:37.333 --> 00:15:39.000 So as you'd imagine, 00:15:39.000 --> 00:15:40.166 you start in terms of 00:15:40.166 --> 00:15:41.291 implementing the top 00:15:41.291 --> 00:15:43.250 with items that score 100, 00:15:43.708 --> 00:15:44.791 and then you go down that 00:15:44.791 --> 00:15:46.166 level of prioritization. 00:15:46.458 --> 00:15:47.541 So this just means it keeps 00:15:47.625 --> 00:15:48.750 everyone focused, 00:15:49.000 --> 00:15:50.041 and it also stops 00:15:50.041 --> 00:15:51.000 those conversations 00:15:51.000 --> 00:15:53.166 where maybe the CMO, head of 00:15:53.166 --> 00:15:54.291 brand, the founder, 00:15:54.625 --> 00:15:55.416 are saying, "No, I 00:15:55.416 --> 00:15:56.458 think we need bundles." 00:15:56.458 --> 00:15:58.250 No, I think we need landing pages. 00:15:58.500 --> 00:16:00.000 This does, it kind of acts as, 00:16:00.208 --> 00:16:01.500 I always say acts as a little bit 00:16:01.500 --> 00:16:03.166 like a UMPs keeper, 00:16:03.500 --> 00:16:04.250 and it comes and says, 00:16:04.291 --> 00:16:05.250 "Here is the data to 00:16:05.250 --> 00:16:06.833 say what the situation is, 00:16:07.041 --> 00:16:07.791 "and here is the 00:16:07.791 --> 00:16:09.083 roadmap that we then have 00:16:09.083 --> 00:16:10.250 "to basically put into play." 00:16:10.583 --> 00:16:11.875 And obviously with AI now, 00:16:12.208 --> 00:16:14.083 this whole process I've described 00:16:14.250 --> 00:16:15.083 has become a lot 00:16:15.083 --> 00:16:16.750 faster, a lot more efficient. 00:16:17.166 --> 00:16:17.791 By the way, that's 00:16:17.791 --> 00:16:19.125 an actual screenshot 00:16:19.125 --> 00:16:21.416 from one of his brands that he's 00:16:21.416 --> 00:16:22.291 actually testing out. 00:16:22.541 --> 00:16:23.416 Everybody can actually 00:16:23.416 --> 00:16:24.875 connect, get an output, 00:16:24.916 --> 00:16:25.791 get the framework, 00:16:26.083 --> 00:16:27.458 but you built a 00:16:27.458 --> 00:16:28.583 proprietary dashboard 00:16:28.916 --> 00:16:30.750 to kind of further streamline it 00:16:30.750 --> 00:16:31.875 to try and see how it works. 00:16:32.041 --> 00:16:33.291 Can you walk us through on, 00:16:33.541 --> 00:16:34.416 you have your idea, 00:16:34.458 --> 00:16:35.375 you have your framework, 00:16:35.541 --> 00:16:36.583 how do you distill it? 00:16:36.625 --> 00:16:37.708 How do you give it to customers? 00:16:37.708 --> 00:16:38.583 Can you tell us a little bit? 00:16:38.875 --> 00:16:39.875 Sure, so we built 00:16:39.875 --> 00:16:40.458 something in the 10, 00:16:40.458 --> 00:16:41.291 they call Nugget Lab, 00:16:41.458 --> 00:16:42.666 and what Nugget Lab 00:16:42.708 --> 00:16:44.291 does is that it plugs in 00:16:44.333 --> 00:16:46.000 all of the different data sources 00:16:46.000 --> 00:16:47.083 that we talked about there, 00:16:47.416 --> 00:16:48.708 and this screenshot 00:16:48.708 --> 00:16:49.625 here is just basically 00:16:49.625 --> 00:16:50.875 the dashboard that happens. 00:16:50.875 --> 00:16:52.250 Now, I'm gonna say 00:16:52.250 --> 00:16:53.416 something controversial. 00:16:53.666 --> 00:16:55.125 I don't like conversion rate. 00:16:56.125 --> 00:16:57.375 And the reason I 00:16:57.375 --> 00:16:58.458 say that is because 00:16:58.500 --> 00:16:59.375 I feel that 00:16:59.666 --> 00:17:01.250 conversion rate is a metric 00:17:01.250 --> 00:17:02.458 that you can hide behind. 00:17:03.083 --> 00:17:03.916 The metric I prefer 00:17:03.916 --> 00:17:04.833 is revenue possession, 00:17:05.250 --> 00:17:06.541 because for me, what revenue 00:17:06.541 --> 00:17:07.666 possession says is 00:17:07.875 --> 00:17:10.125 how efficient is this website 00:17:10.458 --> 00:17:12.208 at bringing cash into the bank 00:17:12.208 --> 00:17:13.208 account of that business? 00:17:13.625 --> 00:17:15.250 And what this does is that it's 00:17:15.250 --> 00:17:16.041 basically tracking, 00:17:16.041 --> 00:17:17.916 number one, the metrics, 00:17:18.041 --> 00:17:18.750 but importantly 00:17:18.750 --> 00:17:19.583 what it does is that 00:17:19.583 --> 00:17:21.125 if we go onto that next slide, 00:17:21.458 --> 00:17:22.291 we'll see here is 00:17:22.291 --> 00:17:23.458 that once obviously 00:17:23.500 --> 00:17:24.125 each of those 00:17:24.125 --> 00:17:25.291 connectors are brought in 00:17:25.291 --> 00:17:26.666 to the different data sources, 00:17:26.958 --> 00:17:27.875 it is then making the 00:17:27.875 --> 00:17:28.916 recommendations. 00:17:29.166 --> 00:17:29.541 Once the 00:17:29.541 --> 00:17:31.041 recommendations are then done, 00:17:31.291 --> 00:17:33.666 the prioritization of Pecte is 00:17:33.666 --> 00:17:34.791 also then done too. 00:17:35.166 --> 00:17:37.041 Now, this is fed in from about 00:17:37.041 --> 00:17:38.125 three and a half thousand 00:17:38.125 --> 00:17:38.875 different A-B tests 00:17:38.875 --> 00:17:39.875 that we've done. Indeed. 00:17:39.916 --> 00:17:40.958 And off the back of that then, 00:17:41.000 --> 00:17:42.291 it creates the prioritization. 00:17:42.291 --> 00:17:42.916 So you can see 00:17:42.958 --> 00:17:44.250 probably very clearly here, 00:17:44.583 --> 00:17:46.125 the work that my team used to do, 00:17:46.125 --> 00:17:47.166 which is very manual, 00:17:47.791 --> 00:17:48.541 or kind of being 00:17:48.541 --> 00:17:49.708 done with data analysis, 00:17:50.083 --> 00:17:51.583 is now being done, you know, 00:17:51.875 --> 00:17:53.125 relatively in a 00:17:53.125 --> 00:17:54.333 short period of time 00:17:54.541 --> 00:17:56.041 compared to what it used to take, 00:17:56.041 --> 00:17:56.708 which was, you know, 00:17:56.708 --> 00:17:57.708 weeks and weeks of work. 00:17:58.041 --> 00:17:59.291 Is it okay for me to say that 00:17:59.291 --> 00:18:00.708 while you're sitting in Saigon, 00:18:00.916 --> 00:18:02.875 there is an adum which people are 00:18:02.875 --> 00:18:04.541 able to chat in London 00:18:04.625 --> 00:18:05.416 through the dashboard, 00:18:05.416 --> 00:18:06.375 is it okay to say that? 00:18:06.541 --> 00:18:07.333 Yeah, absolutely. 00:18:07.458 --> 00:18:08.208 No, I think that's 00:18:08.416 --> 00:18:09.333 pretty interesting. 00:18:09.666 --> 00:18:10.666 Come to think of it, right? 00:18:10.666 --> 00:18:11.875 I'm sure I'm gonna preempt a 00:18:11.875 --> 00:18:13.208 question from the audience. 00:18:13.416 --> 00:18:14.458 I could do this in Claude, 00:18:14.666 --> 00:18:15.541 or I could do this 00:18:15.583 --> 00:18:17.250 probably with a new model 00:18:17.791 --> 00:18:19.041 with OpenAI is gonna 00:18:19.041 --> 00:18:19.750 release, by the way. 00:18:19.750 --> 00:18:21.000 They're gonna announce that in the 00:18:21.000 --> 00:18:22.000 second half of the day. 00:18:22.791 --> 00:18:24.708 How do you, let's say, blend in 00:18:24.708 --> 00:18:26.541 the human and the AI to it? 00:18:26.541 --> 00:18:27.500 How do you ensure 00:18:27.500 --> 00:18:28.583 that, you know what, 00:18:28.666 --> 00:18:30.541 there's a objective thing, 00:18:30.791 --> 00:18:31.916 and there's a subjective thing. 00:18:32.125 --> 00:18:33.875 How do you blend in to ensure 00:18:33.875 --> 00:18:35.166 that your relationship is made in? 00:18:35.583 --> 00:18:36.291 I think that the 00:18:36.291 --> 00:18:37.333 important thing for us is, 00:18:37.333 --> 00:18:38.333 like, when we did 00:18:38.333 --> 00:18:39.791 this, is that we used 00:18:39.875 --> 00:18:41.458 our sort of 11 years of experience 00:18:41.458 --> 00:18:42.833 and put that into the model. 00:18:43.250 --> 00:18:44.791 So, yes, absolutely. 00:18:44.791 --> 00:18:46.041 And I think, look, it depends on 00:18:46.041 --> 00:18:47.125 what kind of size brand 00:18:47.125 --> 00:18:47.958 you are, because when 00:18:47.958 --> 00:18:48.833 you're starting out, 00:18:49.500 --> 00:18:50.416 you're not gonna be able to afford 00:18:50.416 --> 00:18:51.291 to work with an agency. 00:18:51.291 --> 00:18:52.166 And also, I don't 00:18:52.166 --> 00:18:52.791 think you should, 00:18:53.375 --> 00:18:54.833 realistically for me, unless a 00:18:54.833 --> 00:18:56.041 brand is maybe doing 00:18:56.041 --> 00:18:57.166 at least 25, 00:18:57.458 --> 00:18:58.833 30,000 visits a month, 00:18:59.083 --> 00:18:59.791 I don't think you should be 00:18:59.791 --> 00:19:00.791 working with anyone like me. 00:19:01.916 --> 00:19:02.916 So, you can use 00:19:02.916 --> 00:19:04.125 that, your claw models, 00:19:04.125 --> 00:19:04.708 you can use, 00:19:04.791 --> 00:19:07.833 obviously, any of those models. 00:19:08.125 --> 00:19:08.791 But as you don't 00:19:08.791 --> 00:19:09.833 get to a bigger stage, 00:19:09.833 --> 00:19:11.125 it's then good to kind of get some 00:19:11.125 --> 00:19:12.375 real world examples 00:19:12.416 --> 00:19:13.250 of A, B sets that 00:19:13.250 --> 00:19:14.083 have actually been run 00:19:14.375 --> 00:19:15.750 with similar brands to yourself. 00:19:16.291 --> 00:19:17.791 Speaking of real world examples, 00:19:18.208 --> 00:19:19.333 I have spoken to a 00:19:19.333 --> 00:19:20.375 couple of your customers 00:19:20.375 --> 00:19:21.583 while I was there at your event, 00:19:21.625 --> 00:19:22.791 when I was touring and stuff. 00:19:23.000 --> 00:19:25.208 I wanna understand how did this 00:19:25.208 --> 00:19:26.333 help move the needle? 00:19:26.458 --> 00:19:27.250 Do you wanna talk 00:19:27.250 --> 00:19:28.500 about Stone Creek Coffee? 00:19:28.583 --> 00:19:29.666 Yeah, sure. 00:19:29.666 --> 00:19:31.708 So, nine figure coffee brand, two 00:19:31.750 --> 00:19:33.125 brands across the US, 00:19:33.166 --> 00:19:34.250 as well as a string of 00:19:34.250 --> 00:19:35.333 coffee shops as well. 00:19:36.625 --> 00:19:38.541 The beauty, obviously, now with 00:19:38.541 --> 00:19:39.125 the kind of being able 00:19:39.166 --> 00:19:40.208 to use AI is that the 00:19:40.208 --> 00:19:41.708 kind of test that we can run 00:19:41.708 --> 00:19:43.750 and discovering those is a lot 00:19:43.750 --> 00:19:44.833 quicker than it was before. 00:19:45.750 --> 00:19:46.708 It's not quicker in 00:19:46.708 --> 00:19:47.375 the sense, obviously, 00:19:47.375 --> 00:19:48.166 we still have to 00:19:48.166 --> 00:19:49.166 wait for the traffic 00:19:49.166 --> 00:19:50.000 because we need to get 00:19:50.000 --> 00:19:51.375 statistical significance. 00:19:51.750 --> 00:19:52.333 But I think what it 00:19:52.333 --> 00:19:53.583 shows on the next slide here 00:19:53.583 --> 00:19:54.875 is that basically, when we 00:19:54.875 --> 00:19:55.958 inherited the site, 00:19:56.625 --> 00:19:57.500 there was an issue 00:19:57.500 --> 00:19:58.708 was picked up by AI. 00:19:58.708 --> 00:20:00.375 And that was, is that on the 00:20:00.375 --> 00:20:02.125 subscription page on the PDP, 00:20:02.250 --> 00:20:02.916 the subscription 00:20:02.916 --> 00:20:03.833 choices were hidden. 00:20:03.833 --> 00:20:05.333 So what you saw is a dropdown. 00:20:05.833 --> 00:20:07.375 Now, what we were able to say is 00:20:07.375 --> 00:20:08.500 all very quickly 00:20:08.583 --> 00:20:10.416 is identify that this is an issue. 00:20:10.416 --> 00:20:12.625 We can see friction, not only in 00:20:12.625 --> 00:20:14.416 terms of the GA4 data, 00:20:14.416 --> 00:20:15.833 but also the heat map data. 00:20:16.166 --> 00:20:16.958 And then when we go on 00:20:16.958 --> 00:20:18.458 to the next version of it, 00:20:18.458 --> 00:20:20.416 the A-B test, we simplified that 00:20:20.416 --> 00:20:21.416 to be different 00:20:21.416 --> 00:20:22.416 subscription options. 00:20:22.416 --> 00:20:24.833 So much more visual, the 00:20:24.833 --> 00:20:26.291 build of this particular 00:20:26.333 --> 00:20:27.500 front-end part as well 00:20:27.500 --> 00:20:28.750 was again done with AI. 00:20:29.166 --> 00:20:30.125 And then as a result of 00:20:30.166 --> 00:20:31.375 this, what we get is that 00:20:31.375 --> 00:20:33.583 from this very specific change, 00:20:33.875 --> 00:20:35.541 the attribution to that change 00:20:35.541 --> 00:20:36.666 then actually leverages 00:20:36.916 --> 00:20:39.000 about another $44,000 a year. 00:20:39.208 --> 00:20:40.541 So just actually to that 00:20:40.541 --> 00:20:41.791 particular change itself 00:20:42.041 --> 00:20:42.916 on top of what was 00:20:42.916 --> 00:20:43.666 there previously. 00:20:43.833 --> 00:20:44.708 That's insane. 00:20:45.208 --> 00:20:46.666 It's a probably classic case of 00:20:46.666 --> 00:20:48.000 that Pareto principle, right? 00:20:48.166 --> 00:20:50.375 20% gives you that outsized thing 00:20:50.375 --> 00:20:51.625 and having those two changes. 00:20:51.833 --> 00:20:53.208 If it gives that big a delta, 00:20:53.666 --> 00:20:54.500 then we start a brand, 00:20:54.500 --> 00:20:54.833 they're coming to 00:20:54.833 --> 00:20:55.833 you for the playbook. 00:20:57.500 --> 00:20:59.625 Well, how do we know that this is 00:20:59.625 --> 00:21:00.916 not a one-hit wonder 00:21:01.041 --> 00:21:01.666 kind of a model? 00:21:02.875 --> 00:21:04.458 Sure, like it works for one. 00:21:04.708 --> 00:21:06.291 How do we know that it doesn't, 00:21:07.000 --> 00:21:08.583 or it does work for the other? 00:21:08.875 --> 00:21:09.458 What are your thoughts 00:21:09.500 --> 00:21:10.208 with the second case? 00:21:10.250 --> 00:21:12.041 Yeah, so this is a UK-based brand, 00:21:12.291 --> 00:21:13.958 eight figure business selling 00:21:13.958 --> 00:21:14.916 barbecue equipment 00:21:14.916 --> 00:21:15.833 and such like. 00:21:16.708 --> 00:21:18.083 Again, with these guys, the 00:21:18.083 --> 00:21:19.708 situation was similar 00:21:19.708 --> 00:21:21.416 in that they needed them to have 00:21:21.416 --> 00:21:22.500 first of all the analysis 00:21:22.541 --> 00:21:23.666 and check in what was going on. 00:21:24.083 --> 00:21:25.666 And what we saw is that very 00:21:25.666 --> 00:21:27.208 different behaviors 00:21:27.416 --> 00:21:29.458 by forms of payment. 00:21:29.500 --> 00:21:30.541 And payment 00:21:30.541 --> 00:21:32.125 methods obviously in Asia 00:21:32.125 --> 00:21:33.375 is a very important 00:21:33.375 --> 00:21:34.500 thing for us to consider. 00:21:34.875 --> 00:21:36.333 And wherever you are selling 00:21:36.333 --> 00:21:36.958 across the world, 00:21:36.958 --> 00:21:37.625 something important. 00:21:37.625 --> 00:21:38.458 Now what these guys 00:21:38.458 --> 00:21:39.916 did is that they led 00:21:39.916 --> 00:21:41.208 with the buy with shop. 00:21:41.958 --> 00:21:42.958 And what we saw is 00:21:42.958 --> 00:21:44.458 that the lifetime value, 00:21:44.458 --> 00:21:45.416 and we see this a lot, 00:21:45.416 --> 00:21:46.750 the lifetime value of 00:21:46.750 --> 00:21:47.750 different payment methods 00:21:47.750 --> 00:21:48.833 is always very, very different. 00:21:49.291 --> 00:21:50.875 Now what we did with these guys is 00:21:50.875 --> 00:21:51.666 that we actually then 00:21:51.708 --> 00:21:53.875 said actually we need to expand 00:21:53.958 --> 00:21:54.875 those payment options 00:21:54.875 --> 00:21:55.416 that are shown. 00:21:55.791 --> 00:21:56.791 We made this simple 00:21:56.791 --> 00:21:58.125 change on the site, 00:21:58.291 --> 00:21:59.875 which again was assisted by the 00:21:59.875 --> 00:22:01.041 whole AI process, 00:22:01.291 --> 00:22:02.125 making it quicker. 00:22:02.458 --> 00:22:03.666 And as a result of this one, 00:22:03.958 --> 00:22:06.416 but another $400,000 of revenue 00:22:06.666 --> 00:22:08.458 attributed to that store. 00:22:08.458 --> 00:22:09.291 Now, the one thing I 00:22:09.291 --> 00:22:10.166 wanna pick up on here, 00:22:10.166 --> 00:22:11.500 I have in this important is that 00:22:11.541 --> 00:22:12.166 people in the audience 00:22:12.166 --> 00:22:13.791 might say, well, this guy is 00:22:13.791 --> 00:22:14.833 saying he does CRO. 00:22:15.208 --> 00:22:16.041 The conversion rate 00:22:16.041 --> 00:22:17.791 there has gone down 1.4%. 00:22:18.458 --> 00:22:19.166 And that's my point. 00:22:20.083 --> 00:22:22.166 Is that conversion rate alone is a 00:22:22.166 --> 00:22:23.375 very poor measure 00:22:23.583 --> 00:22:24.708 of how successful that 00:22:24.708 --> 00:22:26.250 everyone or unsuccessful 00:22:26.416 --> 00:22:27.500 the particular test is. 00:22:27.750 --> 00:22:29.250 But the revenue per visitor here 00:22:29.541 --> 00:22:30.833 gives us a clear review 00:22:30.833 --> 00:22:31.416 of what we're doing. 00:22:31.416 --> 00:22:32.458 So that would be my 00:22:32.458 --> 00:22:33.625 message to anyone is that 00:22:33.750 --> 00:22:35.875 consider revenue per visitor or 00:22:35.875 --> 00:22:37.041 revenue per session 00:22:37.375 --> 00:22:41.375 as a much more truer version of 00:22:41.375 --> 00:22:42.666 how the site is performing. 00:22:43.166 --> 00:22:44.125 That's the 11 years of 00:22:44.125 --> 00:22:45.083 experience saying that 00:22:45.166 --> 00:22:46.791 don't single out a metric, 00:22:46.791 --> 00:22:48.708 look them through pulse together. 00:22:49.458 --> 00:22:51.541 But how do I know if I'm a brand 00:22:51.583 --> 00:22:52.291 or if I'm a seller 00:22:52.333 --> 00:22:53.375 sitting on the audience, right? 00:22:53.750 --> 00:22:55.250 How do I know that, okay, 00:22:55.875 --> 00:22:57.416 what percentile am I standing 00:22:57.708 --> 00:22:59.041 compared to the different brands 00:22:59.041 --> 00:23:00.916 that's out there in e-commerce? 00:23:01.041 --> 00:23:02.625 How do I kind of benchmark myself? 00:23:02.666 --> 00:23:03.166 Sure. 00:23:03.416 --> 00:23:04.458 The important thing 00:23:04.458 --> 00:23:06.333 is that the way we all, 00:23:06.666 --> 00:23:07.250 regardless of 00:23:07.250 --> 00:23:08.416 whatever business you run, 00:23:08.708 --> 00:23:09.916 we're all here to do three things. 00:23:10.041 --> 00:23:10.750 We were here to get 00:23:10.750 --> 00:23:11.875 customers to buy now. 00:23:12.458 --> 00:23:12.916 We're gonna get 00:23:12.916 --> 00:23:14.166 customers to buy more. 00:23:14.666 --> 00:23:15.125 I'm gonna get 00:23:15.125 --> 00:23:16.250 customers to buy again. 00:23:16.625 --> 00:23:18.041 So buy now, buy more, buy again. 00:23:18.500 --> 00:23:19.375 And that's what we 00:23:19.375 --> 00:23:20.416 call the buy trifecta. 00:23:20.416 --> 00:23:22.333 Now we built this tool that allows 00:23:22.333 --> 00:23:23.416 any Shopify brand 00:23:23.666 --> 00:23:24.625 to then go and pull 00:23:24.625 --> 00:23:26.000 industry benchmarks 00:23:26.083 --> 00:23:27.500 for your specific niche. 00:23:27.833 --> 00:23:29.250 What it will do there is tell you 00:23:29.250 --> 00:23:30.166 your starting point. 00:23:30.500 --> 00:23:32.125 So each of these, buy now, buy 00:23:32.125 --> 00:23:32.791 more, buy again, 00:23:32.833 --> 00:23:34.041 we call as levers. 00:23:34.125 --> 00:23:35.458 So your conversion rate, your AOV, 00:23:36.000 --> 00:23:37.250 and your repeat purchase rate. 00:23:37.666 --> 00:23:38.625 First of all, if you 00:23:38.625 --> 00:23:39.916 use this to work out 00:23:39.916 --> 00:23:40.875 which lever it is 00:23:40.875 --> 00:23:41.833 that you need to pull, 00:23:42.250 --> 00:23:44.166 you can then use that as the 00:23:44.166 --> 00:23:46.333 guidance for the next steps. 00:23:46.416 --> 00:23:47.208 So the steps I 00:23:47.208 --> 00:23:48.291 described previously, 00:23:48.458 --> 00:23:49.583 this happens before that. 00:23:49.583 --> 00:23:50.750 So regardless of 00:23:50.750 --> 00:23:51.875 whatever stage you're at, 00:23:52.166 --> 00:23:54.125 use the trifecta tool on here to 00:23:54.125 --> 00:23:55.291 get the starting point. 00:23:55.625 --> 00:23:56.791 And then you can go forward 00:23:56.791 --> 00:23:58.541 and look into CRO a bit more deep. 00:23:58.916 --> 00:23:59.958 Now I wanna buy more 00:23:59.958 --> 00:24:01.125 time to listen to this, 00:24:01.208 --> 00:24:03.125 but we are capped onto this one. 00:24:03.875 --> 00:24:05.041 What's this interesting slide? 00:24:05.083 --> 00:24:05.958 Tell us more about it. 00:24:06.291 --> 00:24:07.125 Yeah, absolutely. 00:24:07.125 --> 00:24:08.333 So as I mentioned at 00:24:08.333 --> 00:24:09.750 the start that I also run 00:24:09.791 --> 00:24:11.291 an event called Econ collab club. 00:24:11.291 --> 00:24:11.625 I've been running 00:24:11.625 --> 00:24:12.583 about four years. 00:24:13.166 --> 00:24:15.166 We have a community of about 7,000 00:24:15.166 --> 00:24:16.625 e-commerce providers. 00:24:17.291 --> 00:24:19.833 We have merchants and agencies. 00:24:20.166 --> 00:24:21.166 We run events once 00:24:21.166 --> 00:24:21.916 a month in London. 00:24:22.375 --> 00:24:23.125 So if anybody 00:24:23.125 --> 00:24:23.958 would like to join us 00:24:23.958 --> 00:24:24.583 and obviously with our 00:24:24.583 --> 00:24:25.333 collaboration now, 00:24:25.333 --> 00:24:26.875 I have in with obviously ShopX. 00:24:27.416 --> 00:24:28.333 This is something that we're gonna 00:24:28.333 --> 00:24:29.416 be doing together more of. 00:24:29.708 --> 00:24:31.166 And I'd say we have about 180 00:24:31.333 --> 00:24:32.291 people gathering each month. 00:24:32.291 --> 00:24:33.916 So if you are gonna be in London 00:24:33.916 --> 00:24:35.083 or plan to be in London 00:24:35.125 --> 00:24:36.333 over the course of 00:24:36.333 --> 00:24:37.125 the next few months, 00:24:37.333 --> 00:24:38.750 do come and say hi to us. 00:24:39.583 --> 00:24:40.458 Lots of fun. 00:24:40.583 --> 00:24:43.125 8 a.m. movie theater, e-commerce. 00:24:43.250 --> 00:24:45.083 And it's nothing short of a movie. 00:24:45.166 --> 00:24:46.000 I have been there. 00:24:46.041 --> 00:24:46.791 Thank you so much, 00:24:46.791 --> 00:24:47.708 Adam, for doing it. 00:24:47.750 --> 00:24:48.541 A lot of this event 00:24:48.541 --> 00:24:49.666 we drew inspiration 00:24:49.666 --> 00:24:50.625 from Econ collab club. 00:24:50.666 --> 00:24:52.041 Thanks for having us right there. 00:24:52.250 --> 00:24:53.291 Hopefully we keep 00:24:53.291 --> 00:24:54.583 expanding the ShopX network 00:24:54.583 --> 00:24:55.666 to newer geographies. 00:24:56.291 --> 00:24:57.000 That's about it. 00:24:57.041 --> 00:24:58.250 And quickly going on 00:24:58.250 --> 00:24:59.625 to the man to your left. 00:25:00.375 --> 00:25:01.541 We spoke about CRO 00:25:01.583 --> 00:25:02.458 in the full funnel. 00:25:03.166 --> 00:25:06.208 Alex, your product stemmed out of 00:25:06.208 --> 00:25:07.291 you building your brands. 00:25:08.000 --> 00:25:09.166 That's what you said earlier. 00:25:10.250 --> 00:25:11.041 Ads is a very 00:25:11.041 --> 00:25:11.833 interesting part of it. 00:25:11.875 --> 00:25:13.333 I think Bonnie covered a pretty 00:25:13.333 --> 00:25:13.916 interesting thing 00:25:13.916 --> 00:25:14.708 in the previous slide. 00:25:15.916 --> 00:25:17.750 Making ads, I think the benchmark 00:25:17.750 --> 00:25:18.708 has come down lower. 00:25:18.875 --> 00:25:20.583 I think the entry to barrier is 00:25:20.583 --> 00:25:21.416 pretty low right now. 00:25:21.458 --> 00:25:22.750 Creatives, check. 00:25:23.041 --> 00:25:24.416 P-Max, check. 00:25:24.750 --> 00:25:25.875 Can I get enough in more 00:25:25.875 --> 00:25:27.208 performance marketing agency? 00:25:27.375 --> 00:25:27.791 Check. 00:25:28.041 --> 00:25:29.250 Do we get credit cards? 00:25:29.291 --> 00:25:30.291 We have air wallets in the group. 00:25:30.291 --> 00:25:30.625 They're ready to 00:25:30.625 --> 00:25:31.500 give a line of credit. 00:25:31.875 --> 00:25:33.500 How do you still stand out? 00:25:33.916 --> 00:25:34.458 Over to you. 00:25:34.458 --> 00:25:35.250 That's the topic we 00:25:35.250 --> 00:25:35.958 wanna start off with. 00:25:36.291 --> 00:25:37.041 Yeah, for sure. 00:25:37.250 --> 00:25:39.500 It's so easy to create ads now 00:25:39.500 --> 00:25:41.208 with different tools. 00:25:41.208 --> 00:25:43.458 I mean, ChargeGPT has tremendous 00:25:44.166 --> 00:25:45.208 image generation. 00:25:45.375 --> 00:25:47.333 Higgsfield has basically all of 00:25:47.333 --> 00:25:49.541 the LLMs integrated into it. 00:25:49.541 --> 00:25:52.125 And it can generate beautiful, 00:25:52.125 --> 00:25:53.125 amazing creative. 00:25:53.375 --> 00:25:56.541 Very fast and inexpensively. 00:25:56.541 --> 00:25:59.250 So creating ads is not a mode. 00:25:59.833 --> 00:26:02.208 So anyone can create lots of ads. 00:26:02.791 --> 00:26:06.083 But the mode is, on the 00:26:06.083 --> 00:26:07.833 advertising side specifically, 00:26:08.208 --> 00:26:12.333 is in understanding your consumers 00:26:12.541 --> 00:26:13.833 on a deeper level 00:26:14.208 --> 00:26:16.166 and understanding what kind of 00:26:16.250 --> 00:26:18.375 ads, what kind of formats, 00:26:18.375 --> 00:26:21.541 what kind of messaging would 00:26:21.583 --> 00:26:22.916 resonate with them 00:26:23.208 --> 00:26:25.416 on a deeper level to get them to 00:26:26.000 --> 00:26:26.916 buy your product. 00:26:27.166 --> 00:26:28.791 And so that's kind of like how it 00:26:28.791 --> 00:26:31.250 shifted over time. 00:26:31.958 --> 00:26:33.291 Like, when I was starting out, 00:26:33.291 --> 00:26:35.000 like e-commerce, like 2016, 00:26:35.416 --> 00:26:37.125 I was using Canva 00:26:37.125 --> 00:26:38.541 to make banner ads 00:26:38.541 --> 00:26:42.041 and hiring freelancers to make 00:26:42.041 --> 00:26:44.083 pages and stuff like that. 00:26:45.166 --> 00:26:46.458 Now it's so much easier. 00:26:46.833 --> 00:26:47.291 Many of those 00:26:47.291 --> 00:26:48.875 things are just like, 00:26:48.875 --> 00:26:50.750 even one person can go 00:26:50.750 --> 00:26:51.541 through the whole cycle 00:26:51.541 --> 00:26:52.875 from the initial idea to 00:26:52.875 --> 00:26:54.166 actually creating the ad 00:26:54.166 --> 00:26:55.250 and launching the ad 00:26:55.250 --> 00:26:56.916 without any problems. 00:26:57.541 --> 00:26:59.125 And it's like the landing page, 00:26:59.125 --> 00:26:59.875 the conversion site, 00:26:59.916 --> 00:27:03.833 also got simplified from that 00:27:03.833 --> 00:27:04.666 standpoint as well. 00:27:04.666 --> 00:27:07.708 So now it's really understanding, 00:27:08.250 --> 00:27:08.833 like when you are 00:27:08.833 --> 00:27:10.041 advertising, right? 00:27:10.041 --> 00:27:11.666 I think this is a very 00:27:11.666 --> 00:27:12.916 big point for everyone 00:27:12.916 --> 00:27:14.541 who's spending money on ads. 00:27:14.750 --> 00:27:15.958 You're not the only 00:27:16.708 --> 00:27:17.750 person advertising. 00:27:18.250 --> 00:27:19.166 And a lot of people, 00:27:19.166 --> 00:27:20.333 like you look at their ads 00:27:20.333 --> 00:27:20.833 and they're like, 00:27:20.833 --> 00:27:22.125 their ads look like, 00:27:23.000 --> 00:27:24.291 they think they're like alone. 00:27:24.291 --> 00:27:26.000 There's no competition, like no 00:27:26.000 --> 00:27:27.375 one selling similar products. 00:27:27.500 --> 00:27:28.500 Hey, buy this 00:27:29.083 --> 00:27:30.375 magnesium supplement. 00:27:30.375 --> 00:27:31.500 It's like, yeah, like 00:27:31.500 --> 00:27:32.833 there's like 300 other 00:27:32.833 --> 00:27:33.875 magnesium supplements 00:27:33.875 --> 00:27:34.750 that are being sold. 00:27:34.916 --> 00:27:36.458 How is your magnesium 00:27:36.458 --> 00:27:37.375 supplement different? 00:27:37.666 --> 00:27:39.750 And so really like 00:27:39.750 --> 00:27:41.333 what we focus on is 00:27:42.916 --> 00:27:44.750 consistently analyzing that 00:27:44.750 --> 00:27:46.125 competitive landscape. 00:27:47.000 --> 00:27:48.666 Like who's advertising, what 00:27:48.666 --> 00:27:50.291 angles that they're advertising 00:27:50.333 --> 00:27:52.083 with and also micro 00:27:52.083 --> 00:27:54.333 angles, which I'll explain. 00:27:54.625 --> 00:27:56.208 Let's do that fun exercise. 00:27:56.208 --> 00:27:57.041 I wanna do it. 00:27:57.041 --> 00:27:58.916 I'm gonna role play 00:27:58.958 --> 00:27:59.541 right now with you. 00:27:59.541 --> 00:28:01.125 Let's say that I'm a supplement 00:28:01.125 --> 00:28:02.250 brand owner right now. 00:28:02.833 --> 00:28:03.750 Clearly you have a playbook, 00:28:03.750 --> 00:28:04.541 that's why you're minting. 00:28:04.791 --> 00:28:05.541 You stopped counting 00:28:05.541 --> 00:28:06.458 beyond 200 million. 00:28:06.458 --> 00:28:06.833 So something 00:28:06.833 --> 00:28:07.791 that's working for you. 00:28:08.416 --> 00:28:09.375 Let's say I am a 00:28:09.375 --> 00:28:10.458 supplement brand owner. 00:28:10.875 --> 00:28:11.875 That's why I'm making about a 00:28:11.875 --> 00:28:12.916 million dollars already. 00:28:13.458 --> 00:28:15.791 I need to make ads for this BFCM 00:28:15.833 --> 00:28:16.791 or for a sale season. 00:28:17.166 --> 00:28:17.708 Walk me through the 00:28:17.708 --> 00:28:19.458 steps in which you suggest 00:28:19.458 --> 00:28:20.583 that I kind of go ahead because 00:28:20.583 --> 00:28:21.500 the time is huge. 00:28:21.541 --> 00:28:22.541 I'm gonna sell it in all the 00:28:22.583 --> 00:28:23.416 English markets. 00:28:23.625 --> 00:28:24.375 How do I go about it? 00:28:24.416 --> 00:28:25.458 What are my steps? 00:28:25.666 --> 00:28:26.583 Yeah, for sure. 00:28:26.583 --> 00:28:27.291 So number one, 00:28:27.333 --> 00:28:28.583 you have to analyze, 00:28:29.166 --> 00:28:29.833 let's say you're 00:28:29.833 --> 00:28:30.791 already advertising? 00:28:31.041 --> 00:28:32.083 Yeah, I'm already advertising. 00:28:32.083 --> 00:28:33.583 Okay, so step number 00:28:33.583 --> 00:28:35.708 one is really analyzing 00:28:36.000 --> 00:28:38.416 your current ads and seeing which 00:28:38.416 --> 00:28:39.375 parts of the market 00:28:39.375 --> 00:28:40.875 you're covering and not covering. 00:28:40.875 --> 00:28:43.208 So on every market, there is 00:28:43.291 --> 00:28:45.750 different stages of awareness. 00:28:46.666 --> 00:28:47.791 Raise your hand if you understand 00:28:47.875 --> 00:28:49.125 stages of awareness. 00:28:51.041 --> 00:28:52.458 Okay, yeah, it's 00:28:52.458 --> 00:28:53.541 very important concept. 00:28:53.541 --> 00:28:55.041 It's from Eugene Schwartz. 00:28:55.041 --> 00:28:56.708 It's kind of very old, very 00:28:56.708 --> 00:28:58.000 foundational marketing. 00:28:58.375 --> 00:28:59.500 What that means is 00:28:59.500 --> 00:29:00.666 in different market, 00:29:00.666 --> 00:29:03.250 there are people that understand, 00:29:03.833 --> 00:29:04.166 like there are 00:29:04.166 --> 00:29:04.958 people that don't know 00:29:04.958 --> 00:29:05.875 that they have a problem. 00:29:05.958 --> 00:29:07.333 Okay, so first one, know the 00:29:07.333 --> 00:29:08.333 problem awareness. 00:29:08.333 --> 00:29:08.958 Yeah, yeah, so 00:29:08.958 --> 00:29:10.000 they're like unaware. 00:29:10.041 --> 00:29:10.666 They don't even know 00:29:10.666 --> 00:29:11.416 that they have a problem. 00:29:11.416 --> 00:29:11.958 So you have to 00:29:11.958 --> 00:29:13.083 basically educate them 00:29:13.083 --> 00:29:14.000 that they have a problem. 00:29:14.000 --> 00:29:14.958 So then you can 00:29:14.958 --> 00:29:15.916 sell them the solution. 00:29:16.375 --> 00:29:19.458 Then there is problem aware. 00:29:19.458 --> 00:29:20.333 So okay, oh, I 00:29:20.333 --> 00:29:21.416 know I have a problem. 00:29:21.416 --> 00:29:22.416 I'm actively 00:29:22.416 --> 00:29:23.250 looking for a solution. 00:29:23.750 --> 00:29:25.250 Then there is a solution aware. 00:29:25.875 --> 00:29:27.375 So people already know like, oh, I 00:29:27.375 --> 00:29:28.541 need a supplement. 00:29:28.958 --> 00:29:29.375 And so they're 00:29:29.375 --> 00:29:30.750 actively looking and choosing 00:29:30.791 --> 00:29:31.250 which one 00:29:31.250 --> 00:29:32.416 specifically they want to buy. 00:29:32.958 --> 00:29:34.333 And then there's a product aware. 00:29:34.333 --> 00:29:35.041 So this is where 00:29:35.041 --> 00:29:36.250 Black Friday Cybermonet, 00:29:36.250 --> 00:29:39.708 it's like 40% off, special offers. 00:29:39.833 --> 00:29:41.000 People already know they want to 00:29:41.000 --> 00:29:42.541 buy your product specifically. 00:29:42.708 --> 00:29:43.333 You just give them 00:29:43.333 --> 00:29:44.541 the best offer possible 00:29:44.541 --> 00:29:45.500 and they make a purchase. 00:29:45.500 --> 00:29:47.208 And so a lot of the advertising, 00:29:47.208 --> 00:29:47.833 if you look at 00:29:47.833 --> 00:29:48.791 advertising landscape, 00:29:48.791 --> 00:29:49.458 it's like people like, 00:29:49.458 --> 00:29:50.750 hey, buy this product, 00:29:50.750 --> 00:29:51.375 buy this product, 00:29:51.375 --> 00:29:52.041 buy this product. 00:29:52.041 --> 00:29:53.250 But then there's no, 00:29:53.541 --> 00:29:56.125 like the whole kind of like funnel 00:29:56.166 --> 00:29:57.125 infrastructure, right? 00:29:57.166 --> 00:29:58.458 Like where you took that 00:29:58.458 --> 00:30:00.125 consumer from initially 00:30:00.166 --> 00:30:00.791 not even knowing 00:30:00.791 --> 00:30:01.625 they have the problem 00:30:01.625 --> 00:30:03.291 to leading them to that solution. 00:30:03.583 --> 00:30:05.041 They're just hammering those ads. 00:30:05.041 --> 00:30:06.166 And those ads are 00:30:06.166 --> 00:30:07.125 not very scalable. 00:30:07.333 --> 00:30:09.000 You cannot spend a lot on these 00:30:09.000 --> 00:30:10.125 like offer-based 00:30:10.666 --> 00:30:12.208 very product aware ads, 00:30:12.208 --> 00:30:13.208 because those are 00:30:13.250 --> 00:30:14.291 only for people that know 00:30:14.291 --> 00:30:15.333 about their product and 00:30:15.333 --> 00:30:16.083 already want to buy it. 00:30:16.250 --> 00:30:17.000 And probably drops 00:30:17.041 --> 00:30:18.125 credibility also, right? 00:30:18.125 --> 00:30:19.958 When you're a new brand out there, 00:30:19.958 --> 00:30:21.250 you keep like slamming and buy 00:30:21.250 --> 00:30:22.041 more, buy more, buy more. 00:30:22.041 --> 00:30:23.083 That kind of drops on the 00:30:23.083 --> 00:30:24.041 credibility also, right? 00:30:24.250 --> 00:30:25.208 Yeah, yeah, yeah. 00:30:25.250 --> 00:30:27.333 So, and so that's 00:30:27.333 --> 00:30:28.333 one aspect, right? 00:30:28.333 --> 00:30:29.541 Like of the whole kind of 00:30:29.541 --> 00:30:30.500 advertising landscape. 00:30:30.500 --> 00:30:31.875 So let's say you can analyze your 00:30:31.875 --> 00:30:33.041 whole like ad library. 00:30:33.041 --> 00:30:34.750 So I have an example here. 00:30:34.750 --> 00:30:36.125 This is from like 00:30:36.125 --> 00:30:36.791 one of the brands. 00:30:36.791 --> 00:30:37.583 You can see like how 00:30:37.583 --> 00:30:40.000 many ads they launched 00:30:40.875 --> 00:30:43.083 in 12 months, like 00:30:43.083 --> 00:30:45.875 63,000 ads, right? 00:30:45.875 --> 00:30:48.375 And so basically this 00:30:48.375 --> 00:30:50.291 is kind of the collective 00:30:50.291 --> 00:30:51.208 holistic data. 00:30:51.708 --> 00:30:52.833 They're also advertising for 00:30:52.833 --> 00:30:54.291 multiple Facebook pages. 00:30:54.458 --> 00:30:55.791 Who's advertising for more than 00:30:55.791 --> 00:30:56.833 one Facebook page? 00:30:58.666 --> 00:30:59.000 Okay. 00:30:59.291 --> 00:30:59.625 That's quite a 00:30:59.625 --> 00:31:00.500 few in the audience. 00:31:01.041 --> 00:31:01.583 Yeah, that's good. 00:31:02.458 --> 00:31:04.541 So yeah, that's one of the ways to 00:31:04.541 --> 00:31:05.750 also kind of like, 00:31:06.250 --> 00:31:08.625 because advertising and marketing 00:31:08.625 --> 00:31:09.708 gets touch points, right? 00:31:09.708 --> 00:31:10.750 So when you're advertising, 00:31:10.750 --> 00:31:13.125 that's why you want to basically 00:31:13.125 --> 00:31:14.958 hit people multiple times, 00:31:15.041 --> 00:31:16.500 ideal at different stages of 00:31:16.500 --> 00:31:17.583 awareness with your ads. 00:31:18.000 --> 00:31:19.541 And it doesn't like, 00:31:19.541 --> 00:31:20.208 you don't need to do 00:31:20.250 --> 00:31:21.416 any targeting anymore. 00:31:21.416 --> 00:31:22.541 Like targeting is outdated. 00:31:22.541 --> 00:31:23.333 It's that you don't 00:31:23.333 --> 00:31:24.166 need any targeting, 00:31:24.416 --> 00:31:25.666 just like age, 00:31:25.666 --> 00:31:27.625 gender, country, that's it. 00:31:27.666 --> 00:31:28.583 That's all you need. 00:31:28.583 --> 00:31:29.750 And Facebook will figure it out 00:31:29.791 --> 00:31:31.041 because the algorithm is so good. 00:31:31.041 --> 00:31:32.208 Andromeda is that new update, 00:31:32.208 --> 00:31:33.250 which kind of that's that 00:31:33.291 --> 00:31:34.416 targeting, I presume so. 00:31:34.541 --> 00:31:35.125 Yeah, yeah. 00:31:35.125 --> 00:31:36.916 So Facebook now at this stage, 00:31:36.916 --> 00:31:38.083 I mean, they're collecting data 00:31:38.083 --> 00:31:39.208 points from your behavior, 00:31:39.208 --> 00:31:40.041 from every user. 00:31:40.041 --> 00:31:41.416 So they know at which stage of 00:31:41.416 --> 00:31:42.291 awareness your ads. 00:31:42.291 --> 00:31:43.583 So they'll show you the right app, 00:31:43.625 --> 00:31:44.958 even if you don't have any like 00:31:44.958 --> 00:31:45.875 targeting settings. 00:31:46.458 --> 00:31:47.750 And so yeah, let's say we, if we 00:31:47.750 --> 00:31:48.750 took this brand, right? 00:31:48.750 --> 00:31:49.875 And if we took it, we send it 00:31:49.916 --> 00:31:52.875 through like Gethookd MCP, 00:31:53.625 --> 00:31:55.541 which basically pulls all of the 00:31:55.833 --> 00:31:57.500 ads from their library 00:31:57.708 --> 00:31:59.250 and then analyze it based on the 00:31:59.250 --> 00:32:00.833 stages of the awareness, 00:32:01.125 --> 00:32:02.208 analyzes it also 00:32:02.208 --> 00:32:03.750 on the micro angles. 00:32:04.208 --> 00:32:07.083 Micro angle is like, it's 00:32:07.083 --> 00:32:08.708 basically the domain, 00:32:09.541 --> 00:32:11.125 kind of like reason why people buy 00:32:11.125 --> 00:32:12.583 products, right? 00:32:12.583 --> 00:32:15.208 So for example, if, you know, who 00:32:15.208 --> 00:32:16.250 here has an iPhone? 00:32:18.708 --> 00:32:21.708 Okay, so if I'll ask like each 00:32:21.708 --> 00:32:22.791 person here separately, 00:32:22.791 --> 00:32:24.291 like why did you buy iPhone? 00:32:24.333 --> 00:32:25.000 You know, some people 00:32:25.000 --> 00:32:25.791 say, oh, I bought it 00:32:25.791 --> 00:32:27.083 because it has the best camera. 00:32:27.416 --> 00:32:28.791 Some people would say, oh, I 00:32:28.791 --> 00:32:29.791 bought it because, 00:32:30.666 --> 00:32:32.041 I don't know, everyone has iPhone. 00:32:32.041 --> 00:32:33.416 Everyone who's cool has an iPhone. 00:32:33.708 --> 00:32:34.583 And it's like, and 00:32:34.583 --> 00:32:35.083 different people, 00:32:35.083 --> 00:32:35.750 oh, it's like, because 00:32:35.750 --> 00:32:36.625 I could talk with my, 00:32:36.916 --> 00:32:38.083 on the FaceTime, right? 00:32:38.083 --> 00:32:39.291 With my parents, with 00:32:39.291 --> 00:32:40.458 my kids, with my family. 00:32:40.833 --> 00:32:41.541 And so different 00:32:41.541 --> 00:32:43.291 people react to different, 00:32:44.416 --> 00:32:44.833 kind of like emotional. 00:32:44.833 --> 00:32:46.208 Messaging, positioning. 00:32:46.541 --> 00:32:47.291 And so that's 00:32:47.291 --> 00:32:48.208 what we need to have. 00:32:48.208 --> 00:32:49.625 You need to have diversified, 00:32:51.166 --> 00:32:52.541 diversified kind of 00:32:52.541 --> 00:32:53.958 system in your marketing, 00:32:54.208 --> 00:32:55.541 both on the awareness level. 00:32:55.541 --> 00:32:56.916 So you're hitting all of these 00:32:57.375 --> 00:32:58.916 layers from the unaware. 00:32:59.208 --> 00:32:59.833 If you can track 00:32:59.875 --> 00:33:00.916 unaware, that's like, 00:33:01.041 --> 00:33:01.916 that's where you like 00:33:01.958 --> 00:33:03.291 unlock the real scale. 00:33:03.291 --> 00:33:04.291 That's like 50, 00:33:04.291 --> 00:33:05.750 100K a day ad spend. 00:33:06.166 --> 00:33:08.000 That's when you can push that, you 00:33:08.000 --> 00:33:08.791 know, good volume. 00:33:09.583 --> 00:33:10.708 And then on the 00:33:10.750 --> 00:33:11.791 angle side is like, 00:33:11.875 --> 00:33:13.333 you can analyze your whole 00:33:13.333 --> 00:33:14.958 competitive landscape 00:33:15.458 --> 00:33:16.875 in your niche specifically. 00:33:17.208 --> 00:33:17.833 Then you can 00:33:17.833 --> 00:33:19.500 cross-reference it with yours 00:33:19.541 --> 00:33:21.583 and see exactly which angles, 00:33:21.833 --> 00:33:22.666 which stages of 00:33:22.666 --> 00:33:23.791 awareness you're missing. 00:33:23.791 --> 00:33:26.291 So then you see the missing links 00:33:26.625 --> 00:33:27.541 and specifically what 00:33:27.583 --> 00:33:28.916 apps you need to launch 00:33:29.750 --> 00:33:30.583 in order to have 00:33:30.583 --> 00:33:31.916 this complete, you know, 00:33:31.916 --> 00:33:33.541 final infrastructure 00:33:33.541 --> 00:33:34.583 from start to finish. 00:33:34.875 --> 00:33:37.041 And so that has been the biggest 00:33:37.041 --> 00:33:38.250 like unlock for us, 00:33:38.583 --> 00:33:39.333 previously it would 00:33:39.333 --> 00:33:40.416 be just like manually 00:33:40.416 --> 00:33:40.875 and like we would 00:33:40.875 --> 00:33:42.625 have to look at every ad 00:33:42.625 --> 00:33:43.916 and like think about it. 00:33:43.916 --> 00:33:45.458 So AI simplified that 00:33:45.458 --> 00:33:46.666 process of analysis. 00:33:46.916 --> 00:33:48.541 But the good thing is, is that 00:33:49.000 --> 00:33:50.083 very few people, 00:33:50.083 --> 00:33:51.041 I mean, you guys 00:33:51.041 --> 00:33:51.958 now know about this, 00:33:51.958 --> 00:33:53.041 but like very few 00:33:53.041 --> 00:33:54.166 people know about this. 00:33:54.458 --> 00:33:56.166 Very few people have this like 00:33:56.166 --> 00:33:57.583 holistic understanding 00:33:57.583 --> 00:33:59.000 of advertising landscape. 00:33:59.875 --> 00:34:01.458 So if you actually take 00:34:01.458 --> 00:34:02.791 this to your advantage, 00:34:02.791 --> 00:34:04.291 I think the, you know, 00:34:04.291 --> 00:34:05.791 you unlock real skill. 00:34:06.083 --> 00:34:07.333 I mean, that's a lot of things 00:34:07.333 --> 00:34:07.708 that you just 00:34:07.708 --> 00:34:08.500 said for that, right? 00:34:08.500 --> 00:34:09.041 But I see on the 00:34:09.041 --> 00:34:10.291 screen over here, you say, 00:34:10.708 --> 00:34:12.000 what it took back in the day 00:34:12.083 --> 00:34:13.250 versus what it's taking now. 00:34:13.666 --> 00:34:14.791 How true is it? 00:34:14.791 --> 00:34:16.208 Because let's say a 00:34:16.208 --> 00:34:17.083 brand at your scale, 00:34:17.083 --> 00:34:18.000 you're running a $200 million 00:34:18.208 --> 00:34:18.916 brand out there. 00:34:19.166 --> 00:34:19.875 Is it applicable 00:34:19.875 --> 00:34:20.750 even at your level, 00:34:20.750 --> 00:34:23.166 where you replace a team of a 00:34:23.166 --> 00:34:24.708 large team to a smaller team? 00:34:25.125 --> 00:34:26.833 Yeah, I mean, this is real. 00:34:26.833 --> 00:34:28.250 Like, so for example, creative 00:34:28.250 --> 00:34:29.625 strategies on our team, 00:34:29.625 --> 00:34:31.000 like our creative strategies can 00:34:31.541 --> 00:34:32.958 go from the ideation, 00:34:33.000 --> 00:34:33.750 can analyze the 00:34:33.750 --> 00:34:34.666 competitive landscape. 00:34:34.666 --> 00:34:35.458 Okay, we're missing 00:34:35.458 --> 00:34:37.208 these stages of the funnel. 00:34:37.208 --> 00:34:38.541 We don't have underwear ads. 00:34:38.541 --> 00:34:41.583 Okay, he can Gethookd MCP to 00:34:41.583 --> 00:34:42.833 find the underwear ads. 00:34:43.083 --> 00:34:44.166 Boom, he has a concept. 00:34:44.166 --> 00:34:46.208 Then from there, you know, take, 00:34:46.458 --> 00:34:47.458 extract the script, 00:34:47.708 --> 00:34:49.000 rewrite it using, you 00:34:49.000 --> 00:34:50.958 know, Clod or OpenAI, 00:34:51.750 --> 00:34:53.416 and then create like a storyboard 00:34:53.416 --> 00:34:54.291 of his video ad, 00:34:54.291 --> 00:34:56.666 if it's image ad generated using 00:34:56.666 --> 00:34:57.791 ChudgPT and like, 00:34:58.166 --> 00:34:59.291 you know, launch it basically 00:34:59.291 --> 00:35:00.083 within like a few hours. 00:35:00.083 --> 00:35:01.291 So that whole 00:35:01.291 --> 00:35:02.416 cycle that previously, 00:35:02.416 --> 00:35:03.375 oh, you know, you 00:35:03.375 --> 00:35:04.333 send it to designer, 00:35:04.333 --> 00:35:06.291 designer's like, yeah, and you 00:35:06.291 --> 00:35:07.500 know, creates a mock-up, 00:35:07.541 --> 00:35:09.166 sends it back to you, oh, let's do 00:35:09.208 --> 00:35:09.916 these revisions. 00:35:09.916 --> 00:35:10.916 Let's adjust it back 00:35:10.916 --> 00:35:12.000 and forth, back and forth. 00:35:12.000 --> 00:35:13.416 Now it's like, bam, bam, bam. 00:35:13.791 --> 00:35:15.166 But the important thing is like, 00:35:15.166 --> 00:35:15.750 you need to know 00:35:15.750 --> 00:35:17.958 exactly what to focus on. 00:35:17.958 --> 00:35:18.916 And that's kind of, 00:35:18.958 --> 00:35:22.875 it's not, you know, 00:35:22.916 --> 00:35:24.416 the tools and everything makes it 00:35:24.416 --> 00:35:25.416 easier to create stuff, 00:35:25.416 --> 00:35:26.000 but you need to know 00:35:26.000 --> 00:35:27.416 exactly what to create. 00:35:28.291 --> 00:35:29.541 I'm pretty hooked on that line. 00:35:29.541 --> 00:35:30.791 Thank you so much, Alex. 00:35:31.791 --> 00:35:33.291 So how we make 00:35:33.291 --> 00:35:35.333 VLSs with AI, right? 00:35:35.333 --> 00:35:36.666 Like the research and stuff. 00:35:36.666 --> 00:35:37.541 Do you want to quickly tell us 00:35:37.583 --> 00:35:38.833 what is your checklist? 00:35:39.375 --> 00:35:40.958 Yeah, this is just one of the 00:35:40.958 --> 00:35:41.958 formats that we run. 00:35:41.958 --> 00:35:43.541 I mean, there is a, I 00:35:43.541 --> 00:35:44.625 mean, last time I checked, 00:35:44.625 --> 00:35:46.125 there's like 30 plus like 00:35:46.125 --> 00:35:48.375 different ad formats 00:35:48.750 --> 00:35:49.750 that we run. 00:35:50.000 --> 00:35:51.750 This VSL is just one of them. 00:35:51.750 --> 00:35:53.416 VSL is like the video that's 00:35:53.750 --> 00:35:54.583 typically longer 00:35:54.583 --> 00:35:55.791 than like three minutes. 00:35:56.625 --> 00:35:58.333 So they're the most scalable. 00:35:58.375 --> 00:35:59.916 Out of all of the ad formats, 00:35:59.916 --> 00:36:00.416 we found them to 00:36:00.416 --> 00:36:01.291 be the most scalable 00:36:01.291 --> 00:36:02.958 because if you crack a VSL, 00:36:03.166 --> 00:36:05.083 we've had VSLs spend like three, 00:36:05.125 --> 00:36:05.916 more than $3 million. 00:36:06.291 --> 00:36:08.041 Like one VSL, one ad, profitably. 00:36:08.416 --> 00:36:09.875 And so this is just, 00:36:10.000 --> 00:36:11.083 you know, the process, 00:36:11.083 --> 00:36:12.500 all of this process can be done. 00:36:12.875 --> 00:36:14.666 I mean, for this one, because VSL 00:36:14.666 --> 00:36:15.708 is a longer format, 00:36:15.708 --> 00:36:16.250 so you need a 00:36:16.250 --> 00:36:17.291 video editor for this, 00:36:17.291 --> 00:36:18.916 but let's say creative strategies 00:36:18.916 --> 00:36:19.875 can do the research, 00:36:20.000 --> 00:36:21.916 write a script, because 00:36:22.208 --> 00:36:23.708 copier, in my experience, 00:36:23.708 --> 00:36:24.500 copywriters and 00:36:24.541 --> 00:36:25.416 creative strategies 00:36:25.416 --> 00:36:26.625 are kind of like merged together. 00:36:26.833 --> 00:36:28.166 Like you don't need, I don't 00:36:28.166 --> 00:36:28.791 believe you need, 00:36:28.791 --> 00:36:30.291 you don't need like a separate 00:36:30.291 --> 00:36:31.500 like copywriter. 00:36:32.250 --> 00:36:33.666 Creative strategies, the core 00:36:33.666 --> 00:36:35.875 skillset has to be the copy. 00:36:36.333 --> 00:36:37.375 And then they can 00:36:37.375 --> 00:36:38.541 write the script, 00:36:38.541 --> 00:36:42.166 they can create the whole like 00:36:42.166 --> 00:36:43.750 story line, storyboard, 00:36:43.958 --> 00:36:44.541 then create the 00:36:44.541 --> 00:36:45.708 footages, be it all, 00:36:45.750 --> 00:36:48.208 A roll as well using Higgs filter, 00:36:48.208 --> 00:36:49.000 if you have like a real 00:36:49.000 --> 00:36:50.125 footage, real creator. 00:36:50.875 --> 00:36:52.625 And then also there's like 00:36:52.625 --> 00:36:54.750 separate AI workflow 00:36:54.750 --> 00:36:55.791 that we use based on 00:36:55.791 --> 00:36:57.291 all of our previous data 00:36:57.541 --> 00:37:01.291 to rank the ad, the copy, all of 00:37:01.291 --> 00:37:02.583 the factors against, 00:37:02.625 --> 00:37:03.750 you know, all of the previous 00:37:03.750 --> 00:37:04.541 winners we've had 00:37:04.541 --> 00:37:05.708 to see how likely is 00:37:05.708 --> 00:37:06.750 that one to succeed? 00:37:06.958 --> 00:37:08.375 If it requires some revisions, 00:37:08.375 --> 00:37:09.291 then we send it for 00:37:09.291 --> 00:37:10.958 revisions and then launch. 00:37:11.041 --> 00:37:11.541 Yeah, like I was 00:37:11.541 --> 00:37:12.208 catching up with you, 00:37:12.250 --> 00:37:13.500 clearly this format has been 00:37:13.500 --> 00:37:14.500 working pretty well for you. 00:37:14.500 --> 00:37:15.125 That's why I wanted you 00:37:15.125 --> 00:37:16.166 to kind of break it down. 00:37:16.500 --> 00:37:17.041 Makes sense. 00:37:17.666 --> 00:37:18.708 I don't know if you 00:37:18.708 --> 00:37:20.583 guys caught up on the tab, 00:37:20.791 --> 00:37:23.208 hardly from Shopify recently made 00:37:23.250 --> 00:37:24.375 a tweet last night. 00:37:24.375 --> 00:37:26.250 They have announced an 00:37:26.250 --> 00:37:27.541 integration with OpenAI 00:37:27.833 --> 00:37:30.541 where Shopify ads can be run 00:37:30.666 --> 00:37:31.500 within chat GPT, 00:37:32.208 --> 00:37:33.791 like being a 00:37:33.791 --> 00:37:35.666 generative engine optimized, 00:37:35.666 --> 00:37:37.500 being visible in a search engine, 00:37:38.083 --> 00:37:39.458 it's official right now and they 00:37:39.458 --> 00:37:40.416 have some credits unlocked. 00:37:40.416 --> 00:37:42.458 That's a partnership they kind of 00:37:42.750 --> 00:37:43.500 unlocked last night. 00:37:44.208 --> 00:37:44.708 I think that's a 00:37:44.708 --> 00:37:46.125 segue to go to Stefan. 00:37:47.750 --> 00:37:48.500 Two places of money 00:37:48.500 --> 00:37:49.500 already flows in, right? 00:37:49.500 --> 00:37:50.541 When we talk about a funnel, 00:37:50.541 --> 00:37:51.958 I think we spoke about the full 00:37:51.958 --> 00:37:52.791 funnel optimization, 00:37:52.791 --> 00:37:53.666 we spoke about ads. 00:37:54.541 --> 00:37:57.125 What are brands or people who buy 00:37:57.291 --> 00:37:58.166 are looking up on 00:37:58.166 --> 00:37:58.708 the search engine? 00:37:59.000 --> 00:38:00.500 I mean, gone are the days that I 00:38:00.500 --> 00:38:02.041 scroll down the page, 00:38:02.083 --> 00:38:03.541 I go to the bunch of blue links 00:38:03.541 --> 00:38:04.083 that's out there, 00:38:04.083 --> 00:38:04.916 click on a link and 00:38:04.916 --> 00:38:05.750 then get redirected. 00:38:06.375 --> 00:38:07.875 Answers I look up over there. 00:38:08.250 --> 00:38:09.833 What do you think are the places 00:38:09.833 --> 00:38:10.833 where people search 00:38:11.000 --> 00:38:11.791 and how do they 00:38:11.833 --> 00:38:13.083 kind of translate on? 00:38:14.333 --> 00:38:16.291 Yeah, I just want to 00:38:16.291 --> 00:38:17.500 take a step back first 00:38:18.041 --> 00:38:20.083 before I go into this and I want 00:38:20.083 --> 00:38:21.041 to say one thing. 00:38:21.125 --> 00:38:24.083 There's a lot of uncertainty. 00:38:24.958 --> 00:38:26.291 There's a lot of noise. 00:38:26.625 --> 00:38:29.125 There's a lot of fear about what's 00:38:29.125 --> 00:38:30.166 happening with AI 00:38:30.166 --> 00:38:30.708 and how things are 00:38:30.708 --> 00:38:32.125 changing just yesterday. 00:38:32.500 --> 00:38:33.583 In like six hours, 00:38:33.625 --> 00:38:34.875 four different people, 00:38:34.916 --> 00:38:35.333 while we were 00:38:35.333 --> 00:38:36.708 networking, asked me, 00:38:37.791 --> 00:38:38.791 "Do agencies still 00:38:38.791 --> 00:38:40.250 have a future next year? 00:38:40.250 --> 00:38:40.666 "How are things 00:38:40.666 --> 00:38:41.541 changing for apps?" 00:38:41.541 --> 00:38:41.958 And I was like, 00:38:41.958 --> 00:38:42.958 "You're the fourth person 00:38:42.958 --> 00:38:44.458 "that asked me this today." 00:38:44.791 --> 00:38:45.833 And I'll go and I'll explain. 00:38:46.208 --> 00:38:46.708 I'm sure there's 00:38:46.708 --> 00:38:48.166 gonna be many more today. 00:38:48.708 --> 00:38:48.750 (laughing) 00:38:49.166 --> 00:38:49.791 Exactly. 00:38:49.833 --> 00:38:52.291 So the thing is 00:38:52.333 --> 00:38:53.708 with all of this noise, 00:38:54.041 --> 00:38:56.041 we tend to get 00:38:56.041 --> 00:38:57.250 distracted very easily. 00:38:57.833 --> 00:38:59.333 And for my entire 00:38:59.333 --> 00:39:00.333 professional life, 00:39:00.416 --> 00:39:01.875 I've always been 00:39:01.875 --> 00:39:03.458 successful doing one thing, 00:39:03.750 --> 00:39:05.208 which is finding 00:39:05.458 --> 00:39:07.416 the low-hanging fruit. 00:39:08.208 --> 00:39:09.250 Because there's a lot 00:39:09.250 --> 00:39:10.541 of shiny toys around. 00:39:10.958 --> 00:39:12.041 There's Higgs Field. 00:39:12.083 --> 00:39:15.416 There's 64,000 ads in one year. 00:39:15.416 --> 00:39:16.625 It's amazing, it's 00:39:16.666 --> 00:39:17.916 unhuman basically. 00:39:18.000 --> 00:39:19.375 It would not be 00:39:19.375 --> 00:39:20.541 possible two years ago. 00:39:21.291 --> 00:39:22.625 So there's a lot of 00:39:22.625 --> 00:39:24.000 shiny toys out there 00:39:24.083 --> 00:39:25.291 and we tend to forget the 00:39:25.291 --> 00:39:26.125 low-hanging fruit. 00:39:26.125 --> 00:39:28.041 And there's like two things that 00:39:28.083 --> 00:39:29.375 we do for all of our clients 00:39:29.375 --> 00:39:30.875 and everyone here can do. 00:39:31.125 --> 00:39:32.666 And it's really not 00:39:32.666 --> 00:39:33.750 just for e-commerce. 00:39:33.750 --> 00:39:35.208 SaaS businesses can do this for 00:39:35.250 --> 00:39:36.375 their own businesses, 00:39:36.375 --> 00:39:37.291 for their own sites. 00:39:37.666 --> 00:39:39.791 And those two things are focusing 00:39:39.791 --> 00:39:41.125 on traffic and acquisition 00:39:41.125 --> 00:39:43.666 and focusing on personalizing for 00:39:43.666 --> 00:39:44.458 the intelligence. 00:39:44.458 --> 00:39:45.291 So making the 00:39:45.291 --> 00:39:46.833 store legible for LLMs. 00:39:47.375 --> 00:39:49.166 On this one, I actually have a 00:39:49.208 --> 00:39:50.416 very simple example 00:39:50.458 --> 00:39:51.750 from one of our clients. 00:39:54.458 --> 00:39:55.958 We particularly chose a 00:39:55.958 --> 00:39:57.208 mid-sized Shopify brand 00:39:57.208 --> 00:39:58.666 from the US, which 00:39:58.666 --> 00:39:59.791 is seven figures. 00:39:59.875 --> 00:40:03.458 And we chose them because if you 00:40:03.458 --> 00:40:04.625 go for a big brand, 00:40:04.833 --> 00:40:05.583 then the numbers 00:40:05.583 --> 00:40:06.291 are gonna be skewed. 00:40:06.291 --> 00:40:07.625 If you go for a smaller brand, 00:40:07.625 --> 00:40:08.708 the numbers are gonna be skewed. 00:40:09.250 --> 00:40:11.083 Mid-sized is actually gonna show 00:40:11.083 --> 00:40:12.916 what doing the work, 00:40:13.458 --> 00:40:15.666 optimizing your PDPs, 00:40:15.708 --> 00:40:17.166 personalizing the customer 00:40:17.166 --> 00:40:18.458 experience in ways 00:40:19.541 --> 00:40:21.791 in which I'm gonna show very soon, 00:40:22.916 --> 00:40:23.708 actually helps you 00:40:24.208 --> 00:40:25.500 to increase traffic. 00:40:25.500 --> 00:40:27.000 And if we look at the 00:40:27.000 --> 00:40:29.125 Google Search Console, 00:40:29.833 --> 00:40:31.166 everybody knows in August, 00:40:31.208 --> 00:40:31.916 there was like this 00:40:32.416 --> 00:40:33.166 update from Google 00:40:33.208 --> 00:40:34.541 and a lot of websites 00:40:34.541 --> 00:40:36.208 traffic went actually down. 00:40:36.500 --> 00:40:37.291 For this specific 00:40:37.291 --> 00:40:38.458 brand that we did, 00:40:38.583 --> 00:40:39.250 just what I'm about 00:40:39.250 --> 00:40:40.166 to show you right now, 00:40:40.208 --> 00:40:40.875 things right, 00:40:41.458 --> 00:40:42.791 traffic continued to go up 00:40:43.166 --> 00:40:45.125 because it's low-hanging fruit. 00:40:45.125 --> 00:40:46.125 You need to focus on 00:40:46.125 --> 00:40:47.666 improving the actual content 00:40:47.666 --> 00:40:48.416 in a way that's 00:40:48.416 --> 00:40:49.666 legible for humans, 00:40:49.666 --> 00:40:50.875 but most importantly 00:40:51.166 --> 00:40:52.583 for agents as well. 00:40:53.791 --> 00:40:54.708 I like how you put it as a 00:40:54.708 --> 00:40:55.375 low-hanging fruit 00:40:55.375 --> 00:40:56.958 because we are still very early in 00:40:56.958 --> 00:40:57.666 the learning curve 00:40:57.708 --> 00:40:58.791 for us, we are still 00:40:58.791 --> 00:40:59.666 catching up and stuff. 00:40:59.875 --> 00:41:00.958 That's a pretty impressive number 00:41:01.000 --> 00:41:01.916 that you see over there, right? 00:41:02.000 --> 00:41:03.375 A certain amount of citations, 00:41:03.375 --> 00:41:04.708 you've seen the visibility. 00:41:04.916 --> 00:41:05.833 Clearly the numbers 00:41:05.833 --> 00:41:06.750 speak for itself, right? 00:41:07.458 --> 00:41:08.541 Definitely want to 00:41:08.625 --> 00:41:10.166 uncover and try to see 00:41:10.416 --> 00:41:12.541 what was your prep that led you to 00:41:12.541 --> 00:41:13.583 the graph over there 00:41:13.583 --> 00:41:15.458 and how can we translate the same 00:41:15.500 --> 00:41:16.166 for the upcoming 00:41:16.166 --> 00:41:16.875 windows that we have? 00:41:17.750 --> 00:41:18.791 Well, the prep is 00:41:18.833 --> 00:41:20.000 always going to be 00:41:20.041 --> 00:41:20.791 understanding your 00:41:20.791 --> 00:41:22.250 business very well 00:41:22.375 --> 00:41:23.208 and making sure that 00:41:23.208 --> 00:41:24.291 you're not just prompting 00:41:24.583 --> 00:41:27.000 into an LLM of a choice 00:41:27.541 --> 00:41:29.583 and just asking, help me improve 00:41:29.583 --> 00:41:30.458 this product page 00:41:30.458 --> 00:41:32.166 because if the LLM doesn't have 00:41:32.166 --> 00:41:33.500 your business context, 00:41:33.875 --> 00:41:35.541 if the LLM doesn't have your 00:41:35.541 --> 00:41:36.833 organizational DNA, 00:41:36.833 --> 00:41:37.791 which is, I'm gonna 00:41:37.791 --> 00:41:38.666 keep talking about 00:41:38.666 --> 00:41:39.916 this organizational DNA, 00:41:40.541 --> 00:41:41.708 the LLM is gonna give you the 00:41:41.708 --> 00:41:43.166 exact same answers 00:41:43.166 --> 00:41:44.541 and the exact same information 00:41:44.541 --> 00:41:46.750 is giving to everybody else. 00:41:47.041 --> 00:41:49.041 So the prep, I think the 00:41:49.041 --> 00:41:50.166 foundational prep 00:41:50.166 --> 00:41:50.791 is making sure 00:41:50.791 --> 00:41:52.458 the context is okay. 00:41:52.875 --> 00:41:54.583 In our case, being an agency, 00:41:54.583 --> 00:41:56.375 we have the client context, 00:41:56.416 --> 00:41:58.666 which is brand voice, 00:41:58.666 --> 00:42:00.375 products, top sellers, 00:42:00.708 --> 00:42:02.125 history, et cetera. 00:42:02.333 --> 00:42:04.583 It has another layer of context 00:42:04.583 --> 00:42:06.125 which is our 00:42:06.125 --> 00:42:07.750 relationship to the client. 00:42:08.250 --> 00:42:09.250 So who we are, how 00:42:09.250 --> 00:42:10.291 we started working, 00:42:10.916 --> 00:42:11.875 meeting briefs 00:42:11.875 --> 00:42:12.708 and everything else. 00:42:12.708 --> 00:42:13.916 And a third layer of 00:42:13.916 --> 00:42:16.416 context, which is us, 00:42:16.666 --> 00:42:17.791 our organization, 00:42:17.791 --> 00:42:18.875 what makes us different 00:42:18.875 --> 00:42:20.833 from a different agency, what 00:42:20.833 --> 00:42:22.208 gives us a different voice, 00:42:22.291 --> 00:42:23.666 what we're focusing on 00:42:23.666 --> 00:42:24.458 the low hanging fruit. 00:42:24.458 --> 00:42:26.541 So it has to know these things. 00:42:27.416 --> 00:42:28.500 No, speaking of this one, 00:42:28.875 --> 00:42:29.875 one interesting thing that I 00:42:29.875 --> 00:42:31.125 noticed is Itaos, 00:42:31.875 --> 00:42:33.000 what is that, do you wanna quickly 00:42:33.000 --> 00:42:33.625 tell us about it? 00:42:34.000 --> 00:42:35.833 Yeah, Itaos is our 00:42:35.875 --> 00:42:36.833 operating system. 00:42:36.833 --> 00:42:38.291 We just built it 00:42:38.333 --> 00:42:39.708 this year for ourselves 00:42:39.708 --> 00:42:40.458 and whenever I 00:42:40.458 --> 00:42:41.333 showed it to clients, 00:42:41.333 --> 00:42:41.875 they were like, oh 00:42:41.875 --> 00:42:43.041 my God, I want this. 00:42:43.208 --> 00:42:45.416 So it actually spin off into a 00:42:45.416 --> 00:42:46.291 separate business 00:42:46.291 --> 00:42:48.083 and we signed a contract with a 00:42:48.416 --> 00:42:52.125 huge civics organization 00:42:52.458 --> 00:42:54.291 and we're building their own 00:42:54.291 --> 00:42:55.166 operating system. 00:42:55.375 --> 00:42:56.208 I just wanna show 00:42:56.208 --> 00:42:57.000 you guys one thing. 00:42:57.458 --> 00:42:59.958 So this is basically 00:42:59.958 --> 00:43:01.500 asking for a Q4 planner 00:43:01.583 --> 00:43:02.208 for one of the 00:43:02.208 --> 00:43:03.916 clients and this is a demo. 00:43:03.958 --> 00:43:04.791 So all the data you 00:43:04.791 --> 00:43:06.000 see here is non PII. 00:43:06.000 --> 00:43:08.375 I'm from European Union, so GDPR 00:43:08.375 --> 00:43:09.208 is very important. 00:43:09.416 --> 00:43:11.291 So this data is anonymized. 00:43:12.625 --> 00:43:13.166 Just look at the 00:43:13.166 --> 00:43:14.458 recommended Q4 approach. 00:43:14.833 --> 00:43:18.000 Keep the planner standard and then 00:43:18.000 --> 00:43:18.791 when it goes here, 00:43:18.791 --> 00:43:21.333 it says that we need to discuss 00:43:21.333 --> 00:43:22.541 with this client's request 00:43:22.541 --> 00:43:24.875 for a performance linked bonus. 00:43:25.166 --> 00:43:26.208 So the assistant 00:43:26.208 --> 00:43:27.500 knows our contract, 00:43:27.500 --> 00:43:28.208 it knows our 00:43:28.208 --> 00:43:30.208 relationship with the client itself 00:43:30.458 --> 00:43:31.708 and you can do this with 00:43:31.708 --> 00:43:33.000 your own AI assistance. 00:43:33.333 --> 00:43:33.875 Make sure it 00:43:33.875 --> 00:43:35.833 understands your business goals, 00:43:35.833 --> 00:43:36.708 whether those goals 00:43:36.708 --> 00:43:38.541 are like conversion rate 00:43:38.541 --> 00:43:39.958 or revenue per 00:43:39.958 --> 00:43:42.000 session or ads spend. 00:43:42.916 --> 00:43:44.416 If it has that context, it's gonna 00:43:44.416 --> 00:43:45.291 give you better data. 00:43:45.750 --> 00:43:47.250 What's pretty interesting is that 00:43:47.250 --> 00:43:47.875 every single one 00:43:47.875 --> 00:43:48.583 of them over here, 00:43:48.750 --> 00:43:49.291 I think they had a 00:43:49.291 --> 00:43:50.041 problem statement. 00:43:50.333 --> 00:43:51.833 They went into builder mode, 00:43:52.166 --> 00:43:54.250 him building a framework for CRO, 00:43:54.250 --> 00:43:55.208 you building Gethookd 00:43:55.208 --> 00:43:56.208 and you building it has 00:43:56.208 --> 00:43:57.041 and Louis has something 00:43:57.041 --> 00:43:57.833 interesting as well. 00:43:58.041 --> 00:44:00.000 I think that's the best way in me, 00:44:00.916 --> 00:44:02.833 which I think to stand out in this 00:44:02.833 --> 00:44:04.500 commoditized AI world over there. 00:44:05.333 --> 00:44:07.500 Speaking of which, now that you 00:44:07.500 --> 00:44:08.125 have an assistant, 00:44:08.166 --> 00:44:09.083 this cloud, this 00:44:09.083 --> 00:44:10.166 open AI and stuff, 00:44:10.500 --> 00:44:11.875 what questions we 00:44:11.875 --> 00:44:13.333 used to ask before, 00:44:13.625 --> 00:44:14.458 what questions do 00:44:14.458 --> 00:44:16.500 we ask now so that 00:44:16.541 --> 00:44:17.958 there's a personalized way in 00:44:17.958 --> 00:44:18.958 which we get an output 00:44:18.958 --> 00:44:19.708 so that it's not that 00:44:19.708 --> 00:44:20.916 same cookie cutter response 00:44:20.916 --> 00:44:21.708 that we all get, right? 00:44:21.916 --> 00:44:22.958 End of the day, we wanna have that 00:44:22.958 --> 00:44:23.750 amazing citation 00:44:23.750 --> 00:44:24.333 that we wanna have. 00:44:24.708 --> 00:44:26.041 How do you differentiate on that? 00:44:26.541 --> 00:44:27.625 Yeah, so this is a 00:44:27.625 --> 00:44:28.750 very simple experiment 00:44:28.750 --> 00:44:30.416 that anyone in the room can make. 00:44:30.416 --> 00:44:31.708 Whether you're a SaaS business 00:44:31.708 --> 00:44:33.291 or you're an e-commerce business, 00:44:33.625 --> 00:44:35.375 just go to any AI out 00:44:35.375 --> 00:44:37.916 there that you wanna use 00:44:38.250 --> 00:44:40.375 and have it do some deep research. 00:44:40.375 --> 00:44:41.666 What are people 00:44:42.000 --> 00:44:43.958 asking about my brand? 00:44:43.958 --> 00:44:45.750 What are people asking about my 00:44:45.791 --> 00:44:47.416 top selling product X? 00:44:47.791 --> 00:44:50.291 And you're gonna get those 30 00:44:50.333 --> 00:44:52.208 minutes long exercises 00:44:52.500 --> 00:44:53.333 with questions. 00:44:53.416 --> 00:44:54.583 It's gotten longer, right? 00:44:54.583 --> 00:44:55.750 I think with every new model, 00:44:55.791 --> 00:44:56.750 do you feel that Alex, 00:44:57.000 --> 00:44:58.250 yeah, it's getting longer. 00:44:58.625 --> 00:45:00.041 Now I'm like TLDR of 00:45:00.083 --> 00:45:01.208 that also, I just go to it. 00:45:01.875 --> 00:45:03.333 And you pick the top 00:45:03.333 --> 00:45:04.750 three, four, five questions 00:45:04.750 --> 00:45:06.750 over there and then you use a 00:45:06.750 --> 00:45:07.958 prompt in which you're like, 00:45:07.958 --> 00:45:08.750 hey, have a look at 00:45:08.750 --> 00:45:11.750 this product, URL only. 00:45:11.875 --> 00:45:13.708 Do not deep research, do not go 00:45:13.708 --> 00:45:14.666 through the internet, 00:45:14.666 --> 00:45:16.791 just based on the data 00:45:16.791 --> 00:45:18.625 that exists on my PDP. 00:45:19.208 --> 00:45:20.458 Am I answering any of 00:45:20.458 --> 00:45:21.625 these four questions? 00:45:21.625 --> 00:45:24.166 And sadly, most of the times the 00:45:24.166 --> 00:45:25.166 answers are gonna be 00:45:25.416 --> 00:45:27.166 the page does not state this, the 00:45:27.208 --> 00:45:28.458 page does not state that 00:45:28.500 --> 00:45:30.333 over and over again. 00:45:30.375 --> 00:45:32.916 Experiment, if you 00:45:32.916 --> 00:45:34.250 can turn the next slide. 00:45:34.958 --> 00:45:36.833 Okay, so this is 00:45:36.875 --> 00:45:37.708 basically the page 00:45:38.000 --> 00:45:39.833 that does not state this before. 00:45:40.041 --> 00:45:41.250 You're gonna have the image 00:45:41.250 --> 00:45:43.041 information, standard image, 00:45:43.083 --> 00:45:45.833 no title, no alt text, standard 00:45:45.833 --> 00:45:46.708 product description. 00:45:46.708 --> 00:45:48.541 Everybody has the same, 00:45:48.541 --> 00:45:50.000 nobody differentiates. 00:45:50.208 --> 00:45:51.000 And standard 00:45:51.000 --> 00:45:52.166 frequently asked questions, 00:45:52.166 --> 00:45:53.666 actually these are not standard. 00:45:53.666 --> 00:45:54.750 We still see a lot of 00:45:54.750 --> 00:45:56.791 sites with FAQs like, 00:45:56.791 --> 00:45:58.166 how long will it take until my 00:45:58.166 --> 00:45:59.083 order gets here, 00:45:59.083 --> 00:46:02.333 which is completely useless from a 00:46:02.333 --> 00:46:03.583 GEO perspective. 00:46:03.958 --> 00:46:05.541 So this is the page 00:46:05.541 --> 00:46:08.125 that got these prompts back 00:46:08.166 --> 00:46:09.708 and you can see, is 00:46:09.708 --> 00:46:10.666 this good for you? 00:46:10.708 --> 00:46:11.708 Question one, the 00:46:11.750 --> 00:46:13.125 page does not state this. 00:46:13.125 --> 00:46:14.125 It never mentions, 00:46:14.125 --> 00:46:15.375 polls, transition, 00:46:15.416 --> 00:46:17.208 vert and everything else. 00:46:17.458 --> 00:46:19.291 And for every single question, 00:46:19.291 --> 00:46:20.041 I'm gonna speed up 00:46:20.041 --> 00:46:21.125 because I saw that sign. 00:46:21.708 --> 00:46:23.458 For every single question is the 00:46:23.500 --> 00:46:25.125 page does not state this. 00:46:25.166 --> 00:46:27.708 So that was the problem. 00:46:27.708 --> 00:46:29.000 What did we do? 00:46:29.041 --> 00:46:30.375 We asked AI based 00:46:30.375 --> 00:46:31.166 on the questions, 00:46:31.166 --> 00:46:32.791 based on our organizational DNA 00:46:32.791 --> 00:46:33.791 and the client information 00:46:33.791 --> 00:46:35.416 to optimize this product page 00:46:35.416 --> 00:46:35.916 and based on the 00:46:35.916 --> 00:46:36.875 research that we did. 00:46:37.208 --> 00:46:38.541 Optimize the image, optimize 00:46:38.541 --> 00:46:40.625 description, optimize the FAQs. 00:46:40.666 --> 00:46:42.208 And then when we went 00:46:42.250 --> 00:46:44.125 and reran the experiment, 00:46:44.208 --> 00:46:46.083 if you can go to the next slide, 00:46:47.333 --> 00:46:49.208 it does state that the page states 00:46:49.208 --> 00:46:50.000 this information 00:46:50.000 --> 00:46:50.583 about the product. 00:46:50.583 --> 00:46:52.500 The product is a good fit for 00:46:52.625 --> 00:46:53.875 specific individual. 00:46:53.875 --> 00:46:54.750 The product does 00:46:54.750 --> 00:46:55.875 need this and that. 00:46:55.875 --> 00:46:57.333 So it does answer 00:46:57.333 --> 00:46:58.125 to these questions. 00:46:58.125 --> 00:46:59.875 It is visible now to 00:46:59.875 --> 00:47:01.291 the robot, to the agent. 00:47:01.666 --> 00:47:02.250 And it was 00:47:02.250 --> 00:47:03.958 basically zero out of zero 00:47:04.166 --> 00:47:06.250 and now it was four out of four, 00:47:06.666 --> 00:47:07.500 zero out of four 00:47:07.500 --> 00:47:08.666 and then it's four out of four, 00:47:10.166 --> 00:47:12.333 which is something that everyone 00:47:12.333 --> 00:47:13.500 here in the room can do. 00:47:13.500 --> 00:47:14.333 Everyone here in the room 00:47:14.375 --> 00:47:16.250 understands they need to do 00:47:16.708 --> 00:47:18.250 and it's there for the taking. 00:47:18.833 --> 00:47:19.541 Make yourself 00:47:19.791 --> 00:47:21.083 legible for the agents, 00:47:21.083 --> 00:47:22.500 if I may have to summarize that 00:47:22.541 --> 00:47:23.666 and it kind of makes sense. 00:47:24.625 --> 00:47:25.708 Yeah, I think this is 00:47:25.708 --> 00:47:27.166 how the rework happens 00:47:27.166 --> 00:47:28.333 and this is the answer, I presume. 00:47:28.791 --> 00:47:29.708 Yeah, exactly. 00:47:29.750 --> 00:47:32.416 So this is basically just as same 00:47:32.416 --> 00:47:33.125 page in your test. 00:47:33.625 --> 00:47:36.166 We got yes on every single one of 00:47:36.166 --> 00:47:37.041 the most important 00:47:37.041 --> 00:47:38.708 questions that people are asking 00:47:38.708 --> 00:47:40.041 about that product online. 00:47:40.166 --> 00:47:40.708 And again, this 00:47:40.708 --> 00:47:41.750 works for SaaS as well 00:47:41.750 --> 00:47:42.958 for service companies. 00:47:43.583 --> 00:47:44.958 Agency also, I presume. 00:47:45.000 --> 00:47:46.583 Agency can also make them visible. 00:47:46.791 --> 00:47:48.875 We did that for our own website 00:47:48.875 --> 00:47:50.500 and started adding that type of 00:47:50.500 --> 00:47:51.833 information like a year ago 00:47:51.833 --> 00:47:53.875 and it definitely worked. 00:47:54.375 --> 00:47:55.125 Wanna jump 00:47:55.125 --> 00:47:55.916 straight into it, right? 00:47:55.916 --> 00:47:57.666 I think gonna take a cue on this 00:47:57.666 --> 00:47:58.750 one because GEO, 00:47:58.791 --> 00:47:59.250 we didn't cover 00:47:59.250 --> 00:48:00.458 much in the BFCM panel. 00:48:00.750 --> 00:48:01.375 I think we are eight 00:48:01.375 --> 00:48:02.541 weeks away to the calendar. 00:48:03.125 --> 00:48:05.041 If I am running a business, I'm 00:48:05.041 --> 00:48:05.958 not gonna say a seller, 00:48:06.000 --> 00:48:06.541 I'm not gonna say an 00:48:06.541 --> 00:48:07.500 agency or a SaaS owner, 00:48:07.500 --> 00:48:08.041 I'm gonna say I'm 00:48:08.041 --> 00:48:08.791 running a business. 00:48:09.208 --> 00:48:09.625 I wanna be 00:48:09.625 --> 00:48:11.000 visible in the ZI world. 00:48:11.208 --> 00:48:13.333 What are my eight weeks for me to, 00:48:14.041 --> 00:48:15.166 what are my checklists for me to 00:48:15.166 --> 00:48:16.250 be ready in eight weeks? 00:48:17.833 --> 00:48:18.375 It's always gonna 00:48:18.375 --> 00:48:19.458 be the 80-20 rule. 00:48:19.500 --> 00:48:21.083 So focus on, if 00:48:21.083 --> 00:48:21.958 you're in e-commerce, 00:48:22.041 --> 00:48:23.208 20% of your products, 00:48:23.250 --> 00:48:24.541 20% of your collections 00:48:24.541 --> 00:48:26.333 slash category pages, 00:48:26.541 --> 00:48:28.625 20% of your landing pages 00:48:28.666 --> 00:48:29.958 and you want to make 00:48:29.958 --> 00:48:31.791 sure those are optimized 00:48:31.791 --> 00:48:33.375 as soon as possible with content 00:48:33.708 --> 00:48:35.791 that's at everyone's disposal 00:48:35.958 --> 00:48:37.291 as long as you personalize it for 00:48:37.291 --> 00:48:38.958 your own identity. 00:48:39.416 --> 00:48:41.625 So we wanna make sure we do that. 00:48:41.666 --> 00:48:43.458 It's data work but it's AI 00:48:43.458 --> 00:48:44.541 assisted nowadays. 00:48:44.541 --> 00:48:46.875 It no longer takes months to do, 00:48:46.958 --> 00:48:48.208 it takes days to do. 00:48:48.208 --> 00:48:50.416 People are scared about data work. 00:48:50.416 --> 00:48:51.208 They're scared about, 00:48:51.541 --> 00:48:52.875 oh, we have to change this 00:48:52.875 --> 00:48:53.791 for 20,000 00:48:53.916 --> 00:48:55.041 products on our website. 00:48:55.041 --> 00:48:56.666 But when you do the actual math, 00:48:57.041 --> 00:48:58.041 it's not like it's 00:48:58.041 --> 00:49:00.083 2010 or 2015 anymore. 00:49:00.791 --> 00:49:02.458 Operating on 20,000 products is 00:49:02.458 --> 00:49:03.750 much, much, much quicker. 00:49:05.250 --> 00:49:06.416 Then you wanna make sure you get 00:49:06.416 --> 00:49:07.708 your measurement in place 00:49:07.708 --> 00:49:08.583 and there was one thing 00:49:08.625 --> 00:49:10.500 earlier, when I saw Sam, 00:49:11.083 --> 00:49:12.375 he had like zero 00:49:12.375 --> 00:49:14.625 point, under 1% of traffic 00:49:14.625 --> 00:49:17.875 came from a genetic search. 00:49:18.208 --> 00:49:21.041 And that may be correct 00:49:22.041 --> 00:49:24.208 for an entire industry 00:49:24.583 --> 00:49:25.708 but what we've seen for our 00:49:25.708 --> 00:49:26.833 clients is that this year 00:49:26.833 --> 00:49:29.458 we are gonna most likely go over 00:49:29.458 --> 00:49:31.166 10% from Agenic. 00:49:32.458 --> 00:49:34.458 So I think that threshold of 00:49:34.458 --> 00:49:36.416 Agenic referral traffic 00:49:36.416 --> 00:49:38.041 versus SEO referral traffic is 00:49:38.041 --> 00:49:39.125 gonna happen much, much 00:49:39.125 --> 00:49:41.333 sooner than in the next 12 months. 00:49:41.958 --> 00:49:43.166 That's super awesome. 00:49:43.208 --> 00:49:43.750 I know that we have 00:49:43.791 --> 00:49:44.833 little spilled over time. 00:49:45.208 --> 00:49:45.875 Chris, next time we 00:49:45.875 --> 00:49:46.708 should have workshops 00:49:46.708 --> 00:49:47.333 instead of panels. 00:49:47.333 --> 00:49:48.833 Clearly I could spend a whole day 00:49:48.833 --> 00:49:49.750 listening to all of this. 00:49:50.583 --> 00:49:52.541 Well, going on, last 00:49:52.541 --> 00:49:53.333 but not least, right? 00:49:53.333 --> 00:49:53.875 Going on to the 00:49:53.875 --> 00:49:55.291 retention part of the funnel. 00:49:55.500 --> 00:49:56.541 We've done our conversions, we 00:49:56.541 --> 00:49:57.583 have run our ads well, 00:49:57.625 --> 00:49:58.791 we've made ourselves visible. 00:49:59.125 --> 00:50:00.208 The traffic has 00:50:00.208 --> 00:50:00.958 started flowing in. 00:50:01.208 --> 00:50:02.750 But end of the day, if 00:50:02.750 --> 00:50:03.708 I have a leaky bucket, 00:50:04.000 --> 00:50:05.166 I don't know how that's gonna go, 00:50:05.750 --> 00:50:06.625 Xavier actually had 00:50:06.625 --> 00:50:07.416 an interesting quote. 00:50:07.708 --> 00:50:10.625 For every $3, $2 actually comes 00:50:10.625 --> 00:50:11.666 from your email revenue. 00:50:11.708 --> 00:50:12.750 That's the latest quote that he 00:50:12.750 --> 00:50:13.625 got from K-Boss. 00:50:14.458 --> 00:50:16.083 You work very closely with Klaviyo. 00:50:16.083 --> 00:50:16.500 You're one of the 00:50:16.500 --> 00:50:17.541 leading agencies out there. 00:50:18.333 --> 00:50:19.875 Wanna try and understand with you 00:50:19.916 --> 00:50:21.541 about the retention part 00:50:21.541 --> 00:50:22.500 of the funnel, right? 00:50:23.083 --> 00:50:24.000 How do you personalize it? 00:50:24.291 --> 00:50:26.041 If you're going to do that same 00:50:26.041 --> 00:50:27.958 volume of emails, SMS, 00:50:27.958 --> 00:50:29.500 push notifications, or direct 00:50:29.500 --> 00:50:31.000 mails, the same way 00:50:31.041 --> 00:50:31.666 in a cookie cutter 00:50:31.666 --> 00:50:33.291 answer, how do we step out? 00:50:34.708 --> 00:50:35.000 Cool. 00:50:36.208 --> 00:50:37.291 That's a tricky question. 00:50:41.041 --> 00:50:42.541 I'm the guy who, every 00:50:42.541 --> 00:50:43.583 time Klaviyo asks me, hey, 00:50:43.583 --> 00:50:45.333 ask me a question, 00:50:45.375 --> 00:50:46.166 it always depends. 00:50:46.166 --> 00:50:47.583 It kind of depends on your goal. 00:50:47.875 --> 00:50:50.625 And so that's kind of a tricky 00:50:50.625 --> 00:50:51.416 question to answer. 00:50:52.041 --> 00:50:54.416 I have a few, I think what I can 00:50:54.500 --> 00:50:55.250 share is that there are 00:50:55.250 --> 00:50:56.916 a few developments in the 00:50:56.916 --> 00:50:58.458 retention marketing space 00:50:58.500 --> 00:51:00.541 that has changed a 00:51:00.541 --> 00:51:01.541 lot because of AI. 00:51:02.041 --> 00:51:03.625 But I'm gonna say, I probably have 00:51:03.666 --> 00:51:04.458 very little to share 00:51:04.458 --> 00:51:06.375 today because all of them have 00:51:06.375 --> 00:51:06.958 shared everything 00:51:06.958 --> 00:51:07.583 I'm gonna share. 00:51:07.791 --> 00:51:11.000 You need to have a brain, the DNA, 00:51:11.791 --> 00:51:13.333 to basically run your company. 00:51:13.375 --> 00:51:14.333 Now that AI has been 00:51:14.333 --> 00:51:15.416 able to help us that much, 00:51:15.916 --> 00:51:17.083 you kind of still need a really 00:51:17.083 --> 00:51:18.083 smart people on your team 00:51:18.083 --> 00:51:18.833 to make that human 00:51:18.833 --> 00:51:19.916 judgment, you know? 00:51:20.291 --> 00:51:21.166 And you kind of still 00:51:21.166 --> 00:51:22.041 need a matrix system 00:51:22.041 --> 00:51:24.541 to kind of decide, okay, now that 00:51:24.541 --> 00:51:25.166 AI is telling me 00:51:25.166 --> 00:51:26.666 to do these 50 things, 00:51:26.875 --> 00:51:27.625 what should I be doing? 00:51:28.208 --> 00:51:28.500 Right? 00:51:29.208 --> 00:51:31.750 But let me show you one video. 00:51:31.791 --> 00:51:34.458 If I cannot convince you today, 00:51:34.500 --> 00:51:36.500 let the CEO of 8Sleep 00:51:36.666 --> 00:51:39.000 convince you that you 00:51:39.666 --> 00:51:40.708 probably have to rethink 00:51:40.708 --> 00:51:41.583 how you are using AI 00:51:41.583 --> 00:51:42.291 to run your company. 00:51:43.041 --> 00:51:44.041 The tech gods are 00:51:44.041 --> 00:51:44.666 not in our favor. 00:51:44.666 --> 00:51:45.291 I'm not able to play 00:51:45.291 --> 00:51:46.291 the video at the moment, 00:51:46.291 --> 00:51:48.041 but jumping straight 00:51:48.041 --> 00:51:49.583 into the case study, right? 00:51:49.583 --> 00:51:51.666 Like if you wanna actually walk us 00:51:51.666 --> 00:51:53.291 through on the format. 00:51:53.458 --> 00:51:54.916 Sure, so I'll show you 00:51:54.916 --> 00:51:56.000 guys really quickly here. 00:51:56.041 --> 00:51:56.833 This is a case study 00:51:56.875 --> 00:51:58.041 for one of the clients 00:51:58.041 --> 00:51:59.125 called Mobile Tire Shop. 00:51:59.166 --> 00:52:02.250 They're the largest mobile tire 00:52:02.250 --> 00:52:04.375 fitting company in Australia. 00:52:04.833 --> 00:52:05.916 And the four things they're kind 00:52:05.916 --> 00:52:06.625 of seeing on the screen 00:52:06.625 --> 00:52:08.125 is exactly what the rest of the 00:52:08.125 --> 00:52:09.541 panel have mentioned. 00:52:10.041 --> 00:52:12.208 Inputs, playbooks, which are kind 00:52:12.208 --> 00:52:13.250 of the best practices. 00:52:13.625 --> 00:52:15.791 We agencies love to say that 00:52:15.833 --> 00:52:16.583 because we have a 00:52:16.583 --> 00:52:17.500 lot of clients, right? 00:52:17.500 --> 00:52:19.250 So we kind of ingest all 00:52:19.250 --> 00:52:20.291 of these best practices 00:52:20.583 --> 00:52:21.625 into our company brain. 00:52:22.041 --> 00:52:23.458 I think Stefan uses 00:52:23.458 --> 00:52:24.541 a custom software. 00:52:24.541 --> 00:52:25.875 We use Obsidian and Clod, 00:52:25.958 --> 00:52:26.916 whatever that kind of 00:52:27.041 --> 00:52:28.166 floats your boat, right? 00:52:28.875 --> 00:52:30.708 And human discretion, I think 00:52:30.708 --> 00:52:32.041 still think that's important. 00:52:33.333 --> 00:52:34.041 We're hiring a lot of 00:52:34.041 --> 00:52:34.916 strategies right now, 00:52:34.958 --> 00:52:36.291 a really good strategist. 00:52:36.750 --> 00:52:37.791 Like Alex mentioned, 00:52:37.833 --> 00:52:39.083 last time we need like three 00:52:39.083 --> 00:52:40.125 or four or five people. 00:52:40.166 --> 00:52:41.666 Now we just need one 00:52:41.666 --> 00:52:42.583 really good strategist 00:52:42.583 --> 00:52:43.500 to kind of make that happen. 00:52:44.125 --> 00:52:46.791 And the last part, it's AI and 00:52:46.791 --> 00:52:47.791 human execution. 00:52:48.625 --> 00:52:50.833 So the earlier video that I wanted 00:52:50.833 --> 00:52:51.375 to show you guys 00:52:51.375 --> 00:52:52.666 from It's Sleep CEO, 00:52:52.708 --> 00:52:53.666 you guys know It's Sleep. 00:52:53.750 --> 00:52:55.041 I think they are reasonably valued 00:52:55.041 --> 00:52:55.791 at about one and a 00:52:55.791 --> 00:52:56.791 half billion dollars. 00:52:57.458 --> 00:52:59.208 The video was saying, oh, we do a 00:52:59.208 --> 00:52:59.958 hundred million dollars 00:52:59.958 --> 00:53:00.958 from email marketing. 00:53:01.500 --> 00:53:03.625 Last time I have two people, but 00:53:03.625 --> 00:53:04.458 we're gonna quit. 00:53:04.500 --> 00:53:05.916 So we're like, what do we do? 00:53:06.583 --> 00:53:08.041 My co-founder went to, he was 00:53:08.041 --> 00:53:09.208 like, give me three days. 00:53:09.458 --> 00:53:10.916 He managed to automate their 00:53:11.000 --> 00:53:12.375 entire email marketing system 00:53:12.375 --> 00:53:14.000 with bots, just bots only. 00:53:14.250 --> 00:53:15.458 And I believe them because I look 00:53:15.458 --> 00:53:16.166 at their emails, 00:53:16.708 --> 00:53:17.666 It's Sleep is different. 00:53:17.666 --> 00:53:18.083 They already have 00:53:18.083 --> 00:53:18.958 the brand equity. 00:53:19.583 --> 00:53:21.041 Email is something that they can 00:53:21.041 --> 00:53:21.958 probably automate 00:53:21.958 --> 00:53:23.333 a lot with bots. 00:53:23.875 --> 00:53:26.083 It serves that when you are a 00:53:26.083 --> 00:53:27.208 million dollars in revenue 00:53:27.250 --> 00:53:29.208 and you're one and a 00:53:29.208 --> 00:53:30.000 half billion dollars, 00:53:30.458 --> 00:53:31.625 email marketing means very 00:53:31.625 --> 00:53:32.750 different, okay? 00:53:33.125 --> 00:53:34.791 So that's for us as well. 00:53:36.041 --> 00:53:37.791 We were able to generate 50 emails 00:53:37.791 --> 00:53:40.000 for mobile tire shop. 00:53:40.833 --> 00:53:41.875 This used to take three 00:53:41.875 --> 00:53:42.750 months, two years ago. 00:53:43.500 --> 00:53:45.500 We could charge a lot more. 00:53:46.416 --> 00:53:46.833 It's a three 00:53:46.833 --> 00:53:48.166 months project, right? 00:53:48.458 --> 00:53:49.791 But what do we do to 00:53:49.791 --> 00:53:51.000 kind of speed that up? 00:53:51.958 --> 00:53:52.708 Agency, we have 00:53:52.708 --> 00:53:53.875 sales calls, right? 00:53:54.125 --> 00:53:56.000 We do like one or two months of 00:53:56.000 --> 00:53:57.125 sales call to close this. 00:53:57.625 --> 00:53:59.750 These companies are VC bags, 00:53:59.750 --> 00:54:01.208 so they gotta go 00:54:01.208 --> 00:54:02.458 through a lot of layers. 00:54:02.500 --> 00:54:04.291 But every single call recorded. 00:54:04.708 --> 00:54:05.375 Every internal 00:54:05.375 --> 00:54:06.958 call recorded, right? 00:54:07.541 --> 00:54:09.208 We just throw in one prompt 00:54:09.291 --> 00:54:11.958 into our weapon of choices cloud, 00:54:12.000 --> 00:54:13.416 hey, based on all the 00:54:13.416 --> 00:54:14.458 calls that recorded, 00:54:14.583 --> 00:54:15.541 all of the audits 00:54:15.541 --> 00:54:16.208 that we have done, 00:54:16.625 --> 00:54:17.625 the Klaviyo MCP 00:54:17.791 --> 00:54:18.916 that's being connected 00:54:19.166 --> 00:54:20.125 and also connected 00:54:20.125 --> 00:54:21.416 to our company brain 00:54:21.833 --> 00:54:22.291 and whatever 00:54:22.291 --> 00:54:23.500 software that we are using, 00:54:24.208 --> 00:54:25.583 generate 50 emails from me 00:54:25.666 --> 00:54:27.541 according to the pitch, right? 00:54:28.166 --> 00:54:29.666 This sucks because this is making 00:54:29.666 --> 00:54:31.000 my job way too easy 00:54:31.041 --> 00:54:31.916 and everyone can do 00:54:31.916 --> 00:54:32.875 it as well, you know? 00:54:34.166 --> 00:54:36.791 But obviously, why people are 00:54:36.791 --> 00:54:37.583 still coming to us, 00:54:37.583 --> 00:54:38.750 first, not the 00:54:38.750 --> 00:54:39.791 best use of at a time. 00:54:39.833 --> 00:54:42.083 Second, like all of the panels 00:54:42.083 --> 00:54:42.750 here have mentioned, 00:54:42.791 --> 00:54:44.625 that company brain, that DNA is 00:54:44.625 --> 00:54:45.333 really important, 00:54:45.333 --> 00:54:47.208 your best practices, that's what 00:54:47.208 --> 00:54:49.083 people are coming to you for. 00:54:49.208 --> 00:54:51.083 They're like, "Chronos Agency, we 00:54:51.083 --> 00:54:52.583 have over 400 brands, 00:54:52.583 --> 00:54:53.041 they've been doing 00:54:53.041 --> 00:54:53.958 this for eight years." 00:54:54.583 --> 00:54:56.666 They probably know 00:54:56.708 --> 00:54:57.666 that their company brain 00:54:57.666 --> 00:54:58.916 is probably better, right? 00:54:59.750 --> 00:55:02.000 LLMs are great, but 00:55:02.000 --> 00:55:02.833 they're too general. 00:55:04.125 --> 00:55:06.166 If it makes your jobs easy 00:55:06.375 --> 00:55:07.541 and still pays you that figure 00:55:07.541 --> 00:55:08.333 that you quoted, 00:55:08.416 --> 00:55:09.416 I think that's job 00:55:09.416 --> 00:55:10.291 well done over there. 00:55:12.000 --> 00:55:13.583 Lastly on this one, right? 00:55:13.875 --> 00:55:14.916 Rich personalization 00:55:15.041 --> 00:55:16.000 at the speed of light. 00:55:17.125 --> 00:55:18.583 Okay, the Terrebonne, 00:55:18.708 --> 00:55:20.083 because we still have 00:55:20.500 --> 00:55:21.625 the last panel of this 00:55:21.666 --> 00:55:22.458 morning's chat center. 00:55:22.666 --> 00:55:25.791 So, yeah, it's the 00:55:25.791 --> 00:55:26.875 time that I do believe 00:55:26.875 --> 00:55:27.625 that we share a 00:55:27.625 --> 00:55:28.416 lot of thought food 00:55:28.416 --> 00:55:29.958 and insightful information for 00:55:29.958 --> 00:55:31.541 this session's end panel. 00:55:31.541 --> 00:55:33.583 Once again, thank you, you all. 00:55:33.833 --> 00:55:35.000 Oh my God, so productive. 00:55:35.000 --> 00:55:35.750 Can we have a big 00:55:35.750 --> 00:55:36.666 applause for them?