November 18, 2025 • 46 mins
If Amazon ads now feel like a maze of rising CPCs, endless keywords, and dashboards that never sleep, you’re not imagining it. The platform has shifted from organic-first to pay to play—and the only sustainable response is smarter execution that blends learning systems with human strategy.
We sit down with the CEO of Ava Commerce to unpack what “AI-driven” should actually mean for Amazon and D2C brands. Instead of black-box promises, we explore a configurable approach: let AI learn from your data across campaign creation, bid optimization, dayparting, and keyword negation, but keep owners and PPC leads in control with rules at the store, product, and campaign levels. You’ll hear how to tag products into portfolios with different TACOS goals, automate keyword isolation, and sunset underperformers quickly—freeing 65% of the grunt work so humans can focus on decisions that move margin.
Context still wins. We break down when to push for top-of-search visibility over strict ACOS, how to pair ads with pricing and promotions to lift conversion, and why creative and SEO matter more than ever for CTR and CVR. We also dig into audience strategy for DSP, where human insight crafts timing and intent—like targeting competitor buyers when they are most likely to repurchase—while AI handles the execution. Beyond ads, we touch on cash discipline through FBA reimbursements and vendor funds recovery, and why unlimited strategic support helps teams turn software into outcomes.
If you’re ready to scale profit, not just spend, this conversation offers a clear playbook: align creative with conversion, reclaim lost dollars, and deploy configurable AI that learns while you lead. Subscribe, share with a fellow seller who needs a smarter PPC plan, and leave a review with the one lever you’re automating first.
Ready to scale your Amazon business? Click here to book a strategy call. https://calendly.com/firingtheman/amazon
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (01:07):
Welcome everyone to the Firing the Man Podcast, a
show for anyone who wants to betheir own boss.
If you sit in a cubicle everyday and know you are capable of
more, then join us.
This show will help you build abusiness and grow your passive
income streams in just a fewshort hours per day.
And now your hosts, serialentrepreneurs, David Schomer and
(01:29):
Ken Wilson.
SPEAKER_01 (01:32):
Welcome everyone to the Firing the Man podcast.
Today's guest is Hi Meg, thevisionary CEO behind Ava
Commerce, or Ava.guru, one ofthe fastest growing AI platforms
in the Amazon and D2C ecosystem.
Under his leadership, Ava hasbecome a top 10 global
(01:53):
advertising technology solutionpowered by an over 100-person
development team drivinginnovations across PPC
optimizations, reimbursements,and profit analytics.
High is a true pioneer in theAI-driven growth for e-commerce
brands, combining data science,automation, and practical seller
(02:15):
experience to help thousands ofbrands improve margins, boost ad
efficiency, and uncover lostprofits.
Whether you're an Amazon veteranor a D2C founder looking to
outsmart your competitors,you're going to walk away from
this conversation withactionable strategies for
organic growth-drivenadvertising and scaling
(02:37):
profitability in 2025 andbeyond.
Hi, welcome back to the show.
SPEAKER_02 (02:44):
Hey, thanks for having me, David.
It's absolutely a pleasure.
SPEAKER_01 (02:48):
Absolutely.
So we had you on the show backin 2022.
And I remember in in my Rolodexof contacts, you were one of the
first people talking about AI.
And I, at that point in time, itwasn't even really on my radar.
And so uh my question is is howhave things changed?
(03:09):
How has the e-commerce landscapechanged?
And how is AI playing a uh rolein that?
SPEAKER_02 (03:16):
Well, it's kind of a funny story, but I'm talking
about AI, being a computerengineer with an a master's in
AI.
I'm talking about AI from '95almost, like 1995.
And I have to tell you, nobodywas listening to me at the time
because nobody believed in it.
And now it's 2025.
(03:37):
Now everybody believes in it,but there's so many people
talking about it, so much noise.
Now nobody's listening.
Yeah.
Because like too much noise,right?
But I mean, a lot of thingschanged and a lot of things are
changing.
Uh, I just want to give you thisidea.
I'm not gonna expend too much onit, but my son, who is in the
(03:58):
high school now, and he was likelike two years ago, he told me,
hey, maybe I should become acomputer engineer like you.
And I told him, son, it's youit's your decision, but I am not
sure if that particular uh jobwill exist in a decade.
And he was like, What?
(04:19):
You know, and I said, Well, Idon't know.
Because the AI is kind of likeum really going faster and
faster, and it's so amazing likeum how it evolved so far.
But I have to say that sixmonths later, that will be like
two times more things will come.
(04:39):
So that's the evolution of AItoday.
SPEAKER_01 (04:42):
Absolutely.
Uh since 1995, that'sincredible, and uh we'll very
good.
And and to all of our listeners,that's the um that it that is
the the educationalunderpinnings of high and why we
love having him on the show.
So uh I today's conversation, Ireally want to talk about
advertising on Amazon.
(05:03):
And this is something that hasdefinitely changed over over the
years.
Uh I remember when I firststarted selling on Amazon, you
could sell at a you could have ahundred percent organic sales on
certain products.
And it really has has changedinto a pay-to-play platform
where PPC has to be part of yourstrategy.
(05:24):
When you're pricing things,pricing in a taco slime item is
is a necessity.
And so if you know, from fromyour standpoint, hi, how how has
Amazon advertising changed?
SPEAKER_02 (05:38):
Oh, absolutely.
So, first of all, as you saidbefore, it was all about people
who bought this product, alsolike this product, type of nice
stuff that was available onAmazon, no ads.
And when the ads started, likeif you're selling a supplement
brand and you can just uh bid onvitamin C and you can make a lot
(06:01):
of money out of it.
Now, most of this is pretty muchgone.
Uh, what I'm seeing is like theway it evolved today is first of
all, that's the only way to growyour business.
Like, I mean, all the other waysare pretty much like
supplementary to that.
Or the other ways aresupplementing your advertising
(06:22):
strategy because you would liketo increase your improve your
CTR or CVR, definitely you'regonna invest on SEO or graphic
design and all that stuff.
But there are like two mainways.
I wouldn't say the only way, Ijust say that two main ways to
grow the business.
One is definitely theadvertising, and the other one
(06:42):
is the off-Amazon marketing,pretty much not the advertising
angle, but the influencer angleto that, right?
So these are the kind of the twothings that I see.
But when it's on Amazon, becauseit's a very data-driven
platform, now it has to bereally smart.
You know, like the the wholeadvertising needs to be smart,
(07:05):
and uh that's where the AI kicksin, right?
Like, and uh like the the onlychallenge is like maybe you
know, like we should go back toalso explain what it means
AI-driven advertising, becauseone of the things that I feel
like almost every softwarecompany and agency is telling
(07:27):
that they are AI powered,AI-driven.
Is that so?
Like, I think maybe we shouldstart with some definitions.
SPEAKER_01 (07:36):
Absolutely.
Absolutely.
So let's go.
Let's let's get into it.
What um what does AI-driven meanuh in the context of Ava and
what you guys have developedover there?
SPEAKER_02 (07:48):
Well, absolutely.
So, first of all, uh, I mean,from a kind of a terminology
point of view, from a conceptualpoint of view, not from a
marketing point of view,AI-driven system means the
system which is trained by thepast data and learns every day
(08:10):
uh from like the past data andthe decisions and the results of
the decisions.
So, meaning that the maindifference between a kind of a
rule-based system and AI-basedsystem is on a rule-based
system, you basically say, ifthe color is yellow, then do
(08:31):
this.
Now, the rule doesn't change,where on the AI system, you
basically end up seeing thatmaybe the yellow is not like if
it's yellow, do A, and then youend up learning, like the system
learns that A is not correct andswitches to B based on learning
(08:53):
from the outcomes.
So that's the kind of the maindifference.
One is learning and one is notlearning.
Now, the not learning system,which is rule-based, is what we
call AI-based in the market.
Like 98% of the softwares thatare available are rule-based,
(09:15):
which is nothing wrong withthat.
Okay, like pretty much all thecompanies that you know in the
market right now, more and moreI'm seeing that they are not
talking about AI, they sayautomation, etc.
But it's all rule-based.
Nothing wrong with that, becausethe rules make sense, but also
it is uh it also makes sense tobuild an AI-based system which
(09:41):
learns from the data and adjuststhe rules too.
SPEAKER_01 (09:47):
Okay, okay.
I'm excited to to dig into uhinto that in the context of PPC.
And one of the things that Ithink may be helpful well, let's
let's dig into that in in interms of PPC.
What okay, and maybe we coulduse an example, um, say like Dog
Bull or something like that.
(10:08):
When people are are setting upcampaign infrastructure and
they're doing it manually, uh,what are the things that they're
missing out on that an AI-driventool could help them with?
SPEAKER_02 (10:23):
Absolutely.
So I'll come to that point, butbefore that, let me explain uh
what we did with EVA because weare believers of AI, but we
realize that it is not about thesystem only deciding what to do,
because that's kind of a blackbox approach.
(10:45):
And over the last couple ofyears, even we started with uh
our journey with pricing, and uhthree years ago we were talking
about dynamic pricing with you,David.
And uh for many years thecustomers said, we don't like
the black box approach, we don'tlike what like without knowing,
(11:05):
doing things that like AI,basically it's like, why did you
do it?
And the the answer is AI, right?
So, and then it's notsatisfactory for many of the
Amazon sellers because they knowwhat they are doing too.
They want automation, but theyknow.
So we came up with the idea ofconfigurable AI, meaning that
(11:26):
the AI works, but you defineyour own rules, and the rules
override the AI.
We kind of like give thisanalogy of like today the 25 AI
technology is called the AIagents.
Basically, you create your ownagent based on the foundation of
(11:46):
Eva AI, but then you define yourown rules so that you kind of
like address like four mainpillars of automation.
Now, one of them is likecampaign creation, and maybe the
other one is bid optimization.
I'll talk about the other uh twopillars.
But let's think about thecampaign creation and the bid
(12:10):
optimization.
So, first of all, back to thethe example that you gave, the
system already it knowseverything.
Like, I mean, in the past,people use tools like Helium 10
or some other tools, JungleScout, etc.
Try to understand like whichkeywords, what is the search
volume, but this data is nowavailable on the Amazon uh you
(12:35):
know reports.
Now the question is, do youreally want the PPC guy to go
into these reports like searchterm report, SQPR, etc., and
then figure out like whichkeywords are now appearing as a
result of an automated campaign?
And now let's kind of create anew campaign out of that, right?
So that is where the EVA AIautomates based on uh, you know,
(12:59):
the Amazon data, and based onlike um, you know, basically we
have access right now to almost10,000 brands data and uh brings
it together and decides whetherit makes sense to create a
campaign or not.
Now, let's assume that the EvaAI already decides creating
(13:20):
these campaigns and kind ofthinks that if a campaign, if a
keyword has 200 impressions, um,I should create a campaign.
So let's say that's the Eva AIway.
Okay.
Now, as a business owner, youthink that, you know, the profit
level that you have, or thecategory or your experience
(13:44):
basically says that, I mean, youwant to create campaigns when
there is more than 500impressions.
Now, on top of it, you are evenmore cautious.
You want to add, like, also, Iwould like to see at least two
organic orders.
I don't know.
So you basically define whateverrule you want to define based on
(14:06):
50 parameters that are allavailable, anything that is
imaginable on Amazon that youcan define a parameter like C
VR, CTR, ACOS, tacos, revenue,organic sales, whatever.
Like think about all of them asavailable.
You define your own rule and youoverwrite AI.
If there is a rule defined, thenyou overwrite the AI.
(14:29):
So that's kind of our designprinciple.
Uh, you, as the owner, definethe rules.
If you don't define any rules,AI works.
If you define the rules, you'lloverwrite the AI.
The combination of that is yourspecial uh store-specific AI
agent, which works on yourterms.
(14:52):
That's the idea.
SPEAKER_01 (14:53):
I I really like that.
And and the the AI plus humanelement is something that I
think is is really critical.
Um, I've talked about thispublicly on the podcast.
I do it again.
I ran Perpetua for a while.
It's an AI-driven software, andI always felt like I had an
invisible hand in my account.
(15:13):
And when it came to some of thekeywords that it was bidding on,
I I was like, what this hasnothing to do with my product.
And I also I felt as a businessowner like I didn't have a lot
of say in how the automationswere performing.
And so I I like that you get toset your parameters and what
(15:37):
would be you know, I think PVCis very easily an overwhelming
topic.
Like there's it's something thatyou either go very deep in or
you're overwhelmed by it and youhire somebody.
And and so in in terms of youknow coming in and setting up
rules, what do you think aresome of the the most important
(15:57):
things for business owners to belooking at or to to set up?
SPEAKER_02 (16:02):
Absolutely.
So we basically meet with somany different types of business
owners, or even like the PPCexperts that work for other
companies.
Like obviously, their level ofexpertise is very different.
Now, uh one of the designprinciples that we have is like
(16:23):
this the tool needs to serveanyone, meaning that if you're a
business owner and you haveabsolutely no idea about what
you are doing with PPC, yourlevel of expertise is zero out
of ten.
At least you know what's yourbudget, like how much money you
want to spend, what should belike an ACOS or tacos.
(16:46):
At least you can set one of themand you can do it at a store
level, product level, or acampaign level, but probably
campaign level is already toodetailed, but at least on the
store level and product level,you set up your business idea,
like you can set your businessideas and you're done.
Then the rest will be done byAI.
(17:09):
Now, the biggest thing here isthat there is a tree-level
configuration possibility,meaning that it's not only at
the store level, but also at theproduct level and the campaign
level.
Because a lot of businessowners, I always tell them if
you're running your Amazonbusiness, your business is not
at the store level.
It's actually a combination ofsome of your products.
(17:31):
So it's actually the portfoliosare the ones that you are
running, meaning that eachportfolio may have a different
budget and a differenteco-stacos target.
One is like a newly newlylaunched product or it's a
launch product, one is likemaybe a profit maximization uh
portfolio, and the other one islike a couple of products coming
(17:54):
together because they are prettymuch the same.
So, meaning that you need tohave a system that you can
define your portfolios in a veryflexible way.
That's why we have a verysimple-to-use tag management.
So you basically create yourcustom tags on your products,
and now, voila, you have thatportfolio which is tagged, and
(18:16):
now you can define your rolebased on the portfolio.
Now, that is for a businessowner.
It's really two things.
They can tag the products andthey can configure uh the
business parameters at theportfolio level.
But if you're a PPC expert, nowyou want more.
We basically identified fourareas of automation.
(18:39):
One of them is definitelynegative keywords or negation,
like you mentioned before.
And now there is the bidoptimization, um, automatic
campaign creation, day parting,and uh keyword negation.
So these are the the four mainareas of automation that we see.
(19:00):
Now, the system is alreadyautomating uh with AI, okay, so
which is learning andautomating.
But the the biggest thing is nowyou can define any rules as a
business owner or the PPC expertat any of these main four
pillars.
So you can also say that, hey, Iwould like to negate these words
(19:23):
specifically, or you can also goback and like, you know, change
the bid optimization.
Or the system suggests you, forexample, not like you know, I've
seen a lot of agencies liketalking about day parting, they
say, hey, based on myexperience, between, I don't
know, like midnight and 5 a.m.,uh, there is no sales.
(19:44):
Well, but it's a genericstatement.
Now, in our case, the systemsays, based on the data, here is
the suggestion.
Between 2 a.m.
to 4 a.m., you know, like do notuh, you know, spend money on
ads.
But between 6 p.m.
and 7 p.m., do.
And it depends on the product.
(20:05):
It doesn't, it's not like a verygeneric statement.
So that's the day parting thatsystem already works on it, but
the owner or the PPC expert canoverride it by defining that
specific hourly rule as well.
SPEAKER_01 (20:19):
Okay.
Uh you talked a little bit aboutcamp or campaign creation, and
and I'd like to talk about thatand in campaign infrastructure.
I it seems like if you were totake a PPC 101 class, one thing
that may be covered, and maybethis is outdated, uh, but is the
different types of campaigns.
(20:40):
And I would say that in thisclass, this was what was taught
probably three or four yearsago, is when you launch a
product, you set up an autocampaign, and what that's going
to do is it's going to harvestkeywords for you.
And when you have uh so manyconversions or orders, you take
those keywords and you put theminto a manual campaign.
(21:01):
Now, there are so many moreoptions.
You've got exact phrase, uh,match, broad.
And so going through a productlaunch and then just kind of
sustaining a product product'slevel of sales, what do you
think is the ideal campaigninfrastructure?
SPEAKER_02 (21:21):
First of all, as you mentioned, like all this keyword
isolation and uh creating themanual campaigns from the
automated campaigns.
This, I think, alone is like20-25% of a PPC expert's job
because there is always newproducts, there's always new
(21:41):
keywords coming in, and you needto always have that count
discussions and checking at thesystem and everything.
Now, we basically emulated it100%.
So everything happens throughthis automation.
Now, in this case, I wouldn'tcall it specifically AI, but
this is like pure automationbecause it's already clear.
(22:04):
Well, you can still decide onhow this automation is
triggered.
And the other thing is like it'salways not that easy for a
person to, for example, create acampaign and assign just one
keyword to that campaign, andthen also decide like whether it
is an exact or broad and etc.
(22:26):
Now, the system can decide on itbecause when the system sees
like uh somebody is looking fora light bulb, but it's 60 watts.
Okay, now there's always like Ilove this example because that's
one of our customers, and it'svery clear people actually don't
search for light bulbs.
(22:49):
They search with that specificuh information because there's
so many different light bulbs.
So it's an exact search.
Now, the AI also knows that it'san exact search because looking
into the search terms, it'salready pretty obvious that that
is the search term, and there isno light bulb search term.
(23:10):
Maybe even you negate that ifthat makes any sense.
You know, it depends on thecontext.
But the idea here is like thesystem kind of starts if knows
to decide decides like whetherit's exact or broad.
In some cases, we let the systemrun the campaigns for some time,
(23:30):
and then the ones it's kind of anatural selection process in a
way that the ones which are notconverting after a day or two,
they die.
So that's also another way oflike, you know, like the way
that the system kind of decides.
Now, the complexity is huge.
Like if you leave it to a manuallike do done by a PPC expert,
(23:53):
um, well, we still need the PPCexpert.
Don't get me wrong, we're notgetting rid of that PPC expert.
But this particular task or dayparting, like given that um for
a product, there might be ahundred campaigns.
And if you want to do dayparting and optimize the bits,
like we're talking about everyday thousands of updates.
(24:17):
I mean, who is going to takecare of like every single thing?
That's why I think AI is notoptional, it's like almost like
mandatory.
It's it's the the core of thebusiness is done by the AI plus
your rules.
Again, I'm talking about theconfigurable AI, right?
(24:37):
And uh I would talk about thePPC expert and what we think
that guy should do, but I hope Iat least answer your question
for now.
SPEAKER_01 (24:48):
Absolutely.
And and I what one thing I Iwant to clear up is is on on a
particular uh single product,what campaigns are would
typically be running?
Is it an auto, broad, phrase,and exact?
Is it is there or does it justdepend on the product?
SPEAKER_02 (25:11):
Depends on the product and depends on the uh
the success, right?
I mean, I think the system,unless it is pre-defined,
because in some could categoriesand some of the search terms
immediately end up AI to run theexact campaign, but in some of
the other cases, it will makemore sense to run these
(25:33):
campaigns and then see theresults and then kill some of
them.
You know, like that's what likethe AI is doing, uh, which you
know is also another possibilityas well.
Of course, um there is like somany different parameters here,
right?
Like branded versus non-branded,or is it an awareness or
(25:54):
consideration or conversioncampaign?
Because these are different thanlike, you know, like one of the
big mistakes, right?
So people say, oh, ACOS too highon this campaign, kill it.
Like, you know, people do thisadvertising audit.
I love it.
You know, like the agencies doadvertising audits, and then
they basically say, Thiscampaign, we need to kill it
(26:15):
because it's 10%, it's uh 50%ACOS or something, right?
Well, the the question, well,this is an awareness campaign,
right?
Like the ACOS level versus akind of a converting campaign
might be different, right?
So the system kind of looks intoall these uh multi-dimensions
like uh broad, exact, match,plus, branded, non-branded, plus
(26:41):
um where is it like what type ofa campaign is that in terms of
the marketing funnel, and thendecides based on that?
It's where it becomes reallythat's the main complexity
because of the multidimensionalum decision-making process, it
becomes impossible for a humanto decide.
(27:03):
So, what human does is onlydecide based on maximum two
dimensions, and like we we callthat guy like the best PPC
expert, but even that guy willignore some of the other
dimensions when deciding onthat.
SPEAKER_01 (27:21):
Absolutely, absolutely.
You were talking about thedifferent pillars of PPC, and in
bid optimization was one.
And if if you talk about just amanual way of optimizing an
account, you know, you'll you'llclick on a manual campaign,
there'll be 20 keywords there,and there'll be 10 of them that
aren't getting any impressions.
(27:42):
And uh let's just use dog bowlas an example, it's not getting
any impressions.
Amazon suggests that you bidbetween 75 cents and a dollar,
and you ask yourself thequestion, well, if I could get
it for 75 cents, why would Iadjust my bid to a dollar?
Um and so you make youradjustment and then you come
back two weeks later and see ifyou have any impressions.
(28:04):
And so how does how does thetool optimize bids and how does
it look at increasing bids ordecreasing bids?
Because that is kind of a scarything.
You are giving a machine yourcredit card to go spend money on
your behalf.
And and so um how does thesystem work with bid
(28:27):
optimization?
SPEAKER_02 (28:28):
Absolutely.
So I mean, I agree with you.
Um obviously we always thinkthat the best thing here is to
combine the PPC expert and theAI agent, right?
I mean, I'm not I don't believethe software alone approach.
And I would like, I mean, as aguy in the industry, a computer
(28:50):
engineer who is like very happywith AI, you know, taking care
of a lot of tasks.
But I don't believe that if youjust let your system to run by
AI, you will not go anywhere.
First of all, let me explainthis one.
Because a lot of this, you know,you asked the question of the
bid optimization, right?
(29:11):
But a lot of these decisions isthat are these decisions really
depend on your data or thecontext?
Meaning that think about likewhat can go wrong.
Here you go.
Like, let's say your competitoris out of stock, and another one
just increased the price.
(29:32):
As a result of that, now you areselling more, and people are
even clicking more to your ads,and the conversion even
increased, everything is great.
And think about the AI solution.
The AI solution will think that,hey, I am doing great.
I mean, you are not doing great,it's the context that is doing
(29:54):
that's why the uh autonomous carcompany, that thing, already
like you know, companies likeWaimo, Google, etc., they spent
trillions of dollars and stillit's not there yet, right?
There is like maybe a couple ofyears to go, like, but it's
trillions of dollars.
(30:15):
In the beginning, it was like,hey, in one year, like the
autonomous cars will beeverywhere.
Why?
Because the context parameters,now it's the same thing.
The context is changing, it'snot the same context.
And the AI is looking at yourdata.
It's it's hard for AI to knowthe other contexts unless there
is some scraping, etc.
(30:36):
Fine.
You know, we do, I'm not sayingwe do it, but we are able to get
third-party data and do someranking-based campaigns because
we want to do organic growth,etc.
But still, you cannot controlall the other parameters around
you.
Like, even maybe somebody islike creating an attack campaign
and reduce the price one dollarbelow you and attacking you, how
(30:59):
do you know?
Like if you're the AI system,like you don't know.
So, what I'm saying is that'swhere the PPC expert should.
Focus on.
We are pretty much taking careof 65% of all the tasks of the
can the PPC expert shouldn't betoo much focused on the bid
(31:21):
optimization on everything, butreally focus on that competitive
intelligence.
Because that's where thedifferentiator of the PPC expert
is from our point of view.
When it comes to the bidoptimization, again, in the
system, it's not like you know,just because you say, hey, spend
$10 on this bet, the system cando it, right?
(31:43):
Because even the Amazon systemdoesn't work like that.
Like if you spend a lot of, likeif you say$10 per bid, but the
CPC is$2, like probably you'llend up somewhere spending$2,
anyways.
But then again, uh the mainthing, and by the way, you know,
this is an evolving strategyhere.
(32:05):
And uh the the problem withoptimizations is what if like
you spend a bit more and you'llbe in the top of the search?
And what if you are doing uh asuboptimal strategy, meaning
that you are suboptimizing thebit, meaning your ACOS is fine,
(32:26):
but you're you're in the you'rein the third row.
And maybe if you put a bit more,you'll be on the first row, and
that's where I am very muchfocused right now, because um I
can tell you that is like Idon't see any risks today, like
even a human as human itself ismore risky than the Eva AI tool,
(32:49):
and Amazon doesn't allow manythings, but you always have a
PPC expert, or the system hasall types of alerts, like you're
not gonna spend like a hundredthousand dollars.
There is no way.
But uh the main thing that I amfocusing on is like even the
people, the the PPC specialistor the system always looks into
(33:11):
Acos, tacos, C VR, CTR, but kindof ignores what happens if I am
on Amazon.com, I search for thatkeyword, and where am I?
Like, where am I exactly?
And how can I go maybe on top?
Like, because that's not what AIdoes, okay?
It's kind of a differentstrategy here.
(33:32):
Like it's not optimization, it'slike I need to bid for that and
then get to the top of thesearch, right?
So that's a bit different.
That's where we want to, we arefocusing our RND right now.
Uh we are almost there, butthat's a big deal to achieve
that.
SPEAKER_01 (33:48):
Absolutely, absolutely.
And and I have to say, it'srefreshing to hear a company
founder that's that's AI driventalk about the human element,
uh, and that that as we standhere, it's a necessity.
Um you had made the comment onat the beginning of the episode
how there's a lot of AI-poweredcompanies, uh, and to people
(34:09):
tuning in on audio, I'm doingair quotes here.
Uh, but what is that what doesthat really mean?
And I I I would use the exampleof just just because you give
somebody a hammer doesn't meanthey can build you a house.
And and in in this case, um, youknow, uh just having the the
hammer being the technology, uhthat there need there needs to
(34:34):
be the human element as as westand here talking.
And so um we promised to deepdive on PPC, and I think we're
delivering.
Hi, are there any other uhPPC-related topics that you
think would be worth talkingabout?
SPEAKER_02 (34:49):
I just want to uh emphasize the point of human,
right?
Like, especially like now, likethere's of course the
search-based campaigns.
I mean, there's also likeretargeting and remarketing
stuff, but then there's theaudience-based campaigns.
I mean, when it comes to theaudiences and creating the
audiences, um again, you needthe human element because that
(35:15):
is not done automatically.
What is automatic is like theAmazon Marketing Cloud or even
our system can give you theideas, but at the end, the human
decides like what makes sense uhto bid for when it comes to an
audience-based marketing.
And that's why, for example, weare also a DSP provider, and uh
(35:38):
like uh our customers say, hey,great, the PPC automation is
great, we're gonna have this.
We understand now our PPC expertcan look after, I don't know,
instead of um one account, maybefive accounts.
What about DSP automation or DSPAI with DSP?
And I tell them, like, hey, Ithink pretty much all the DSP
(36:00):
automation uh is kind of likerubbish from my point of view
because that's all aboutremarketing, retargeting stuff,
which you can also easily do,which makes sense.
But the top the point is, do youreally need to buy a tool to do
that?
And then the real DSP value isgenerating that audience.
(36:20):
For example, you are a peanutbutter um like selling, like
you're selling peanut butters,and you're targeting the
audiences which purchased peanutbutter from your competitors
four weeks ago, with theassumption that now the peanut
butter is kind of like uh youknow it's not enough, and like
(36:41):
the the the consumer will buy apeanut butter in the next one
week, let's assume.
Now, this is kind of like thescenario that you generate, and
that is where we need human.
I mean, we need human to tothink about all these
differences.
That's why the competitivecontext, I think, is the most
(37:02):
like the biggest differentiatorof the human.
I just want to give anotherexample.
What if it makes more sense togive a discount or a strike
through price or a coupon orsome type of a deal instead of
maybe pushing more advertising?
Again, that's a human decision.
(37:24):
We are building the connectionbetween we are the only company,
we have the dynamic pricingmodule works with advertising,
but I still feel like the humancan have a better uh decision
making uh against thecompetition when it comes to a
decision of maybe I should do alittle bit less advertising and
(37:47):
maybe I should reduce the pricenow.
I mean, that's kind of adecision of that human.
That's why I'm telling the PPCexperts now, you need to run
like you need to be responsiblefor pricing and promotion.
You cannot just go away withlike I I did my best with the
ACOS.
Well, maybe not.
Maybe this tool, this product,price sensitive.
(38:09):
Maybe you need to reduce theprice a bit.
And your ACOS then go downbecause your conversion rate
goes up as a result.
That's why that's myrecommendation.
The PPC expert, now you need tobe the pricing and promotion uh
expert too.
Like you need to combine thisbecause we are doing 65% of all
the mundane tasks for you.
You don't need to deal withthat.
SPEAKER_01 (38:30):
I like I really like that.
And in a situation I think a lotof people are in is they'll hire
a PPC manager for 20 hours aweek, and 18 of those hours are
spent spent doing bidoptimizations.
And and so what we're what we'redoing by using technology is is
allowing them to work on higherreturn on investment things.
(38:51):
And and that's that's really,really good.
Now, hi, the Ava is doesn't justdo PPC.
There are several more featureson the platform.
Can you talk about those alittle bit?
SPEAKER_02 (39:04):
Absolutely.
So thanks for uh you know askingme that question, too.
So one of the things that we doreally well for more than 10,000
sellers right now is therecovery services.
Uh for the uh sellers, uh itmeans the FBA reimbursements,
and we have a uh success uh feerate of 9%.
(39:27):
So basically, whatever werecover, like whenever we
recover the funds, we take 9% ofit.
We also do that for Amazonvendors, uh, which is the vendor
funds recovery, which is nothingto do with like FBA, but it's
more the financial disputes andkind of um you know raising that
again like to Amazon and gettingthe vendors' money back.
(39:50):
So that's kind of a big deal.
Uh we also provide like um youknow SEO graphic design services
for some of the customers.
Like the reason why we do thatis again, I just want to, you
know, like as the AI guy, well,when we work with a company uh
or an agency, and uh if like ifI don't feel like, for example,
(40:14):
the listing is not optimized orgraphic design doesn't make
sense to me, then I mean thechances to be successful with
with advertising is limited,right?
Like that's what I said in thebeginning.
That's why I came up like evenyesterday, I was having this
huge contract conversation witha supplement brand.
(40:35):
And uh at some point I said, youknow what, I'm gonna give you
the graphic design for free.
And like the guy was like superuh excited, like you know,
because he thought like thenegotiation is done and now I'm
giving more.
Because I said, I'm gonna giveyou like a hundred hours of
graphic design.
And he's like, Are you crazy?
Like, why are you doing this?
(40:56):
I said, because if I don't do itwith this listing, you're gonna
fail anyways, and then you'regonna blame me.
So, I mean, I'm getting enoughmoney for you from you for the
advertising.
I'll give this to you because weneed at least like all the six,
seven agents.
We need to do all this um imagedesign, you know, visualization
and all that stuff, like brandstory, A content.
(41:19):
We need to do that.
So we need to increase thisconversion rate, right?
So that's why we have thesupplementary services around
advertising and recovery, whichare the two main things that we
focus on.
When it comes to advertising, wealso believe that there is some
good software tools too, butthere is no support.
So we have this unlimitedstrategic calls.
(41:41):
We assign a dedicated person toevery single brand or agency.
So even the agency like they gotlike, why are you assigning it?
We say, no, we need to assignyou a PPC expert because we need
to do this right for you so thatyou can then go and you know,
kind of like do it for yourcustomers better.
Or if you work with a brand, thesame, they sometimes say, hey, I
(42:04):
know enough.
And then no, like we need tohave this expert for you.
So it's a it's a combination oflike two things, three things
when it comes to advertising,the configurable AI, uh, the
unlimited support, and theorganic focus, organic
growth-focused advertising.
This comes like thedifferentiators of our software
(42:25):
plus the people or humanapproach.
SPEAKER_01 (42:29):
Absolutely.
Absolutely.
Well, no, good to hear about allthe other aspects of Ava.
And um you have a special offerfor all the firing the man
audience.
Uh, can you share that?
SPEAKER_02 (42:43):
Sure, yeah.
So I think, you know, I'll gofor two special offers here, not
even one.
So the first one is the easything, this reimbursements.
Let's, you know, like we havethe 9% uh commission rate that's
already the best in the market.
We do it for 10,000 customers,right?
But on top of that, I would liketo offer an additional$500 that
(43:10):
is kind of reimbursed for free,just for them to get the taste
of it.
And the rest we can handle that.
You know, you don't need to doit internally or anybody else
with anybody else who is you youpay 20-25%.
Uh my second offer is on theadvertising side.
So up to$10,000 at spend, ourplatform plus the experices is
(43:35):
$499.
That covers like a lot of thesellers, but then if you're
doing more than that, it stillmakes sense.
But we're gonna give a$500 offon that software services too
for anyone who wants to use ourum Eva AI platform for
advertising.
SPEAKER_01 (43:55):
Outstanding, outstanding.
And I'll post links to all thatin the show notes.
Hi, before we wrap up the show,we have something called the
fire round.
It's four questions we ask everyguest at the end of the episode.
Are you ready?
Sure.
All right, what is your favoritebook?
SPEAKER_02 (44:12):
Yeah, so my uh favorite book is The Capital
from Thomas Pickett.
Very nice.
SPEAKER_01 (44:20):
What are your hobbies?
SPEAKER_02 (44:23):
Um, biggest hobbies are bodybuilding and swimming.
I do uh, you know, like I doswim every second day or and
order bodybuilding, you know,like weightlifting and things
like that.
SPEAKER_01 (44:34):
Very nice.
What is one thing you do notmiss about working for the man?
SPEAKER_02 (44:41):
Um, you know, it's um I I I just don't miss to
report something on a specifictime frame, you know, which I
don't miss at all.
Yeah.
SPEAKER_01 (44:58):
Oh, very good.
And and final question what doyou think sets apart successful
entrepreneurs from those whogive up, fail, or never get
started?
SPEAKER_02 (45:08):
Um success from my point of view is is there any
progression today?
Now, if you feel like you areprogressing, don't stick like
don't uh fixate on some success,some target that you need to
achieve.
Ignore that.
The key point is is there anyprogression?
(45:31):
Because if there is noprogression, you need to think
twice.
There is progression, thenforget it.
You will progress, you willcome, Louis, you will reach any
target if you are progressing.
SPEAKER_01 (45:43):
Very nice, very nice.
Well, hi, thank you so much foryour time today and looking
forward to staying in touch.
SPEAKER_02 (45:50):
Thanks, David.
Thanks for having me.
Welcome everyone to the Firing the Man Podcast, a
show for anyone who wants to betheir own boss.
If you sit in a cubicle everyday and know you are capable of
more, then join us.
This show will help you build abusiness and grow your passive
income streams in just a fewshort hours per day.
And now your hosts, serialentrepreneurs, David Schomer and
(01:29):
Ken Wilson.
SPEAKER_01 (01:32):
Welcome everyone to the Firing the Man podcast.
Today's guest is Hi Meg, thevisionary CEO behind Ava
Commerce, or Ava.guru, one ofthe fastest growing AI platforms
in the Amazon and D2C ecosystem.
Under his leadership, Ava hasbecome a top 10 global
(01:53):
advertising technology solutionpowered by an over 100-person
development team drivinginnovations across PPC
optimizations, reimbursements,and profit analytics.
High is a true pioneer in theAI-driven growth for e-commerce
brands, combining data science,automation, and practical seller
(02:15):
experience to help thousands ofbrands improve margins, boost ad
efficiency, and uncover lostprofits.
Whether you're an Amazon veteranor a D2C founder looking to
outsmart your competitors,you're going to walk away from
this conversation withactionable strategies for
organic growth-drivenadvertising and scaling
(02:37):
profitability in 2025 andbeyond.
Hi, welcome back to the show.
SPEAKER_02 (02:44):
Hey, thanks for having me, David.
It's absolutely a pleasure.
SPEAKER_01 (02:48):
Absolutely.
So we had you on the show backin 2022.
And I remember in in my Rolodexof contacts, you were one of the
first people talking about AI.
And I, at that point in time, itwasn't even really on my radar.
And so uh my question is is howhave things changed?
(03:09):
How has the e-commerce landscapechanged?
And how is AI playing a uh rolein that?
SPEAKER_02 (03:16):
Well, it's kind of a funny story, but I'm talking
about AI, being a computerengineer with an a master's in
AI.
I'm talking about AI from '95almost, like 1995.
And I have to tell you, nobodywas listening to me at the time
because nobody believed in it.
And now it's 2025.
(03:37):
Now everybody believes in it,but there's so many people
talking about it, so much noise.
Now nobody's listening.
Yeah.
Because like too much noise,right?
But I mean, a lot of thingschanged and a lot of things are
changing.
Uh, I just want to give you thisidea.
I'm not gonna expend too much onit, but my son, who is in the
(03:58):
high school now, and he was likelike two years ago, he told me,
hey, maybe I should become acomputer engineer like you.
And I told him, son, it's youit's your decision, but I am not
sure if that particular uh jobwill exist in a decade.
And he was like, What?
(04:19):
You know, and I said, Well, Idon't know.
Because the AI is kind of likeum really going faster and
faster, and it's so amazing likeum how it evolved so far.
But I have to say that sixmonths later, that will be like
two times more things will come.
(04:39):
So that's the evolution of AItoday.
SPEAKER_01 (04:42):
Absolutely.
Uh since 1995, that'sincredible, and uh we'll very
good.
And and to all of our listeners,that's the um that it that is
the the educationalunderpinnings of high and why we
love having him on the show.
So uh I today's conversation, Ireally want to talk about
advertising on Amazon.
(05:03):
And this is something that hasdefinitely changed over over the
years.
Uh I remember when I firststarted selling on Amazon, you
could sell at a you could have ahundred percent organic sales on
certain products.
And it really has has changedinto a pay-to-play platform
where PPC has to be part of yourstrategy.
(05:24):
When you're pricing things,pricing in a taco slime item is
is a necessity.
And so if you know, from fromyour standpoint, hi, how how has
Amazon advertising changed?
SPEAKER_02 (05:38):
Oh, absolutely.
So, first of all, as you saidbefore, it was all about people
who bought this product, alsolike this product, type of nice
stuff that was available onAmazon, no ads.
And when the ads started, likeif you're selling a supplement
brand and you can just uh bid onvitamin C and you can make a lot
(06:01):
of money out of it.
Now, most of this is pretty muchgone.
Uh, what I'm seeing is like theway it evolved today is first of
all, that's the only way to growyour business.
Like, I mean, all the other waysare pretty much like
supplementary to that.
Or the other ways aresupplementing your advertising
(06:22):
strategy because you would liketo increase your improve your
CTR or CVR, definitely you'regonna invest on SEO or graphic
design and all that stuff.
But there are like two mainways.
I wouldn't say the only way, Ijust say that two main ways to
grow the business.
One is definitely theadvertising, and the other one
(06:42):
is the off-Amazon marketing,pretty much not the advertising
angle, but the influencer angleto that, right?
So these are the kind of the twothings that I see.
But when it's on Amazon, becauseit's a very data-driven
platform, now it has to bereally smart.
You know, like the the wholeadvertising needs to be smart,
(07:05):
and uh that's where the AI kicksin, right?
Like, and uh like the the onlychallenge is like maybe you
know, like we should go back toalso explain what it means
AI-driven advertising, becauseone of the things that I feel
like almost every softwarecompany and agency is telling
(07:27):
that they are AI powered,AI-driven.
Is that so?
Like, I think maybe we shouldstart with some definitions.
SPEAKER_01 (07:36):
Absolutely.
Absolutely.
So let's go.
Let's let's get into it.
What um what does AI-driven meanuh in the context of Ava and
what you guys have developedover there?
SPEAKER_02 (07:48):
Well, absolutely.
So, first of all, uh, I mean,from a kind of a terminology
point of view, from a conceptualpoint of view, not from a
marketing point of view,AI-driven system means the
system which is trained by thepast data and learns every day
(08:10):
uh from like the past data andthe decisions and the results of
the decisions.
So, meaning that the maindifference between a kind of a
rule-based system and AI-basedsystem is on a rule-based
system, you basically say, ifthe color is yellow, then do
(08:31):
this.
Now, the rule doesn't change,where on the AI system, you
basically end up seeing thatmaybe the yellow is not like if
it's yellow, do A, and then youend up learning, like the system
learns that A is not correct andswitches to B based on learning
(08:53):
from the outcomes.
So that's the kind of the maindifference.
One is learning and one is notlearning.
Now, the not learning system,which is rule-based, is what we
call AI-based in the market.
Like 98% of the softwares thatare available are rule-based,
(09:15):
which is nothing wrong withthat.
Okay, like pretty much all thecompanies that you know in the
market right now, more and moreI'm seeing that they are not
talking about AI, they sayautomation, etc.
But it's all rule-based.
Nothing wrong with that, becausethe rules make sense, but also
it is uh it also makes sense tobuild an AI-based system which
(09:41):
learns from the data and adjuststhe rules too.
SPEAKER_01 (09:47):
Okay, okay.
I'm excited to to dig into uhinto that in the context of PPC.
And one of the things that Ithink may be helpful well, let's
let's dig into that in in interms of PPC.
What okay, and maybe we coulduse an example, um, say like Dog
Bull or something like that.
(10:08):
When people are are setting upcampaign infrastructure and
they're doing it manually, uh,what are the things that they're
missing out on that an AI-driventool could help them with?
SPEAKER_02 (10:23):
Absolutely.
So I'll come to that point, butbefore that, let me explain uh
what we did with EVA because weare believers of AI, but we
realize that it is not about thesystem only deciding what to do,
because that's kind of a blackbox approach.
(10:45):
And over the last couple ofyears, even we started with uh
our journey with pricing, and uhthree years ago we were talking
about dynamic pricing with you,David.
And uh for many years thecustomers said, we don't like
the black box approach, we don'tlike what like without knowing,
(11:05):
doing things that like AI,basically it's like, why did you
do it?
And the the answer is AI, right?
So, and then it's notsatisfactory for many of the
Amazon sellers because they knowwhat they are doing too.
They want automation, but theyknow.
So we came up with the idea ofconfigurable AI, meaning that
(11:26):
the AI works, but you defineyour own rules, and the rules
override the AI.
We kind of like give thisanalogy of like today the 25 AI
technology is called the AIagents.
Basically, you create your ownagent based on the foundation of
(11:46):
Eva AI, but then you define yourown rules so that you kind of
like address like four mainpillars of automation.
Now, one of them is likecampaign creation, and maybe the
other one is bid optimization.
I'll talk about the other uh twopillars.
But let's think about thecampaign creation and the bid
(12:10):
optimization.
So, first of all, back to thethe example that you gave, the
system already it knowseverything.
Like, I mean, in the past,people use tools like Helium 10
or some other tools, JungleScout, etc.
Try to understand like whichkeywords, what is the search
volume, but this data is nowavailable on the Amazon uh you
(12:35):
know reports.
Now the question is, do youreally want the PPC guy to go
into these reports like searchterm report, SQPR, etc., and
then figure out like whichkeywords are now appearing as a
result of an automated campaign?
And now let's kind of create anew campaign out of that, right?
So that is where the EVA AIautomates based on uh, you know,
(12:59):
the Amazon data, and based onlike um, you know, basically we
have access right now to almost10,000 brands data and uh brings
it together and decides whetherit makes sense to create a
campaign or not.
Now, let's assume that the EvaAI already decides creating
(13:20):
these campaigns and kind ofthinks that if a campaign, if a
keyword has 200 impressions, um,I should create a campaign.
So let's say that's the Eva AIway.
Okay.
Now, as a business owner, youthink that, you know, the profit
level that you have, or thecategory or your experience
(13:44):
basically says that, I mean, youwant to create campaigns when
there is more than 500impressions.
Now, on top of it, you are evenmore cautious.
You want to add, like, also, Iwould like to see at least two
organic orders.
I don't know.
So you basically define whateverrule you want to define based on
(14:06):
50 parameters that are allavailable, anything that is
imaginable on Amazon that youcan define a parameter like C
VR, CTR, ACOS, tacos, revenue,organic sales, whatever.
Like think about all of them asavailable.
You define your own rule and youoverwrite AI.
If there is a rule defined, thenyou overwrite the AI.
(14:29):
So that's kind of our designprinciple.
Uh, you, as the owner, definethe rules.
If you don't define any rules,AI works.
If you define the rules, you'lloverwrite the AI.
The combination of that is yourspecial uh store-specific AI
agent, which works on yourterms.
(14:52):
That's the idea.
SPEAKER_01 (14:53):
I I really like that.
And and the the AI plus humanelement is something that I
think is is really critical.
Um, I've talked about thispublicly on the podcast.
I do it again.
I ran Perpetua for a while.
It's an AI-driven software, andI always felt like I had an
invisible hand in my account.
(15:13):
And when it came to some of thekeywords that it was bidding on,
I I was like, what this hasnothing to do with my product.
And I also I felt as a businessowner like I didn't have a lot
of say in how the automationswere performing.
And so I I like that you get toset your parameters and what
(15:37):
would be you know, I think PVCis very easily an overwhelming
topic.
Like there's it's something thatyou either go very deep in or
you're overwhelmed by it and youhire somebody.
And and so in in terms of youknow coming in and setting up
rules, what do you think aresome of the the most important
(15:57):
things for business owners to belooking at or to to set up?
SPEAKER_02 (16:02):
Absolutely.
So we basically meet with somany different types of business
owners, or even like the PPCexperts that work for other
companies.
Like obviously, their level ofexpertise is very different.
Now, uh one of the designprinciples that we have is like
(16:23):
this the tool needs to serveanyone, meaning that if you're a
business owner and you haveabsolutely no idea about what
you are doing with PPC, yourlevel of expertise is zero out
of ten.
At least you know what's yourbudget, like how much money you
want to spend, what should belike an ACOS or tacos.
(16:46):
At least you can set one of themand you can do it at a store
level, product level, or acampaign level, but probably
campaign level is already toodetailed, but at least on the
store level and product level,you set up your business idea,
like you can set your businessideas and you're done.
Then the rest will be done byAI.
(17:09):
Now, the biggest thing here isthat there is a tree-level
configuration possibility,meaning that it's not only at
the store level, but also at theproduct level and the campaign
level.
Because a lot of businessowners, I always tell them if
you're running your Amazonbusiness, your business is not
at the store level.
It's actually a combination ofsome of your products.
(17:31):
So it's actually the portfoliosare the ones that you are
running, meaning that eachportfolio may have a different
budget and a differenteco-stacos target.
One is like a newly newlylaunched product or it's a
launch product, one is likemaybe a profit maximization uh
portfolio, and the other one islike a couple of products coming
(17:54):
together because they are prettymuch the same.
So, meaning that you need tohave a system that you can
define your portfolios in a veryflexible way.
That's why we have a verysimple-to-use tag management.
So you basically create yourcustom tags on your products,
and now, voila, you have thatportfolio which is tagged, and
(18:16):
now you can define your rolebased on the portfolio.
Now, that is for a businessowner.
It's really two things.
They can tag the products andthey can configure uh the
business parameters at theportfolio level.
But if you're a PPC expert, nowyou want more.
We basically identified fourareas of automation.
(18:39):
One of them is definitelynegative keywords or negation,
like you mentioned before.
And now there is the bidoptimization, um, automatic
campaign creation, day parting,and uh keyword negation.
So these are the the four mainareas of automation that we see.
(19:00):
Now, the system is alreadyautomating uh with AI, okay, so
which is learning andautomating.
But the the biggest thing is nowyou can define any rules as a
business owner or the PPC expertat any of these main four
pillars.
So you can also say that, hey, Iwould like to negate these words
(19:23):
specifically, or you can also goback and like, you know, change
the bid optimization.
Or the system suggests you, forexample, not like you know, I've
seen a lot of agencies liketalking about day parting, they
say, hey, based on myexperience, between, I don't
know, like midnight and 5 a.m.,uh, there is no sales.
(19:44):
Well, but it's a genericstatement.
Now, in our case, the systemsays, based on the data, here is
the suggestion.
Between 2 a.m.
to 4 a.m., you know, like do notuh, you know, spend money on
ads.
But between 6 p.m.
and 7 p.m., do.
And it depends on the product.
(20:05):
It doesn't, it's not like a verygeneric statement.
So that's the day parting thatsystem already works on it, but
the owner or the PPC expert canoverride it by defining that
specific hourly rule as well.
SPEAKER_01 (20:19):
Okay.
Uh you talked a little bit aboutcamp or campaign creation, and
and I'd like to talk about thatand in campaign infrastructure.
I it seems like if you were totake a PPC 101 class, one thing
that may be covered, and maybethis is outdated, uh, but is the
different types of campaigns.
(20:40):
And I would say that in thisclass, this was what was taught
probably three or four yearsago, is when you launch a
product, you set up an autocampaign, and what that's going
to do is it's going to harvestkeywords for you.
And when you have uh so manyconversions or orders, you take
those keywords and you put theminto a manual campaign.
(21:01):
Now, there are so many moreoptions.
You've got exact phrase, uh,match, broad.
And so going through a productlaunch and then just kind of
sustaining a product product'slevel of sales, what do you
think is the ideal campaigninfrastructure?
SPEAKER_02 (21:21):
First of all, as you mentioned, like all this keyword
isolation and uh creating themanual campaigns from the
automated campaigns.
This, I think, alone is like20-25% of a PPC expert's job
because there is always newproducts, there's always new
(21:41):
keywords coming in, and you needto always have that count
discussions and checking at thesystem and everything.
Now, we basically emulated it100%.
So everything happens throughthis automation.
Now, in this case, I wouldn'tcall it specifically AI, but
this is like pure automationbecause it's already clear.
(22:04):
Well, you can still decide onhow this automation is
triggered.
And the other thing is like it'salways not that easy for a
person to, for example, create acampaign and assign just one
keyword to that campaign, andthen also decide like whether it
is an exact or broad and etc.
(22:26):
Now, the system can decide on itbecause when the system sees
like uh somebody is looking fora light bulb, but it's 60 watts.
Okay, now there's always like Ilove this example because that's
one of our customers, and it'svery clear people actually don't
search for light bulbs.
(22:49):
They search with that specificuh information because there's
so many different light bulbs.
So it's an exact search.
Now, the AI also knows that it'san exact search because looking
into the search terms, it'salready pretty obvious that that
is the search term, and there isno light bulb search term.
(23:10):
Maybe even you negate that ifthat makes any sense.
You know, it depends on thecontext.
But the idea here is like thesystem kind of starts if knows
to decide decides like whetherit's exact or broad.
In some cases, we let the systemrun the campaigns for some time,
(23:30):
and then the ones it's kind of anatural selection process in a
way that the ones which are notconverting after a day or two,
they die.
So that's also another way oflike, you know, like the way
that the system kind of decides.
Now, the complexity is huge.
Like if you leave it to a manuallike do done by a PPC expert,
(23:53):
um, well, we still need the PPCexpert.
Don't get me wrong, we're notgetting rid of that PPC expert.
But this particular task or dayparting, like given that um for
a product, there might be ahundred campaigns.
And if you want to do dayparting and optimize the bits,
like we're talking about everyday thousands of updates.
(24:17):
I mean, who is going to takecare of like every single thing?
That's why I think AI is notoptional, it's like almost like
mandatory.
It's it's the the core of thebusiness is done by the AI plus
your rules.
Again, I'm talking about theconfigurable AI, right?
(24:37):
And uh I would talk about thePPC expert and what we think
that guy should do, but I hope Iat least answer your question
for now.
SPEAKER_01 (24:48):
Absolutely.
And and I what one thing I Iwant to clear up is is on on a
particular uh single product,what campaigns are would
typically be running?
Is it an auto, broad, phrase,and exact?
Is it is there or does it justdepend on the product?
SPEAKER_02 (25:11):
Depends on the product and depends on the uh
the success, right?
I mean, I think the system,unless it is pre-defined,
because in some could categoriesand some of the search terms
immediately end up AI to run theexact campaign, but in some of
the other cases, it will makemore sense to run these
(25:33):
campaigns and then see theresults and then kill some of
them.
You know, like that's what likethe AI is doing, uh, which you
know is also another possibilityas well.
Of course, um there is like somany different parameters here,
right?
Like branded versus non-branded,or is it an awareness or
(25:54):
consideration or conversioncampaign?
Because these are different thanlike, you know, like one of the
big mistakes, right?
So people say, oh, ACOS too highon this campaign, kill it.
Like, you know, people do thisadvertising audit.
I love it.
You know, like the agencies doadvertising audits, and then
they basically say, Thiscampaign, we need to kill it
(26:15):
because it's 10%, it's uh 50%ACOS or something, right?
Well, the the question, well,this is an awareness campaign,
right?
Like the ACOS level versus akind of a converting campaign
might be different, right?
So the system kind of looks intoall these uh multi-dimensions
like uh broad, exact, match,plus, branded, non-branded, plus
(26:41):
um where is it like what type ofa campaign is that in terms of
the marketing funnel, and thendecides based on that?
It's where it becomes reallythat's the main complexity
because of the multidimensionalum decision-making process, it
becomes impossible for a humanto decide.
(27:03):
So, what human does is onlydecide based on maximum two
dimensions, and like we we callthat guy like the best PPC
expert, but even that guy willignore some of the other
dimensions when deciding onthat.
SPEAKER_01 (27:21):
Absolutely, absolutely.
You were talking about thedifferent pillars of PPC, and in
bid optimization was one.
And if if you talk about just amanual way of optimizing an
account, you know, you'll you'llclick on a manual campaign,
there'll be 20 keywords there,and there'll be 10 of them that
aren't getting any impressions.
(27:42):
And uh let's just use dog bowlas an example, it's not getting
any impressions.
Amazon suggests that you bidbetween 75 cents and a dollar,
and you ask yourself thequestion, well, if I could get
it for 75 cents, why would Iadjust my bid to a dollar?
Um and so you make youradjustment and then you come
back two weeks later and see ifyou have any impressions.
(28:04):
And so how does how does thetool optimize bids and how does
it look at increasing bids ordecreasing bids?
Because that is kind of a scarything.
You are giving a machine yourcredit card to go spend money on
your behalf.
And and so um how does thesystem work with bid
(28:27):
optimization?
SPEAKER_02 (28:28):
Absolutely.
So I mean, I agree with you.
Um obviously we always thinkthat the best thing here is to
combine the PPC expert and theAI agent, right?
I mean, I'm not I don't believethe software alone approach.
And I would like, I mean, as aguy in the industry, a computer
(28:50):
engineer who is like very happywith AI, you know, taking care
of a lot of tasks.
But I don't believe that if youjust let your system to run by
AI, you will not go anywhere.
First of all, let me explainthis one.
Because a lot of this, you know,you asked the question of the
bid optimization, right?
(29:11):
But a lot of these decisions isthat are these decisions really
depend on your data or thecontext?
Meaning that think about likewhat can go wrong.
Here you go.
Like, let's say your competitoris out of stock, and another one
just increased the price.
(29:32):
As a result of that, now you areselling more, and people are
even clicking more to your ads,and the conversion even
increased, everything is great.
And think about the AI solution.
The AI solution will think that,hey, I am doing great.
I mean, you are not doing great,it's the context that is doing
(29:54):
that's why the uh autonomous carcompany, that thing, already
like you know, companies likeWaimo, Google, etc., they spent
trillions of dollars and stillit's not there yet, right?
There is like maybe a couple ofyears to go, like, but it's
trillions of dollars.
(30:15):
In the beginning, it was like,hey, in one year, like the
autonomous cars will beeverywhere.
Why?
Because the context parameters,now it's the same thing.
The context is changing, it'snot the same context.
And the AI is looking at yourdata.
It's it's hard for AI to knowthe other contexts unless there
is some scraping, etc.
(30:36):
Fine.
You know, we do, I'm not sayingwe do it, but we are able to get
third-party data and do someranking-based campaigns because
we want to do organic growth,etc.
But still, you cannot controlall the other parameters around
you.
Like, even maybe somebody islike creating an attack campaign
and reduce the price one dollarbelow you and attacking you, how
(30:59):
do you know?
Like if you're the AI system,like you don't know.
So, what I'm saying is that'swhere the PPC expert should.
Focus on.
We are pretty much taking careof 65% of all the tasks of the
can the PPC expert shouldn't betoo much focused on the bid
(31:21):
optimization on everything, butreally focus on that competitive
intelligence.
Because that's where thedifferentiator of the PPC expert
is from our point of view.
When it comes to the bidoptimization, again, in the
system, it's not like you know,just because you say, hey, spend
$10 on this bet, the system cando it, right?
(31:43):
Because even the Amazon systemdoesn't work like that.
Like if you spend a lot of, likeif you say$10 per bid, but the
CPC is$2, like probably you'llend up somewhere spending$2,
anyways.
But then again, uh the mainthing, and by the way, you know,
this is an evolving strategyhere.
(32:05):
And uh the the problem withoptimizations is what if like
you spend a bit more and you'llbe in the top of the search?
And what if you are doing uh asuboptimal strategy, meaning
that you are suboptimizing thebit, meaning your ACOS is fine,
(32:26):
but you're you're in the you'rein the third row.
And maybe if you put a bit more,you'll be on the first row, and
that's where I am very muchfocused right now, because um I
can tell you that is like Idon't see any risks today, like
even a human as human itself ismore risky than the Eva AI tool,
(32:49):
and Amazon doesn't allow manythings, but you always have a
PPC expert, or the system hasall types of alerts, like you're
not gonna spend like a hundredthousand dollars.
There is no way.
But uh the main thing that I amfocusing on is like even the
people, the the PPC specialistor the system always looks into
(33:11):
Acos, tacos, C VR, CTR, but kindof ignores what happens if I am
on Amazon.com, I search for thatkeyword, and where am I?
Like, where am I exactly?
And how can I go maybe on top?
Like, because that's not what AIdoes, okay?
It's kind of a differentstrategy here.
(33:32):
Like it's not optimization, it'slike I need to bid for that and
then get to the top of thesearch, right?
So that's a bit different.
That's where we want to, we arefocusing our RND right now.
Uh we are almost there, butthat's a big deal to achieve
that.
SPEAKER_01 (33:48):
Absolutely, absolutely.
And and I have to say, it'srefreshing to hear a company
founder that's that's AI driventalk about the human element,
uh, and that that as we standhere, it's a necessity.
Um you had made the comment onat the beginning of the episode
how there's a lot of AI-poweredcompanies, uh, and to people
(34:09):
tuning in on audio, I'm doingair quotes here.
Uh, but what is that what doesthat really mean?
And I I I would use the exampleof just just because you give
somebody a hammer doesn't meanthey can build you a house.
And and in in this case, um, youknow, uh just having the the
hammer being the technology, uhthat there need there needs to
(34:34):
be the human element as as westand here talking.
And so um we promised to deepdive on PPC, and I think we're
delivering.
Hi, are there any other uhPPC-related topics that you
think would be worth talkingabout?
SPEAKER_02 (34:49):
I just want to uh emphasize the point of human,
right?
Like, especially like now, likethere's of course the
search-based campaigns.
I mean, there's also likeretargeting and remarketing
stuff, but then there's theaudience-based campaigns.
I mean, when it comes to theaudiences and creating the
audiences, um again, you needthe human element because that
(35:15):
is not done automatically.
What is automatic is like theAmazon Marketing Cloud or even
our system can give you theideas, but at the end, the human
decides like what makes sense uhto bid for when it comes to an
audience-based marketing.
And that's why, for example, weare also a DSP provider, and uh
(35:38):
like uh our customers say, hey,great, the PPC automation is
great, we're gonna have this.
We understand now our PPC expertcan look after, I don't know,
instead of um one account, maybefive accounts.
What about DSP automation or DSPAI with DSP?
And I tell them, like, hey, Ithink pretty much all the DSP
(36:00):
automation uh is kind of likerubbish from my point of view
because that's all aboutremarketing, retargeting stuff,
which you can also easily do,which makes sense.
But the top the point is, do youreally need to buy a tool to do
that?
And then the real DSP value isgenerating that audience.
(36:20):
For example, you are a peanutbutter um like selling, like
you're selling peanut butters,and you're targeting the
audiences which purchased peanutbutter from your competitors
four weeks ago, with theassumption that now the peanut
butter is kind of like uh youknow it's not enough, and like
(36:41):
the the the consumer will buy apeanut butter in the next one
week, let's assume.
Now, this is kind of like thescenario that you generate, and
that is where we need human.
I mean, we need human to tothink about all these
differences.
That's why the competitivecontext, I think, is the most
(37:02):
like the biggest differentiatorof the human.
I just want to give anotherexample.
What if it makes more sense togive a discount or a strike
through price or a coupon orsome type of a deal instead of
maybe pushing more advertising?
Again, that's a human decision.
(37:24):
We are building the connectionbetween we are the only company,
we have the dynamic pricingmodule works with advertising,
but I still feel like the humancan have a better uh decision
making uh against thecompetition when it comes to a
decision of maybe I should do alittle bit less advertising and
(37:47):
maybe I should reduce the pricenow.
I mean, that's kind of adecision of that human.
That's why I'm telling the PPCexperts now, you need to run
like you need to be responsiblefor pricing and promotion.
You cannot just go away withlike I I did my best with the
ACOS.
Well, maybe not.
Maybe this tool, this product,price sensitive.
(38:09):
Maybe you need to reduce theprice a bit.
And your ACOS then go downbecause your conversion rate
goes up as a result.
That's why that's myrecommendation.
The PPC expert, now you need tobe the pricing and promotion uh
expert too.
Like you need to combine thisbecause we are doing 65% of all
the mundane tasks for you.
You don't need to deal withthat.
SPEAKER_01 (38:30):
I like I really like that.
And in a situation I think a lotof people are in is they'll hire
a PPC manager for 20 hours aweek, and 18 of those hours are
spent spent doing bidoptimizations.
And and so what we're what we'redoing by using technology is is
allowing them to work on higherreturn on investment things.
(38:51):
And and that's that's really,really good.
Now, hi, the Ava is doesn't justdo PPC.
There are several more featureson the platform.
Can you talk about those alittle bit?
SPEAKER_02 (39:04):
Absolutely.
So thanks for uh you know askingme that question, too.
So one of the things that we doreally well for more than 10,000
sellers right now is therecovery services.
Uh for the uh sellers, uh itmeans the FBA reimbursements,
and we have a uh success uh feerate of 9%.
(39:27):
So basically, whatever werecover, like whenever we
recover the funds, we take 9% ofit.
We also do that for Amazonvendors, uh, which is the vendor
funds recovery, which is nothingto do with like FBA, but it's
more the financial disputes andkind of um you know raising that
again like to Amazon and gettingthe vendors' money back.
(39:50):
So that's kind of a big deal.
Uh we also provide like um youknow SEO graphic design services
for some of the customers.
Like the reason why we do thatis again, I just want to, you
know, like as the AI guy, well,when we work with a company uh
or an agency, and uh if like ifI don't feel like, for example,
(40:14):
the listing is not optimized orgraphic design doesn't make
sense to me, then I mean thechances to be successful with
with advertising is limited,right?
Like that's what I said in thebeginning.
That's why I came up like evenyesterday, I was having this
huge contract conversation witha supplement brand.
(40:35):
And uh at some point I said, youknow what, I'm gonna give you
the graphic design for free.
And like the guy was like superuh excited, like you know,
because he thought like thenegotiation is done and now I'm
giving more.
Because I said, I'm gonna giveyou like a hundred hours of
graphic design.
And he's like, Are you crazy?
Like, why are you doing this?
(40:56):
I said, because if I don't do itwith this listing, you're gonna
fail anyways, and then you'regonna blame me.
So, I mean, I'm getting enoughmoney for you from you for the
advertising.
I'll give this to you because weneed at least like all the six,
seven agents.
We need to do all this um imagedesign, you know, visualization
and all that stuff, like brandstory, A content.
(41:19):
We need to do that.
So we need to increase thisconversion rate, right?
So that's why we have thesupplementary services around
advertising and recovery, whichare the two main things that we
focus on.
When it comes to advertising, wealso believe that there is some
good software tools too, butthere is no support.
So we have this unlimitedstrategic calls.
(41:41):
We assign a dedicated person toevery single brand or agency.
So even the agency like they gotlike, why are you assigning it?
We say, no, we need to assignyou a PPC expert because we need
to do this right for you so thatyou can then go and you know,
kind of like do it for yourcustomers better.
Or if you work with a brand, thesame, they sometimes say, hey, I
(42:04):
know enough.
And then no, like we need tohave this expert for you.
So it's a it's a combination oflike two things, three things
when it comes to advertising,the configurable AI, uh, the
unlimited support, and theorganic focus, organic
growth-focused advertising.
This comes like thedifferentiators of our software
(42:25):
plus the people or humanapproach.
SPEAKER_01 (42:29):
Absolutely.
Absolutely.
Well, no, good to hear about allthe other aspects of Ava.
And um you have a special offerfor all the firing the man
audience.
Uh, can you share that?
SPEAKER_02 (42:43):
Sure, yeah.
So I think, you know, I'll gofor two special offers here, not
even one.
So the first one is the easything, this reimbursements.
Let's, you know, like we havethe 9% uh commission rate that's
already the best in the market.
We do it for 10,000 customers,right?
But on top of that, I would liketo offer an additional$500 that
(43:10):
is kind of reimbursed for free,just for them to get the taste
of it.
And the rest we can handle that.
You know, you don't need to doit internally or anybody else
with anybody else who is you youpay 20-25%.
Uh my second offer is on theadvertising side.
So up to$10,000 at spend, ourplatform plus the experices is
(43:35):
$499.
That covers like a lot of thesellers, but then if you're
doing more than that, it stillmakes sense.
But we're gonna give a$500 offon that software services too
for anyone who wants to use ourum Eva AI platform for
advertising.
SPEAKER_01 (43:55):
Outstanding, outstanding.
And I'll post links to all thatin the show notes.
Hi, before we wrap up the show,we have something called the
fire round.
It's four questions we ask everyguest at the end of the episode.
Are you ready?
Sure.
All right, what is your favoritebook?
SPEAKER_02 (44:12):
Yeah, so my uh favorite book is The Capital
from Thomas Pickett.
Very nice.
SPEAKER_01 (44:20):
What are your hobbies?
SPEAKER_02 (44:23):
Um, biggest hobbies are bodybuilding and swimming.
I do uh, you know, like I doswim every second day or and
order bodybuilding, you know,like weightlifting and things
like that.
SPEAKER_01 (44:34):
Very nice.
What is one thing you do notmiss about working for the man?
SPEAKER_02 (44:41):
Um, you know, it's um I I I just don't miss to
report something on a specifictime frame, you know, which I
don't miss at all.
Yeah.
SPEAKER_01 (44:58):
Oh, very good.
And and final question what doyou think sets apart successful
entrepreneurs from those whogive up, fail, or never get
started?
SPEAKER_02 (45:08):
Um success from my point of view is is there any
progression today?
Now, if you feel like you areprogressing, don't stick like
don't uh fixate on some success,some target that you need to
achieve.
Ignore that.
The key point is is there anyprogression?
(45:31):
Because if there is noprogression, you need to think
twice.
There is progression, thenforget it.
You will progress, you willcome, Louis, you will reach any
target if you are progressing.
SPEAKER_01 (45:43):
Very nice, very nice.
Well, hi, thank you so much foryour time today and looking
forward to staying in touch.
SPEAKER_02 (45:50):
Thanks, David.
Thanks for having me.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Las Culturistas with Matt Rogers and Bowen Yang
Ding dong! Join your culture consultants, Matt Rogers and Bowen Yang, on an unforgettable journey into the beating heart of CULTURE. Alongside sizzling special guests, they GET INTO the hottest pop-culture moments of the day and the formative cultural experiences that turned them into Culturistas. Produced by the Big Money Players Network and iHeartRadio.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.