May 31, 2026 • 51 mins
We talk through what “AI visibility” really means and why marketers are racing ahead of the data, even as consumer behavior shifts toward AI search and AI assistants. We connect SEO fundamentals to GEO and agentic commerce so you can take practical steps instead of getting stuck in paralysis.
• AI visibility tools and why the industry lacks a baseline standard
• How context and personalization change AI search results
• Why LLM analytics is immature and often modelled on simulations
• The importance of data hygiene when AI tools disagree
• How AI assistants compress the customer journey into one surface
• Google AI Mode, AI Overviews, and why SEO still matters
• Agentic commerce and what it means when AI buys for people
• Simple starting steps like tracking, schema, and product feed readiness
• Avoiding low-value AI-generated content and focusing on usefulness
Guest Contact Information:
Linkedin: linkedin.com/in/andrewmhiggins
Website: parsnipp.com
More from EWR Digital and Matthew:
Leave us a review wherever you listen: Spotify, Apple Podcasts, or Amazon Podcast
Free SEO Consultation: www.ewrdigital.com/discovery-call
With over 5 million downloads, The Best SEO Podcast has been the go-to show for digital marketers, business owners, and entrepreneurs wanting real-world strategies to grow online.
Now, host Matthew Bertram — creator of the LLM Visibility Stack™, and Lead Strategist at EWR Digital — takes the conversation beyond traditional SEO into the AI era of discoverability.
Each week, Matthew dives into the tactics, frameworks, and insights that matter most in a world where search engines, large language models, and answer engines are reshaping how people find, trust, and choose businesses. From SEO and AI-driven marketing to executive-level growth strategy, you’ll hear expert interviews, deep-dive discussions, and actionable strategies to help you stay ahead of the curve.
Find more episodes here:
Rate the show in your podcast app. Apple ratings are how new listeners decide whether to press play — an honest one takes about thirty seconds. Steps for iPhone, Mac and Spotify:
www.bestseopodcast.com/review
Follow us on:
Facebook: @bestseopodcast
Instagram: @thebestseopodcast
Tiktok: @bestseopodcast
LinkedIn: @bestseopodcast
Connect With Matthew Bertram:
Website: www.matthewbertram.com
Instagram: @matt_bertram_live
LinkedIn: @mattbertramlive
Powered by: ewrdigital.com
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:03):
This is the unknown secrets of internet marketing.
Your insider guide to thestrategies top marketers use to
crush the competition.
Ready to unlock your businessfull potential.
Let's get started.
SPEAKER_01 (00:18):
Howdy, welcome back to another fun-filled episode of
The Unknown Secrets of InternetMarketing.
I'm your host, Matt Bertram.
We are talking about everythingAI marketing, AI visibility,
SEO, um, kind of how you dealwith the gentic.
Everything's very fluid.
Uh, wanted to bring in somebodyI know people have been asking
about e-commerce and agenticstuff.
(00:38):
That is not my wheelhouse.
So I want to be clear.
Um, I do know and have greatpartners uh that do a lot of uh
e-commerce and agentic stuff.
We're doing more B2B stuff, leadgen, that sort of thing.
Uh, and and so uh all right,Andrew Higgins.
Andrew Higgins, welcome to theshow.
He's got an AI visibility tool,he's doing e-commerce agentic.
(01:01):
Uh, we're having some differentkinds of conversations around
like, is LLM visibility likebaseline analytics, right?
Like, I think that that's uh areally interesting point that
everybody's trying to add AIvisibility, whether it be GA4,
uh, all the different uh goldstandard tools are offering um
(01:24):
LLM visibility, everybody'sdoing it a little bit
differently.
Um, but everybody's racing to doit.
And to your point, I feel likethere needs to be a baseline.
It would be great if there waslike a standard.
I don't know if Google's gonnacome out and say like it is.
I mean, they're switchingeverybody over to AI search uh
shortly.
So um I know it's gonna be moreof a conversation.
(01:47):
Um, but I but I would love tokind of get your landscape,
Andrew, on what what you'reseeing in the marketplace.
I know you launched a tool um uha while back around L
invisibility.
I know they're popping upeverywhere.
So I'd love to see what you'reseeing with that, with that, as
well as like kind of what isgoing on with e-commerce.
SPEAKER_02 (02:08):
Totally.
Well, thank you for having me.
Um I think everyone is isscrambling to try and figure out
the problem at the same time.
And to some extent, likemarketers are trying to build
the plane while they're flyingit, like pardon the old adage,
but I think the the reality iswhat we know, like if we just
figure out things we know,consumer behavior and the way
(02:29):
that people are digesting andusing the internet has wildly
changed on the back of AI.
The way we figure out whatrestaurant to go to or vacation
or what to buy our moms fortheir birthdays, like that
behavior, that customer journeyhas changed a ton in the past
few years and is changingrapidly on an almost weekly
basis.
(02:49):
So I think when we think aboutwhether it's AI visibility or
e-commerce or honestly any partof the digital marketing stack,
it all is being retooled and theplaybook is being rewritten.
So I think the the firstconversation that at an industry
level we're having is GEO or AIvisibility or AIO or AI SEO.
(03:12):
Like we don't even have thenomenclature right yet.
Uh I'm not convinced that's themarket.
I think that's where theconversation today is.
Uh I think that's for a couplereasons.
One is because we understand SEOand we had this monolith that
was Google search and governed alot of eyeballs and the
beginning of a lot of shopperjourneys.
(03:33):
Uh today, one, we have it's nota monolith.
There are lots of players, thereare lots of labs in large
language models, and even underthe hood of you know, Gemini as
an example, there's multiplemodels that behave differently,
that collect data differently.
And the biggest piece is it's nolonger just 10 links.
(03:54):
No context matters.
So if you and I have a similarinteraction with a large
language model, we will getdifferent results because of the
context of who we are.
And all of that changes the waywe start to optimize.
If we know people areinteracting with AI instead of a
brand or a retailer or socialmedia, we have to marketers, we
(04:16):
need to figure out how toinfluence those interactions,
but all of that is changing.
SPEAKER_01 (04:20):
Well, I mean, you you see it with like Rand
Fishkin and like what he'sdoing, right?
Like he like he made the pivotto like ICP.
So it's just like who who is theaudience?
Where are they at?
What is happening um before theyeven start searching, right?
Like what like it goes back,like pendulum back to you know,
(04:43):
mad men, right?
Like, like where what what isthe creative?
What is the brand?
What is the brand image?
Like, where is that brand showup?
Where does it want to be around?
Like, I I kind of even feel likeyou're like running the track.
Like you got to do reps, you gotto do reps, and you got to be in
those same areas consistently tobe viewed as that authority,
right?
And so it it's a full funneloptimization of where does your
(05:07):
brand show up?
Where is your audience at?
How are you speaking to them?
What questions are youanswering?
It all comes together with areyou communicating with your
target audience correctly andwherever they're at, uh, at
whatever their need is.
And and that's a much bigger askthan optimizing for top 10 links
(05:28):
in Google.
SPEAKER_02 (05:29):
Well, and I think the all the principles that you
just walked through are thesame.
Like AI doesn't change the coreprinciples of marketing.
We have new channels, we havenew use cases, we have new ways
we need to try and startmeasuring data.
Like, so principles are theexact same.
We just understand consumerbehavior is different, and it's
(05:49):
that's the reason I think AIvisibility is a big conversation
today.
I'm not sure it will betomorrow.
And reason being is step one islike understand the usage, under
start to try and figure out thechannel, understand how AI talks
about my brand, how it'sdifferent to my competitors.
And then the biggest thing is ifyou go, if you go all the way
(06:10):
back, you mentioned the Mad Menera.
Well, there used to be 10 TVchannels and some a few radio
stations and a few largepublications.
People read the newspaper.
So the ability as technology hascontinued to advance, we can
target our user.
And I think AI is another leapin the direction of your ICP is
becoming less and less broadbecause your ability to
(06:33):
personalize is stronger andstronger.
So today, when contextinfluences, you know, if you
know, how an LLM interacts witha shopper, then we as marketers
need to be able to get down tothat level of specificity as
well.
So I don't I think coreprinciples are still the same.
I think the playbook is beingrewritten, the tools, the stack
looks different.
But the hot topic today is thedata, which is AI visibility,
(06:58):
which I don't think will be along-term category.
I think it of course willmatter, but also the data is not
accurate yet.
Like I've that's the data isthis is benchmark data.
And if you use 20 differenttools doing something related,
you'll actually get 20 differentanswers, which is a problem.
SPEAKER_01 (07:15):
Well, you I mean, that even goes into like why the
models are hallucinating orwhatever.
It's like bad data, right?
And like even when I'm likedoing some programming stuff,
I'm going, like, why did yougrab that?
And then I'm like, go down alittle bit of rabbit hole to
like understand why it broughtthat in.
And like people aresensationalizing, like, oh, this
AI got um what was it?
(07:36):
Like I got got really interestedin like the Golden Gate Bridge,
or there was that like stuff,and it was like, well, if you
keep feeding the model, there'slike a decay rate and attention,
and like, of course, that's whatthe models do if you understand
the models.
But if you're looking ateverything black box, then you
know that that data hygiene andand understanding like the math
(07:59):
that goes into this is reallyquite important.
And I I believe I believe whatyou're saying because the
visibility pieces, like, we'rejust trying to figure out how to
measure, and then once wemeasure, it goes back into like,
well, is your marketing good?
Because that's the debate rightnow with everybody on social
(08:21):
media that I'm seeing is likegeo is completely different, and
then no, it's the exact same,it's just good marketing, right?
And it's it's this back andforth.
Then it's like, okay, it is goodmarketing, but how how to do it
and understanding that this is acompletely different technology
than Google, like Google worksdifferently on how it gets to
(08:43):
the answer, it might get to thesame answer, but how it gets to
it and understanding the theunder the hood is how you get
the competitive advantage toexecute, but good marketing is
always gonna win.
And and really, I feel likemarket feedback and and I think
that that's leads intoe-commerce is understanding the
(09:05):
user, the the the the customerhas changed is how how has
informing you on what you needto do, on how you need to speak
to them.
And if you have e-commerceversus like a B2B where it's
offline, online, like whatever,and you can't track everything,
you get cleaner data.
(09:25):
So data hygiene is like likefirst, right?
Are you measuring everythingright?
And e-commerce, you get to seekind of the full experience, so
you can optimize what'shappening and you get more data
in that process because it's youknow, you're you're targeting
the user, you're engaging withthe user, they're buying online,
(09:46):
and then you're drop shippingthem something or whatever,
right?
SPEAKER_02 (09:49):
Well, and I you're I agree completely.
I think we we don't have afeedback loop.
So it's if this is a newdiscipline of marketing, we've
got to create the feedback loopso you can truly understand what
works and what doesn't.
Um, but the interesting piece iswe don't have usage data yet.
So if we yes, we have tounderstand the technology and
understand how large languagemodels work so that we can start
(10:12):
to reverse engineer them tofigure out how as a brand we
optimize ourselves to bevisible, to be present, to be
discoverable.
But the state of the market,like the marketing discipline of
you know, AI search, GEO,whatever we're gonna call it, is
still quite immature in thatwhat we're actually doing under
the hood is it's predictive.
(10:33):
You're building a model of ashopper and having simulated
interactions.
That's kind of like buying aFacebook ad and meta saying, I'm
not gonna tell you the number ofimpressions, I'm gonna guess,
I'm not gonna tell you yourclick-through rate, and I'm not
gonna pass through a tag, butbuy my ads anyway.
And we wouldn't do that.
I mean, there was an era youknow, of TikTok where it was the
(10:54):
usage was so high that peoplewere willing to invest
advertising dollars in anexploratory way, but of course
you started getting B2B APIs soyou could measure these things,
and so you could close thefeedback loop.
We're still in the infancy wherewe don't have the feedback loop.
And that's I think what a lot ofmarketers are trying to
establish as a step one.
SPEAKER_01 (11:13):
Okay, so so one of the interesting conversations,
which this might be out outsideof this discussion, but I've
I've started to see it is I knowFacebook got in trouble.
I mean, they're getting introuble for a lot of stuff right
now, but um, they got in troublefor really faking impressions
like early on, like like they'reand and you wonder why like
(11:34):
these ads are not working theway they should.
Like, my I always went back tolike, are they showing them to
the right people?
Right, because they need to getthere, they're trying to hit
spin numbers, right?
So they keep expanding it out,expand it out.
Google does the same thing.
I saw a study recently thatcompared um running ads on the
Google platform versus usinglike a DSP or something like
(11:58):
that to buy ads for you of thetarget audience set.
Interesting, because Google isshowing.
I mean, you're paying like apercentage to run through the
DSP, but if you took the samedollar amounts, right, and go,
well, I'm getting 12% more uhjust doing Google and I'm not
using the DSP.
(12:19):
What's gonna be the result?
Yeah, the results from thisstudy again, like I don't have
all the details and don't quoteme on it, but it looked to be
really, really accurate thatGoogle was serving ads to
properties that benefited themmore than benefited the user,
and they had a limitedinventory, so like they're
(12:42):
forcing things that are are notthe best use of your money.
And if you if if if you havemore inventory to select from
and you kind of lock in whoyou're trying to reach a little
bit tighter, you'll always getyou're gonna get better
targeting, right?
Like it was it was kind of likeyou knew the the setup, but it's
(13:04):
so true, right?
And I feel like I mean now toyour point, measuring LLMs, like
different use cases.
Like when I see an LM and I'mlike, ah, we're in Claude and
like, oh my gosh, like peopleare stealing all our stuff, like
you know, like that's what Ithink, like, but like like you
you all of these differentchannels, right, or AIs, people
(13:26):
are gonna have differentrelationships, and I think the
battle is like which one's gonnaown, like, which one's gonna be
the one that is the one thateveryone uses.
SPEAKER_02 (13:36):
I don't care it might not matter though, it
might it that who knows?
I think we we there will be alot of evolution that might be
like saying you have to pick onesocial media channel to market
across, right?
That these might be use casespecific.
SPEAKER_01 (13:50):
Well, yeah, but uh so this is what I would tell you
these AIs store okay, chat GBT,for example, they came out and
said they're gonna remember,like I think like it's like
government surveillance,whatever, but like they're gonna
every single thing, just ifyou're listening, look in the
terms of use, every single thingyou put into chat GBT, they're
(14:11):
keeping forever.
I don't know if everybody knowsthat.
So if you're listening, thatthat they came out and said
that, not so like FISA warrant,whatever.
Like so, just know that ifyou're using Chat GBT or using
codex, every single thing you dois going in that now.
Then you got people, Andrew,that are uh taking open source
(14:32):
LMs, putting them on their ownservers, air gapping it, like
all their data is not leaving,it's not getting processed
because anything you put in anyof these tools is getting
processed off site.
So I think anthropic, I uh youknow, it was 30 days at one
point.
I don't know what they changedit to, but they they they clean
all the data every 30 daysunless like you store it.
But but perpetual memory, likememory is a big conversation,
(14:57):
uh, I feel like around thesedifferent AIs and how it recalls
information and like the ragpulls and all that.
So what people are finding thatare using these tools that are
consumers, not builders uh inthese tools, is they're getting
really comfortable with one AIbecause it knows them and it
learns them, right?
And they're comfortable with theinterface and they know how to
(15:19):
use it.
I I mean, I feel like I knowpeople that are jumping around
between different tools, yeah.
But the people that are not inthe marketing space, like like I
would consider just straightconsumers, yeah, they've started
to use one tool, they haven'tdone it, and right, and they get
comfortable with that, and andlike even there's like little
(15:41):
debates on like, well, I likeClaude better and I like ChatGPT
better, and then they're liketalking about how they like it
better, but no one's tried theother one.
And I feel like there'singrainment in that interface,
and the more things you pluginto it, right, and it gets
better data, uh the theswitching costs become harder,
right?
And right, and so I I meanagain, like we don't know what
(16:04):
the future holds, like whoknows?
Like, I mean, we don't even knowif there are aliens yet or not,
right?
But like AI, like I I would tellyou that I I think the battle
honestly is who builds the bestconsumer tool that everybody has
an AI that they filtereverything through because it
(16:26):
knows them so well and it helpsthem make decisions.
I mean, you're even seeing kids,you know, and like don't talk to
it like a human, and like Imean, it's causing like all
kinds of things, bad things tohappen.
Um, I I think that there's gonnabe a trusted agent, like or
whatever that you're workingwith that that is your primary
(16:48):
like assistant that you filtereverything through, which of
course you could build someadvanced things that plug in and
grab the different models andall that.
Um, and there's a tool outthere, guys.
I think it's called Poe.
I I don't really use it, but itit grabs all the different LLMs
and you can ask it a question ifyou're using it just as a chat
bot, not like as a developmenttool.
(17:08):
But um, like I just think peopleare gonna have their go-to and
they're gonna filter everythingthrough it.
So if that I I can tell youright now, I was looking at um
to buy something on Amazon orwhatever, and I was like,
screenshot it, put it in there,look at the reviews, like do
some additional research.
(17:29):
Yeah, tell me this and tell meif there's something else that I
um that I'm missing, and then itwas like, Oh, well, uh, you
know, you're you're wanting todo this.
Uh like it was like a supplementor something, right?
It was like, um, you're wantingto do a gummy, like that might
not be as whatever, and and Iwas trying to decide between
these like mushroom gummies orwhatever, right?
(17:51):
And then it basically was like,No, you need a capsule, and then
these are the brands, and thenthese are the like like I was
down this path, and like thesewere like through that like
consideration phase, yeah.
Like I was there, and then itjust pivoted me to go like, no,
you didn't think about thesethings, you need to figure out,
and then these are the threebrands that you should select
(18:12):
from, and here are the linkslike that.
Like, because you're okay, goingback to L invisibility, like if
like how did I get prompted withthat question?
And then it represented this,and someone answered it, and
then these three brands answerthat question the most
(18:33):
effectively and had the mostreviews, and like whatever else,
that it got shown to me,provided me links, and then I
bought something over here thatI wasn't even thinking about,
right?
Like, that's the world thatwe're trying to measure.
SPEAKER_02 (18:49):
Well, and think about how much of that journey
sat on one surface to your pointof like it sat, whereas the
fragmented like digital customerjourney, yes, you still
transacted on Amazon or whereveryou ended up purchasing, but
pretty soon that transactionwill also be able to take place
with uh with the AI buying onbehalf of the shopper.
(19:09):
So I think about you know, thecon I think about how long the
customer journey is is gonna beon one single surface.
So is this visibility is not theright question.
This is, and that there will beso many use cases built into one
channel.
You know, this is yes,visibility will always matter,
(19:30):
and you yeah, but you will haveto figure out how to stay
visible, you will have to figureout how to, as you get into
consideration and you get intocompetitive comparison and
return policies, you'll have tomake transacting easy.
You know, when you think aboutlike currency, you know,
currency across, you know, abunch of different models in
different markets, you know,where you have context on the
(19:51):
shopper and pricing promotionsin different regions, like this
stuff gets very complicated.
And we know the drop-offs we seeand conversion and engagement
when you do those things wrong.
SPEAKER_01 (20:02):
Yeah, well, I mean, yeah, I mean, relevancy is going
to be a big component, butyou're right, visibility
honestly doesn't matter.
It's like at what stage of thebuying process are you at?
Correct.
And now here's the thing to havewho like the work is what was
done to make that recall happenthe way it happened, right?
(20:24):
Like, so that's where all thelike invisible work comes in.
I um what was I gonna say?
Uh you were you were talkingabout something, and I was like,
man, I need to write this down,but um keep going, keep going,
and I'll think of it.
SPEAKER_02 (20:40):
But but that that is where I think the challenge
becomes exponentially harderbecause we're not optimizing for
one click and then owning thatexperience on our site, it's
very multifaceted.
SPEAKER_01 (20:49):
Oh, this is what I was gonna say.
So, Google, if you watch like umtheir, you know, every every
every big brand, uh, and I wouldencourage anybody that's a
marketer.
Uh, if you can't go to theseevents, um, and and I I I don't
go to all of them, but like youcan go watch them on YouTube or
whatever.
But Google's last like yearlyunveil, it's kind of like the
(21:10):
talk where Steve Jobs wouldalways come out and unveil like
the new product or whateverthey're doing.
So Google's big event, I I uhit's escaping me what it's what
it's called right now, but theywere unveiling AI search and
they were unveiling what we'retalking about, is and I think
that the reason they're doing itnow is they don't want to lose
(21:31):
uh customers to another surface,like or another, and so they're
switching to AI mode, right?
And that's why they're doing itbecause they already know that
the whole uh value funnel or orthe search funnel is gonna
collapse, and they want to keepit on Google.
And so if everybody uses uh youknow uh Gemini or whatever's
(21:55):
powering AI overviews, and thatyou can search on Google.
Google, ask all the questionsand buy, and they have this, you
know, huge market share, they'reprobably gonna win this next
evolution, right?
Correct.
Because there's not I thinkyou're muted by the way, Andrew.
I don't know.
Um, but uh there you go.
(22:16):
Okay.
And uh and uh I I I mean I Ithink they're making the switch,
they were prepping it, butbefore too many people are doing
the buying journey and some ofthese other platforms, they they
still have the market share.
People are still saying they'reusing AI because they're looking
at the overviews, right?
And like everybody feels good.
I think I think they're lookingat the metrics and and they're
(22:41):
gonna hard switch it because ifthey can capture everybody just
to do it on Google, um, they'llthey'll win the next uh round,
right?
And then all that customerjourney will happen there.
So I still think optimizing forGoogle is uh what needs to
happen.
SPEAKER_02 (22:58):
Um I also yeah, and I don't think they're
conflicting like tactics, whichis the I think Optima, even the
SEO versus geo discussion, likethey are they belong in the same
discussion, and there are thingsthat are uniquely positioned for
optimizing for AI that aredifferent, do not matter with
the way Google indexestraditional search, but there's
(23:20):
also a huge overlap, right?
And so I don't think they'remutually exclusive.
I think the the reality is thatlike the core hygiene uh that is
SEO will continue to be true.
There's now just a bunch ofother elements of building a
website, marketing online,working with PR and you know
that that matter and areingested in a different way,
(23:41):
whether through RAG or throughyou know, the way large language
models interpret information.
And also the way we search, justwith when it was when you had
one search and you had to winthat initial keyword or that
keyword phrase, and you youoptimize for fewer pieces.
Now it's you have to optimize,you know, along that whole
journey, and the way the way youdo that changes based on where
(24:02):
you're at in the funnel.
SPEAKER_01 (24:04):
Yeah, no, I I totally agree.
So, what are some other insightsthat you're seeing on the
e-commerce side that would behelpful to bring into this
conversation?
SPEAKER_02 (24:13):
Um, that it's coming first and foremost.
I think agent agentic commercealso even by itself is a broad
conversation.
SPEAKER_01 (24:20):
Like are you talking agent to agents?
SPEAKER_02 (24:22):
That's exactly what I think.
Okay.
So most of what I deal with on aday-to-day basis when I think
agente commerce, it is how ise-commerce going to be
integrated into large languagemodels?
Like, which is I think only asliver of the discussion because
the way in which agents aregonna work, you know, work on
behalf of shoppers is much moreplentiful than that.
SPEAKER_01 (24:43):
Okay, so for a while, anybody that listens to
me, I kind of went down this umuh cryptocurrency path.
Uh, and and we have some bigclients in the space uh that
we're doing we're doing geo workfor and all that.
But um I really see so there wasa protocol back in like 91-92
(25:05):
that the internet do you knowwhat I'm talking about?
Like the what it what sobasically there was a protocol
on the internet where you couldtransfer money uh online, but it
was never used.
I think this year they justturned it on, um, and they're
getting ready to move to digitalcurrency uh in the banks, and
(25:26):
there's some laws being passedand everything like that.
Because, well, I think that theagents are gonna want to use
like a Solana or like somethingfast to be able to transact and
uh fractionally buy.
And you're seeing it with APItoken usage as well, but I I
think it's even broader, right?
(25:47):
Like agent to agent commerce,which what I've read, it's gonna
be like 75% like eclipsing thecurrent e-commerce.
SPEAKER_02 (25:55):
The TAM projections are astronomical, yeah.
SPEAKER_01 (25:59):
Right.
And so, like, it's not even it'slike the agentic e-commerce.
Okay, that's like I I guess agood framework for I I would
just say the agent economy, likeagents to agents talking on your
behalf based on like what you'vedone.
And I was I was doing a littlebit of stuff on eBay where I was
like trying to get it to bid onstuff, and yeah, and then I
didn't put in some guardrailsand I bought a bunch of comic
(26:20):
books essentially.
My wife was like, Oh my gosh,uh, but but I mean, I wanted to
test it out, right?
Of like what's happening.
Um, and it was it was a reallygood experience uh to know
what's possible and like kind ofwhere things are going.
And I think there's a lot ofcatch up, but yeah, I I I mean,
I you're plugging in every APIand MPC to these things, like
(26:43):
they're gonna be connected in,like they're gonna they're gonna
know, like, okay, you put like Iwas I was I spoke at a
conference and it was likecameras, sensors, yeah, LLM
models are on all this stuff.
So, you know, okay, we'retalking um direct to consumer,
your refrigerator, the internetof things, knows like where your
milk's at, and then knows whatyou buy, and then it sends you
(27:06):
an alert hey, would you like meto order this milk?
You you know, it orders it, andthen people deliver it to your
house, right?
Like that that is gonna behappening very soon.
And and so I have been lookingat like, okay, where are we
going as an agency?
And um, I I bought another kindof governance AI, it's called
(27:28):
Modal Point.
Um, haven't been talking aboutit a lot, but it's it's really
like about data hygiene and likewhat is your data doing and
where is it going and likewhat's happening?
And the other piece that I'mkind of like looking at is
going, how are people gonna sellto these agents?
How are they gonna integratetheir business and services?
And I even interviewed a guyfrom Google right before COVID,
(27:50):
which was weird that it wasright before COVID.
But what he told me is he said,every business, and you know,
he's talking to like I wastalking to P B, he said every
business should have a shoppingcart on their website and be
able to do commerce throughtheir website as much as
possible because he saideverything's going that way, and
(28:13):
if you just think about how it'sall going to be interconnected,
it is 100%.
SPEAKER_02 (28:18):
And I the yeah, so I try and simplify this at least
so I have like language that Iunderstand in a as a career, as
a marketer, and so like I justthink of them as use cases and
channels still.
It's so easy to get like mireddown in the language and the
technology of AI, but like youget lost.
And so, to your point, likethere will be a bunch of use
(28:41):
cases of agenda commerce.
There will be like, is thisagent to agent?
Is this as simple as how do wemake e-commerce storefronts and
transactions work in a uh in alarge experience within a large
language model?
Is this it or is this an eightyou're a human's own agent
working and you're still buyingdirectly on e-commerce sites?
Like, I think we'll see allversions and permutations of,
(29:04):
but the nice thing is at leastas this matures, especially at
least from the large labs andthe LLMs that have you know the
majority of usage, protocols arestarting to be released, you
know.
Like Stripe is Stripe has donetheir integration with Chat GPT.
You're starting to get thefoundations and be able to see
at least a little bit into thefuture, and it allows you to
(29:26):
start building tools andstrategies and figure out what
to do.
SPEAKER_01 (29:29):
Well, I I want to I want to buy into Stripe and it's
like privately held, right?
Like I'm like Stripe's in themiddle of all of it, right?
And I I even think uh what a theuh what's his name?
Uh started square.
Um uh what is it?
Yeah, yeah.
Dorsey is I'm telling you, likethey they're right in the middle
(29:50):
of of of of and and I think ElonMusk knows what's going on too.
All the PayPal guys, right?
All the people guys were tryingto do this, right?
They were trying to do this, andthe technology just wasn't
caught up yet.
And to your point, thefragmented like growth is gonna
happen, like certain businessesand people are gonna get it, and
(30:12):
there's gonna be a huge tail onthis thing, but I mean it's all
eventually gonna end up there,and and and to your point, like
I just think that knowing thatthe the I mean, do you remember
when QR codes came out?
SPEAKER_02 (30:26):
Like, no one was using QR codes, they were
popular in Asia, but the NorthAmerican market for some reason
QR codes didn't work inmarketing, yeah.
SPEAKER_01 (30:36):
It didn't it didn't work, and then COVID happened,
and that use case came, andyou're like you can't touch
anything, and now everybody'sgot a QR code, everybody knows
how to use it.
Every menu, yes, yeah, yeah,everything, right?
And and so you know, maybethere's gonna be some kind of
catalyst like that, but I I seea gentic, whatever, you know,
(30:57):
people have different names too,right?
Like autonomous, and likethey're everybody's using their
own branded, like connectedwhatever, uh connected worker,
like and so but I I seecryptocurrency as kind of this
like frontier technology, uh,and payments.
Let's just call it payments,really.
Yeah, absolutely is is what Ithink it is, uh, more than
anything else.
(31:17):
Like, there's a couple coolthings, but you different use
cases, but but I think paymentsis the big thing, and then you
got agentic, and and then andthen you got well like decision
infrastructure essentially.
Like, how do you which is data,yeah, which is data, but big
data, data lakes, like how doyou hook it all together?
And all these things are justcoming together, and I I I feel
(31:40):
like there's there's just somuch in that space in any
direction, it just is going inall these different directions,
and it could turn into all thesedifferent use cases.
And and I think that that wherethose points meet is where uh
where everything's goingeventually.
Uh, and as those things kind ofcollide into each other and
(32:01):
people develop new things.
Like, here's the coolest thingfor anybody that hasn't done any
programming, you don't need toknow how to program, you just
need to know how to know whatyou built and kind of
understand, like maybe theblueprint piece of it.
But like, I got a 3D printerright here.
I I mean, I can create whateverin my head and then it can make
it like the like thepossibilities of where things
(32:27):
are going, and then like I mean,you I could just I mean, there's
technology that I could justdrop ship that to somebody super
easy, or you could probably have3D printers that would make
stuff, and then someone has abunch of them working and they
drop ship them for you.
Like, I mean, that's and thenyou talk about organs, they're
doing it with organs, like thewhere everything's going is just
(32:48):
um mind blowing, and marketingis right in the middle of it.
And marketers, I went to thisconference, Andrew.
That that is all agenciesowners, and what they were
telling everybody is you need tolearn to be the AI expert, and
clients are gonna outsource thatto you, and like and that's what
(33:08):
the data is showing, andeverybody was like, I did not go
to college for this.
SPEAKER_02 (33:14):
Like I think it was I think it was Mark Cuban the
other day on TV PN, but likethis is the new category, right?
Is is AI integration.
We had digital transformation,we we had to move into the cloud
at one point.
But like, how do you build newdigitally native businesses?
And those businesses that arenot digitally native, how do you
(33:34):
make that integration seamless?
Like, that is a massivecategory.
And I think that brands andbusinesses don't have enough
internal expertise yet.
Like that will change over time,and then there will, but right
now there's a big chasm.
And I at least for marketersspecifically, I think there's
one luxury is you don't need youdon't need to be all the way at
(33:55):
the bleeding edge because markthe marketing doesn't become
relevant until the consumerbehavior like substantiates it.
So we actually don't need it's aone, we should understand the
systems and talk about itbecause it's important, but like
we sort of get the fast followbecause we don't need to beat
our customers to a channel or toa use case, we just have to be
incredibly responsive andnimble.
SPEAKER_01 (34:16):
And and the clients have a lot of regulation and
internal stuff that they have todeal with, so they want to push
it to us right now because theycan't do it yet, right?
Like, literally, they're tryingto outsource it because they
can't bring it in-house yet,like they haven't all figured it
out.
So now is like this kind ofgolden sweet spot, whatever, for
(34:40):
anybody that's listening that'sin marketing to take advantage
of it and become that expert.
And I mean, there was peoplewhen when chat GPT really came
came on the uh scene, evenpeople in our organization that
didn't know how they felt aboutit and kind of was like, I'm not
sure about this, and people areconcerned.
And I'm like, this is not likejust a fact, this is a trend,
(35:04):
this is a movement, and we gottaget on the span wagon and in
front of this and become AIfirst, AI native as quickly as
we can, because that experiencecompounds.
Like, I see it, and even peopleon my team, I got power users, I
got people that are dabbling,and the power users are
compounding their quick, right?
(35:26):
Yeah, and then like if you're ifyou're getting into it, yeah,
you're gonna compound too, buthow much how much far ahead is
this person or this companygonna be ahead of you before you
jump on it, you know?
SPEAKER_02 (35:40):
100%.
I mean, I I think, and also theit's hard to build really great
marketing when you also havethrashing like at the
foundational level.
I mean, as as we have theability to advertise across AIs,
as we have the ability, ascommerce really like gets its
footing and becomes a real thingthat shoppers rely on, like
(36:01):
there's gonna be a bunch ofchanges.
So I think it actually it's veryhard.
You have to sort of thread theneedle by like this matters.
Brands, marketers, businessesneed to start like understanding
this is how they're chopper andhow businesses buy and search
and use the internet, but at thesame time, there is such a thing
as being too far out in front ofyour skis because things are
(36:21):
gonna change on a dime.
SPEAKER_01 (36:24):
So it's called parsnip, right?
SPEAKER_02 (36:27):
Is that like the vegetable?
Yeah, yeah.
SPEAKER_01 (36:29):
Like everybody that so that's interesting.
Okay, I want to talk about thatfor a second before we get into
it.
Is I feel like I'm seeing, andwe're we're actually launching a
campaign um the next month, nextmonth, we're launching a
campaign next month.
But I'm seeing so many AIcompanies find a a thing, like
(36:52):
what whatever it is, uh bear oflike like so.
In this space, there's bears,there's carrots, there's yeah,
they're they're like any kind oflike uh thing that everybody
knows what it is and canremember it, but then they like
put their spin on it so you youyou can uh subconsciously, I I
forget what it's called, butyou're you're creating a peg to
(37:13):
something that they know about,and so it's helping you with uh
memory recall, but it'ssomething different enough that
it's not AI is gonna know thedifference in whatever it is.
Does that make sense?
Uh 100.
SPEAKER_02 (37:28):
Naming is hard.
Like, I mean, I this it's it'slike one of those silly things
that like entrepreneurs likeit's actually a it's actually a
hurdle.
Um, and I think it'schallenging.
I think it would be to ourbenefit if people liked
parsnips, the vegetable more.
It's a pretty inconsequentialvegetable, so we have to make it
into a more consequentialsoftware business.
But uh, no, but like namings,naming is hard.
SPEAKER_01 (37:50):
So just to finish out kind of what I was saying is
um so we re we used to be callede web results.
We've been around 26 years, ouragency's been around a long
time.
We did web development.
I don't talk about it a lot onthe podcast.
Our audience set of who we workwith and who listens, completely
different.
EWR sponsors podcast, probablyshould do a commercial, but
(38:11):
essentially, um the brand therewas an e in front of everything,
right?
With the internet, right?
Yeah, so e web results, e webstyle, e-web results were really
in vogue, but you know, 10 yearsafter no one uses the e, it was
kind of like we need to updateit.
So we did the KFC thing, that'sall I could think of.
(38:31):
The KFC Kentucky Fried Chickenwent to KFC.
We shortened it to EWR digital,but nobody can remember EWR,
they can't remember it.
SPEAKER_02 (38:41):
Interesting.
SPEAKER_01 (38:43):
Okay, right, and so I'm trying to peg it to
something that people remember.
Yeah, so we're launching acampaign, we're doing a big
rebrand.
EWR uh is an airport tickersymbol.
Anything that's government, youhave no like marketing rights
to, there's huge volume.
Anybody that goes through EWR isusually going international, um,
(39:05):
right?
So our business users aremultinational companies.
We work with a lot of energycompanies, they know of EWR, and
people at business conferenceswill be like, EWR, like are you
out of New York or Newark orlike yeah, like whatever.
So I was just like, Okay, thisis the cloak.
I'm not gonna change the nameagain because a rebrand, and
(39:26):
I've done a number of podcastsabout this.
I would not encourage anybody todo a rebrand if you don't have
to.
It is tons of work, painful,like there's so much legacy
stuff out there.
You're talking about datahygiene, clean, it's bad.
And and and even these thememory sets on the search
engines and now AI, like it'sgonna be difficult.
SPEAKER_02 (39:45):
AI learn faster though, so that's to your
benefit.
As in, it takes a lot longer toclimb the you know the blue
links in traditional search,it's faster.
SPEAKER_01 (39:53):
Yeah, fair enough.
So, so we're gonna be doing thiswhole campaign on like kind of
take yourself to the skies, likelike as a hub, uh, for whatever.
Like, we're gonna do like thisbig play on it.
Yeah, and and then I get andthen my team, we were kind of
debating it.
And uh, sorry, I'm going on atangent, everybody, but we were
doing that.
We're it's interesting, I think.
Hopefully, you'll find itinteresting.
(40:13):
So basically, they're like, No,we're gonna get confused at the
airport.
And I was like, is that a badthing necessarily?
Because AI, I can tell you itlistens to us and it knows like
who we're around and they knowyour behaviors.
So if you're a businesstraveling person and you're
looking for a marketing company,it will probably know, hey, I
(40:37):
don't know which one to showthis to.
But if you're just like aregular user and you're not
looking for a marketing agency,it's gonna know to serve EWR, so
there's gonna be some confusion.
We're gonna get a little bit oflift, I think, temporarily from
EWR.
SPEAKER_03 (40:52):
Pun intended.
SPEAKER_01 (40:53):
Yeah, well, yeah, pun intended, I guess.
Yeah, I gotta uh there's a lotof fun things actually.
I'm thinking from a marketingstandpoint, we could do, but but
same thing, like I just I seethis consistently uh with new
companies as they're tagging itto something of a word or a
topic or an object that peopleare familiar with, and then you
(41:16):
add AI or whatever you're doingon it, and then and now you're
like, Oh, I remember that, andyou know, like it's totally
different than that thing.
So I I I think I'm I'm that wasa long-winded way to
congratulate you on uh thenaming of it is difficult, and
you did something that peoplealready have in their brain.
SPEAKER_02 (41:35):
That's the help.
And we have we have an emoji wecan use because there's a
carrot, so it we can shorthandit to the emoji keyboard if we
need to.
SPEAKER_01 (41:43):
Yeah, so tell tell me a little bit about kind of
how you started and then umwhere you see the opportunity
now because as as we starttalking about at the beginning
of this, AI um visibility maynot be a category, it's just
kind of a standardinfrastructure as we move
forward.
SPEAKER_02 (42:00):
Yeah, I absolutely think that there will a peer
play AI visibility platform willnot be the type of tools we're
talking about.
And I don't know the timehorizon is this six months, this
12 months.
I think you have to understandAI visibility.
I think you need to optimize forit, but that dovetails into the
use cases as a marketer that youactually need to perform, right?
(42:21):
Yeah, like a big piece of thatis content, right?
One of the one of the core waysto go combat traditional SEO and
improve GEO visibility isthrough developing great
content.
Like Parsnip is not a contentcompany, right?
There are some really great AIvisibility tools that in my
mind, they're these are contentcompanies, right?
There will be when I think ofParsnip, we have an AI
(42:45):
visibility tool that doesexactly what brands need.
You figure out how you indextoday, you go through, you
figure out how you rank againstcompetitors, and then you spit
out recommendations on how toget started.
That's cool.
I think the core competency,though, is for us is agenda
commerce.
So figuring out if we have thisreally long customer journey
(43:05):
where you have to build content,you have to update parts of like
you know, website schema andproduct feeds, and you're gonna
be able to advertise across AI.
You will have software toolsthat help, I think, and have
competencies in each of those.
For us, it's figuring out if youunderstand your visibility, how
do you move, how do youintegrate that with the ability
(43:26):
to sell on the surface?
So that and we're e-commercefolks, you know, historically.
That's that's our background.
We built kind of software at theintersection of social media
marketing and e-commerce for thelast decade.
So in many ways, it's the same,you know, just the usage that
we're helping marketers respondto is not people on social
media, it's people using largelanguage models.
SPEAKER_01 (43:48):
Yeah.
And I what you started talkingabout where I keyed in on it
specifically, and I haven't donea podcast since the new Google
Core update release, whichContent AI generated content, a
bunch of uh directory sites gottaken down.
Um there are things that youneed to do when you're
(44:11):
generating content and and toolslike Parsnet um help you uh make
sure you're checking all theboxes, like make sure you're if
you are using AI to helpgenerate the content, um there
there are fundamentals andthings that you need to be doing
to make sure that you don't gettaken down and usually attaching
(44:33):
it to a human, there's kind ofeats that you need to focus on,
expertise, authority, trust,experience.
Um, I mean, Google came out andI thought they waved the light
flag and said, hey, if it'suseful content, it's fine.
I feel like AI generatedgenerated content was pretty
targeted uh in this last update.
Um, but there's things that youcan do to mitigate it, there's
(44:55):
things that you need to bethinking about, and it's
constantly evolving uh to yourto your point, Andrew.
And and the consulting andstrategy piece, I think is big
of what people are looking forand in tools providing that that
that knowledge to say, what do Ineed to be focused on?
What do I need to be learningat?
Because I'm not going to be theexpert in this.
(45:16):
I'm trying to do my job and allthis other stuff.
And so I I think these kind oftools are really, really helpful
to give you almost like achecklist or or synthesize it
down or abstract it out so thatyou kind of have a direction of
like these are the buoys that Ineed to start to hit or go in
that direction.
SPEAKER_02 (45:35):
But it's because most folks, like we like we nerd
out and talk about these thingsall day, but most folks go,
okay, I understand this, youknow, I understand people are
using chat GPT every day andhaving conversations I care
about for my business, but theydon't know where to start.
So when I think of when I thinkof what Parsnip looks like
today, it makes it as easy aspossible to understand where you
(45:56):
stand and how you get started.
What are the 30 things youshould start doing?
I don't think that's the fullpicture.
And I all and so I think we'regonna see the evolution really
fast.
Like we made the decision not togo deep down the content route.
I think there will be greatcontent companies that support
GEO and AI visibility.
And I'm also a little, these areput these are my words, you
(46:18):
know, this wasn't part of theGoogle announcement, but like
why AI content was targeted.
I don't think it was targetedbecause AI can't help build
great useful content.
I think the problem was some ofthe early GEO strategies are
kind of like when we used to putclear text on websites for
keywords in old school search.
You know, you'd have yourwebsite and then you'd have your
(46:39):
keyword 30 times in clear font,you know, so that you know
Google pick.
I think that we there's a therewas a little bit of we're gonna
have use AI to generatederivative content, answer the
same thing 50 different waysthat was not useful.
So, like yeah, yeah, it's spamfilter.
SPEAKER_01 (46:57):
Yeah, yeah, it's just a spam filter.
It's like it's not genuinelyuseful content.
Yeah, yeah.
Hey, so uh I uh I was talking tosomebody and I I use the term AI
slop and they got a theyactually got offended.
It was like a big like, and thenlike I'm like, this is the word
of the year for like uh whatwhat what what uh what um
(47:20):
dictionary was it?
Was it Webster or something?
That AI slop was the word of theyear.
Oh, where they do word of theyear, yes.
Yeah, it was AI slop.
I was like, I was like, sorry,like I didn't think that this
was offensive, but um, you gotto be careful AI slop.
SPEAKER_02 (47:36):
So um it's it's about usefulness and to your
point, like e like AI.
So by the way, there's great AIcontent out there, there's also
really horrible AI content outthere, and so I think using it's
like using the right tool tosolve the right problem, like AI
will be the answer in a lot ofthose contexts, but it doesn't
mean that you flood the system.
(47:57):
It's also not the way theinternet can work.
If agents are havingconversations with themselves
proactively on behalf of theshopper to infinity, like nobody
wins ultimately.
Like that's a that's a badformat.
SPEAKER_01 (48:10):
So, so Andrew, we gotta we're getting close to
time here.
I gotta wrap up.
Um, I would love to hear likewhat is the biggest takeaway
that with companies you'veworked with consistently on the
e-commerce side that they'vejust failed to do, or there's a
gap, a blind spot in that theycould help improve uh their
(48:31):
their whatever they need to do,ranking, visibility, access to
the right customers, kind of inthat vein.
And then share a little bit moreabout uh where you're posting
online, like kind of where yourconversations are happening, so
people can follow you and uhplug whatever you would like to
plug.
SPEAKER_02 (48:47):
Absolutely.
Well, thank you.
Thank you for having me as well.
Been a fun conversation.
Um, I think the biggest, thebiggest issue most folks have is
it's kind of like analysis, it'sparalysis through analysis, is
everyone is having these typesof conversations, like the one
we had today, and assume theyhave to like hop over the chasm.
Specifically with commerce,there are baby steps, and there
(49:09):
are things that if you market orsell online, things you
absolutely need to be doing thatare not overly challenging.
It's just more likely than not,this isn't a marketing
discipline that your team is upto speed on.
So it there are ways to takebite-sized chunks, is the
biggest feedback.
There are ways to make sure youhave a correct tracking pixel
for Chat GPT, or integrate aproduct feed, or make sure that
(49:33):
your product catalog isorganized with a schema that
LLMs can read.
Like very simple stuff thatevery e-commerce professional or
digital marketer, theyunderstand the tactics, but I
think there's a disconnectbetween all the things they're
really good at and they've donetheir whole careers and AI.
And a lot there's not enoughconversation that's connecting
(49:54):
them.
So they go, well, AI is this bigquagmire, we need to figure it
out.
And then they understand taggingand product feeds and tracking
and content marketing.
And so figure out the way totake a few small steps and start
working in the direction.
SPEAKER_01 (50:08):
Love it, love it.
SPEAKER_02 (50:09):
Um then obviously we, you know, part snip is we
are we launched the business inNovember, uh, but we do a lot on
LinkedIn uh and our website arethe easiest places, but we've
tried to make it's a it's a freetool to try.
We've tried to kind of put ourmoney where our mouth is in
terms of make it really easy toget started and just start to
figure out what the roadmapshould look like.
SPEAKER_01 (50:30):
Very cool.
Well, Andrew, thank you so muchfor coming on.
I think this is a greatdiscussion.
Um, I think that you know, youget a bunch of PhDs in the room
and they're talking about abunch of AI stuff uh on both
sides, and nothing gets done.
And you gotta you gotta havesomebody that comes in and
translates it.
And hopefully we we did that foryou today, uh, for people
(50:51):
listening, like understandingwhere things are going, maybe
some action steps and kind ofgiving you that general
direction to take the next stepand to move forward.
We do do coaching and classes ifyou want to reach out, uh find
out more information about that.
Uh, bestseopodcast.com.
Um until the next time, guys.
(51:13):
My name is Matt Bertram.
Uh, if you want to grow yourbusiness with the largest, most
powerful tool on the planet,which maybe is AI, not the
internet anymore, uh, reach outto EWR for more revenue in your
business.
Thank you so much, Andrew, forcoming on.
Thank you, everybody, forlistening.
Talk to you later.
Bye bye for now.
This is the unknown secrets of internet marketing.
Your insider guide to thestrategies top marketers use to
crush the competition.
Ready to unlock your businessfull potential.
Let's get started.
SPEAKER_01 (00:18):
Howdy, welcome back to another fun-filled episode of
The Unknown Secrets of InternetMarketing.
I'm your host, Matt Bertram.
We are talking about everythingAI marketing, AI visibility,
SEO, um, kind of how you dealwith the gentic.
Everything's very fluid.
Uh, wanted to bring in somebodyI know people have been asking
about e-commerce and agenticstuff.
(00:38):
That is not my wheelhouse.
So I want to be clear.
Um, I do know and have greatpartners uh that do a lot of uh
e-commerce and agentic stuff.
We're doing more B2B stuff, leadgen, that sort of thing.
Uh, and and so uh all right,Andrew Higgins.
Andrew Higgins, welcome to theshow.
He's got an AI visibility tool,he's doing e-commerce agentic.
(01:01):
Uh, we're having some differentkinds of conversations around
like, is LLM visibility likebaseline analytics, right?
Like, I think that that's uh areally interesting point that
everybody's trying to add AIvisibility, whether it be GA4,
uh, all the different uh goldstandard tools are offering um
(01:24):
LLM visibility, everybody'sdoing it a little bit
differently.
Um, but everybody's racing to doit.
And to your point, I feel likethere needs to be a baseline.
It would be great if there waslike a standard.
I don't know if Google's gonnacome out and say like it is.
I mean, they're switchingeverybody over to AI search uh
shortly.
So um I know it's gonna be moreof a conversation.
(01:47):
Um, but I but I would love tokind of get your landscape,
Andrew, on what what you'reseeing in the marketplace.
I know you launched a tool um uha while back around L
invisibility.
I know they're popping upeverywhere.
So I'd love to see what you'reseeing with that, with that, as
well as like kind of what isgoing on with e-commerce.
SPEAKER_02 (02:08):
Totally.
Well, thank you for having me.
Um I think everyone is isscrambling to try and figure out
the problem at the same time.
And to some extent, likemarketers are trying to build
the plane while they're flyingit, like pardon the old adage,
but I think the the reality iswhat we know, like if we just
figure out things we know,consumer behavior and the way
(02:29):
that people are digesting andusing the internet has wildly
changed on the back of AI.
The way we figure out whatrestaurant to go to or vacation
or what to buy our moms fortheir birthdays, like that
behavior, that customer journeyhas changed a ton in the past
few years and is changingrapidly on an almost weekly
basis.
(02:49):
So I think when we think aboutwhether it's AI visibility or
e-commerce or honestly any partof the digital marketing stack,
it all is being retooled and theplaybook is being rewritten.
So I think the the firstconversation that at an industry
level we're having is GEO or AIvisibility or AIO or AI SEO.
(03:12):
Like we don't even have thenomenclature right yet.
Uh I'm not convinced that's themarket.
I think that's where theconversation today is.
Uh I think that's for a couplereasons.
One is because we understand SEOand we had this monolith that
was Google search and governed alot of eyeballs and the
beginning of a lot of shopperjourneys.
(03:33):
Uh today, one, we have it's nota monolith.
There are lots of players, thereare lots of labs in large
language models, and even underthe hood of you know, Gemini as
an example, there's multiplemodels that behave differently,
that collect data differently.
And the biggest piece is it's nolonger just 10 links.
(03:54):
No context matters.
So if you and I have a similarinteraction with a large
language model, we will getdifferent results because of the
context of who we are.
And all of that changes the waywe start to optimize.
If we know people areinteracting with AI instead of a
brand or a retailer or socialmedia, we have to marketers, we
(04:16):
need to figure out how toinfluence those interactions,
but all of that is changing.
SPEAKER_01 (04:20):
Well, I mean, you you see it with like Rand
Fishkin and like what he'sdoing, right?
Like he like he made the pivotto like ICP.
So it's just like who who is theaudience?
Where are they at?
What is happening um before theyeven start searching, right?
Like what like it goes back,like pendulum back to you know,
(04:43):
mad men, right?
Like, like where what what isthe creative?
What is the brand?
What is the brand image?
Like, where is that brand showup?
Where does it want to be around?
Like, I I kind of even feel likeyou're like running the track.
Like you got to do reps, you gotto do reps, and you got to be in
those same areas consistently tobe viewed as that authority,
right?
And so it it's a full funneloptimization of where does your
(05:07):
brand show up?
Where is your audience at?
How are you speaking to them?
What questions are youanswering?
It all comes together with areyou communicating with your
target audience correctly andwherever they're at, uh, at
whatever their need is.
And and that's a much bigger askthan optimizing for top 10 links
(05:28):
in Google.
SPEAKER_02 (05:29):
Well, and I think the all the principles that you
just walked through are thesame.
Like AI doesn't change the coreprinciples of marketing.
We have new channels, we havenew use cases, we have new ways
we need to try and startmeasuring data.
Like, so principles are theexact same.
We just understand consumerbehavior is different, and it's
(05:49):
that's the reason I think AIvisibility is a big conversation
today.
I'm not sure it will betomorrow.
And reason being is step one islike understand the usage, under
start to try and figure out thechannel, understand how AI talks
about my brand, how it'sdifferent to my competitors.
And then the biggest thing is ifyou go, if you go all the way
(06:10):
back, you mentioned the Mad Menera.
Well, there used to be 10 TVchannels and some a few radio
stations and a few largepublications.
People read the newspaper.
So the ability as technology hascontinued to advance, we can
target our user.
And I think AI is another leapin the direction of your ICP is
becoming less and less broadbecause your ability to
(06:33):
personalize is stronger andstronger.
So today, when contextinfluences, you know, if you
know, how an LLM interacts witha shopper, then we as marketers
need to be able to get down tothat level of specificity as
well.
So I don't I think coreprinciples are still the same.
I think the playbook is beingrewritten, the tools, the stack
looks different.
But the hot topic today is thedata, which is AI visibility,
(06:58):
which I don't think will be along-term category.
I think it of course willmatter, but also the data is not
accurate yet.
Like I've that's the data isthis is benchmark data.
And if you use 20 differenttools doing something related,
you'll actually get 20 differentanswers, which is a problem.
SPEAKER_01 (07:15):
Well, you I mean, that even goes into like why the
models are hallucinating orwhatever.
It's like bad data, right?
And like even when I'm likedoing some programming stuff,
I'm going, like, why did yougrab that?
And then I'm like, go down alittle bit of rabbit hole to
like understand why it broughtthat in.
And like people aresensationalizing, like, oh, this
AI got um what was it?
(07:36):
Like I got got really interestedin like the Golden Gate Bridge,
or there was that like stuff,and it was like, well, if you
keep feeding the model, there'slike a decay rate and attention,
and like, of course, that's whatthe models do if you understand
the models.
But if you're looking ateverything black box, then you
know that that data hygiene andand understanding like the math
(07:59):
that goes into this is reallyquite important.
And I I believe I believe whatyou're saying because the
visibility pieces, like, we'rejust trying to figure out how to
measure, and then once wemeasure, it goes back into like,
well, is your marketing good?
Because that's the debate rightnow with everybody on social
(08:21):
media that I'm seeing is likegeo is completely different, and
then no, it's the exact same,it's just good marketing, right?
And it's it's this back andforth.
Then it's like, okay, it is goodmarketing, but how how to do it
and understanding that this is acompletely different technology
than Google, like Google worksdifferently on how it gets to
(08:43):
the answer, it might get to thesame answer, but how it gets to
it and understanding the theunder the hood is how you get
the competitive advantage toexecute, but good marketing is
always gonna win.
And and really, I feel likemarket feedback and and I think
that that's leads intoe-commerce is understanding the
(09:05):
user, the the the the customerhas changed is how how has
informing you on what you needto do, on how you need to speak
to them.
And if you have e-commerceversus like a B2B where it's
offline, online, like whatever,and you can't track everything,
you get cleaner data.
(09:25):
So data hygiene is like likefirst, right?
Are you measuring everythingright?
And e-commerce, you get to seekind of the full experience, so
you can optimize what'shappening and you get more data
in that process because it's youknow, you're you're targeting
the user, you're engaging withthe user, they're buying online,
(09:46):
and then you're drop shippingthem something or whatever,
right?
SPEAKER_02 (09:49):
Well, and I you're I agree completely.
I think we we don't have afeedback loop.
So it's if this is a newdiscipline of marketing, we've
got to create the feedback loopso you can truly understand what
works and what doesn't.
Um, but the interesting piece iswe don't have usage data yet.
So if we yes, we have tounderstand the technology and
understand how large languagemodels work so that we can start
(10:12):
to reverse engineer them tofigure out how as a brand we
optimize ourselves to bevisible, to be present, to be
discoverable.
But the state of the market,like the marketing discipline of
you know, AI search, GEO,whatever we're gonna call it, is
still quite immature in thatwhat we're actually doing under
the hood is it's predictive.
(10:33):
You're building a model of ashopper and having simulated
interactions.
That's kind of like buying aFacebook ad and meta saying, I'm
not gonna tell you the number ofimpressions, I'm gonna guess,
I'm not gonna tell you yourclick-through rate, and I'm not
gonna pass through a tag, butbuy my ads anyway.
And we wouldn't do that.
I mean, there was an era youknow, of TikTok where it was the
(10:54):
usage was so high that peoplewere willing to invest
advertising dollars in anexploratory way, but of course
you started getting B2B APIs soyou could measure these things,
and so you could close thefeedback loop.
We're still in the infancy wherewe don't have the feedback loop.
And that's I think what a lot ofmarketers are trying to
establish as a step one.
SPEAKER_01 (11:13):
Okay, so so one of the interesting conversations,
which this might be out outsideof this discussion, but I've
I've started to see it is I knowFacebook got in trouble.
I mean, they're getting introuble for a lot of stuff right
now, but um, they got in troublefor really faking impressions
like early on, like like they'reand and you wonder why like
(11:34):
these ads are not working theway they should.
Like, my I always went back tolike, are they showing them to
the right people?
Right, because they need to getthere, they're trying to hit
spin numbers, right?
So they keep expanding it out,expand it out.
Google does the same thing.
I saw a study recently thatcompared um running ads on the
Google platform versus usinglike a DSP or something like
(11:58):
that to buy ads for you of thetarget audience set.
Interesting, because Google isshowing.
I mean, you're paying like apercentage to run through the
DSP, but if you took the samedollar amounts, right, and go,
well, I'm getting 12% more uhjust doing Google and I'm not
using the DSP.
(12:19):
What's gonna be the result?
Yeah, the results from thisstudy again, like I don't have
all the details and don't quoteme on it, but it looked to be
really, really accurate thatGoogle was serving ads to
properties that benefited themmore than benefited the user,
and they had a limitedinventory, so like they're
(12:42):
forcing things that are are notthe best use of your money.
And if you if if if you havemore inventory to select from
and you kind of lock in whoyou're trying to reach a little
bit tighter, you'll always getyou're gonna get better
targeting, right?
Like it was it was kind of likeyou knew the the setup, but it's
(13:04):
so true, right?
And I feel like I mean now toyour point, measuring LLMs, like
different use cases.
Like when I see an LM and I'mlike, ah, we're in Claude and
like, oh my gosh, like peopleare stealing all our stuff, like
you know, like that's what Ithink, like, but like like you
you all of these differentchannels, right, or AIs, people
(13:26):
are gonna have differentrelationships, and I think the
battle is like which one's gonnaown, like, which one's gonna be
the one that is the one thateveryone uses.
SPEAKER_02 (13:36):
I don't care it might not matter though, it
might it that who knows?
I think we we there will be alot of evolution that might be
like saying you have to pick onesocial media channel to market
across, right?
That these might be use casespecific.
SPEAKER_01 (13:50):
Well, yeah, but uh so this is what I would tell you
these AIs store okay, chat GBT,for example, they came out and
said they're gonna remember,like I think like it's like
government surveillance,whatever, but like they're gonna
every single thing, just ifyou're listening, look in the
terms of use, every single thingyou put into chat GBT, they're
(14:11):
keeping forever.
I don't know if everybody knowsthat.
So if you're listening, thatthat they came out and said
that, not so like FISA warrant,whatever.
Like so, just know that ifyou're using Chat GBT or using
codex, every single thing you dois going in that now.
Then you got people, Andrew,that are uh taking open source
(14:32):
LMs, putting them on their ownservers, air gapping it, like
all their data is not leaving,it's not getting processed
because anything you put in anyof these tools is getting
processed off site.
So I think anthropic, I uh youknow, it was 30 days at one
point.
I don't know what they changedit to, but they they they clean
all the data every 30 daysunless like you store it.
But but perpetual memory, likememory is a big conversation,
(14:57):
uh, I feel like around thesedifferent AIs and how it recalls
information and like the ragpulls and all that.
So what people are finding thatare using these tools that are
consumers, not builders uh inthese tools, is they're getting
really comfortable with one AIbecause it knows them and it
learns them, right?
And they're comfortable with theinterface and they know how to
(15:19):
use it.
I I mean, I feel like I knowpeople that are jumping around
between different tools, yeah.
But the people that are not inthe marketing space, like like I
would consider just straightconsumers, yeah, they've started
to use one tool, they haven'tdone it, and right, and they get
comfortable with that, and andlike even there's like little
(15:41):
debates on like, well, I likeClaude better and I like ChatGPT
better, and then they're liketalking about how they like it
better, but no one's tried theother one.
And I feel like there'singrainment in that interface,
and the more things you pluginto it, right, and it gets
better data, uh the theswitching costs become harder,
right?
And right, and so I I meanagain, like we don't know what
(16:04):
the future holds, like whoknows?
Like, I mean, we don't even knowif there are aliens yet or not,
right?
But like AI, like I I would tellyou that I I think the battle
honestly is who builds the bestconsumer tool that everybody has
an AI that they filtereverything through because it
(16:26):
knows them so well and it helpsthem make decisions.
I mean, you're even seeing kids,you know, and like don't talk to
it like a human, and like Imean, it's causing like all
kinds of things, bad things tohappen.
Um, I I think that there's gonnabe a trusted agent, like or
whatever that you're workingwith that that is your primary
(16:48):
like assistant that you filtereverything through, which of
course you could build someadvanced things that plug in and
grab the different models andall that.
Um, and there's a tool outthere, guys.
I think it's called Poe.
I I don't really use it, but itit grabs all the different LLMs
and you can ask it a question ifyou're using it just as a chat
bot, not like as a developmenttool.
(17:08):
But um, like I just think peopleare gonna have their go-to and
they're gonna filter everythingthrough it.
So if that I I can tell youright now, I was looking at um
to buy something on Amazon orwhatever, and I was like,
screenshot it, put it in there,look at the reviews, like do
some additional research.
(17:29):
Yeah, tell me this and tell meif there's something else that I
um that I'm missing, and then itwas like, Oh, well, uh, you
know, you're you're wanting todo this.
Uh like it was like a supplementor something, right?
It was like, um, you're wantingto do a gummy, like that might
not be as whatever, and and Iwas trying to decide between
these like mushroom gummies orwhatever, right?
(17:51):
And then it basically was like,No, you need a capsule, and then
these are the brands, and thenthese are the like like I was
down this path, and like thesewere like through that like
consideration phase, yeah.
Like I was there, and then itjust pivoted me to go like, no,
you didn't think about thesethings, you need to figure out,
and then these are the threebrands that you should select
(18:12):
from, and here are the linkslike that.
Like, because you're okay, goingback to L invisibility, like if
like how did I get prompted withthat question?
And then it represented this,and someone answered it, and
then these three brands answerthat question the most
(18:33):
effectively and had the mostreviews, and like whatever else,
that it got shown to me,provided me links, and then I
bought something over here thatI wasn't even thinking about,
right?
Like, that's the world thatwe're trying to measure.
SPEAKER_02 (18:49):
Well, and think about how much of that journey
sat on one surface to your pointof like it sat, whereas the
fragmented like digital customerjourney, yes, you still
transacted on Amazon or whereveryou ended up purchasing, but
pretty soon that transactionwill also be able to take place
with uh with the AI buying onbehalf of the shopper.
(19:09):
So I think about you know, thecon I think about how long the
customer journey is is gonna beon one single surface.
So is this visibility is not theright question.
This is, and that there will beso many use cases built into one
channel.
You know, this is yes,visibility will always matter,
(19:30):
and you yeah, but you will haveto figure out how to stay
visible, you will have to figureout how to, as you get into
consideration and you get intocompetitive comparison and
return policies, you'll have tomake transacting easy.
You know, when you think aboutlike currency, you know,
currency across, you know, abunch of different models in
different markets, you know,where you have context on the
(19:51):
shopper and pricing promotionsin different regions, like this
stuff gets very complicated.
And we know the drop-offs we seeand conversion and engagement
when you do those things wrong.
SPEAKER_01 (20:02):
Yeah, well, I mean, yeah, I mean, relevancy is going
to be a big component, butyou're right, visibility
honestly doesn't matter.
It's like at what stage of thebuying process are you at?
Correct.
And now here's the thing to havewho like the work is what was
done to make that recall happenthe way it happened, right?
(20:24):
Like, so that's where all thelike invisible work comes in.
I um what was I gonna say?
Uh you were you were talkingabout something, and I was like,
man, I need to write this down,but um keep going, keep going,
and I'll think of it.
SPEAKER_02 (20:40):
But but that that is where I think the challenge
becomes exponentially harderbecause we're not optimizing for
one click and then owning thatexperience on our site, it's
very multifaceted.
SPEAKER_01 (20:49):
Oh, this is what I was gonna say.
So, Google, if you watch like umtheir, you know, every every
every big brand, uh, and I wouldencourage anybody that's a
marketer.
Uh, if you can't go to theseevents, um, and and I I I don't
go to all of them, but like youcan go watch them on YouTube or
whatever.
But Google's last like yearlyunveil, it's kind of like the
(21:10):
talk where Steve Jobs wouldalways come out and unveil like
the new product or whateverthey're doing.
So Google's big event, I I uhit's escaping me what it's what
it's called right now, but theywere unveiling AI search and
they were unveiling what we'retalking about, is and I think
that the reason they're doing itnow is they don't want to lose
(21:31):
uh customers to another surface,like or another, and so they're
switching to AI mode, right?
And that's why they're doing itbecause they already know that
the whole uh value funnel or orthe search funnel is gonna
collapse, and they want to keepit on Google.
And so if everybody uses uh youknow uh Gemini or whatever's
(21:55):
powering AI overviews, and thatyou can search on Google.
Google, ask all the questionsand buy, and they have this, you
know, huge market share, they'reprobably gonna win this next
evolution, right?
Correct.
Because there's not I thinkyou're muted by the way, Andrew.
I don't know.
Um, but uh there you go.
(22:16):
Okay.
And uh and uh I I I mean I Ithink they're making the switch,
they were prepping it, butbefore too many people are doing
the buying journey and some ofthese other platforms, they they
still have the market share.
People are still saying they'reusing AI because they're looking
at the overviews, right?
And like everybody feels good.
I think I think they're lookingat the metrics and and they're
(22:41):
gonna hard switch it because ifthey can capture everybody just
to do it on Google, um, they'llthey'll win the next uh round,
right?
And then all that customerjourney will happen there.
So I still think optimizing forGoogle is uh what needs to
happen.
SPEAKER_02 (22:58):
Um I also yeah, and I don't think they're
conflicting like tactics, whichis the I think Optima, even the
SEO versus geo discussion, likethey are they belong in the same
discussion, and there are thingsthat are uniquely positioned for
optimizing for AI that aredifferent, do not matter with
the way Google indexestraditional search, but there's
(23:20):
also a huge overlap, right?
And so I don't think they'remutually exclusive.
I think the the reality is thatlike the core hygiene uh that is
SEO will continue to be true.
There's now just a bunch ofother elements of building a
website, marketing online,working with PR and you know
that that matter and areingested in a different way,
(23:41):
whether through RAG or throughyou know, the way large language
models interpret information.
And also the way we search, justwith when it was when you had
one search and you had to winthat initial keyword or that
keyword phrase, and you youoptimize for fewer pieces.
Now it's you have to optimize,you know, along that whole
journey, and the way the way youdo that changes based on where
(24:02):
you're at in the funnel.
SPEAKER_01 (24:04):
Yeah, no, I I totally agree.
So, what are some other insightsthat you're seeing on the
e-commerce side that would behelpful to bring into this
conversation?
SPEAKER_02 (24:13):
Um, that it's coming first and foremost.
I think agent agentic commercealso even by itself is a broad
conversation.
SPEAKER_01 (24:20):
Like are you talking agent to agents?
SPEAKER_02 (24:22):
That's exactly what I think.
Okay.
So most of what I deal with on aday-to-day basis when I think
agente commerce, it is how ise-commerce going to be
integrated into large languagemodels?
Like, which is I think only asliver of the discussion because
the way in which agents aregonna work, you know, work on
behalf of shoppers is much moreplentiful than that.
SPEAKER_01 (24:43):
Okay, so for a while, anybody that listens to
me, I kind of went down this umuh cryptocurrency path.
Uh, and and we have some bigclients in the space uh that
we're doing we're doing geo workfor and all that.
But um I really see so there wasa protocol back in like 91-92
(25:05):
that the internet do you knowwhat I'm talking about?
Like the what it what sobasically there was a protocol
on the internet where you couldtransfer money uh online, but it
was never used.
I think this year they justturned it on, um, and they're
getting ready to move to digitalcurrency uh in the banks, and
(25:26):
there's some laws being passedand everything like that.
Because, well, I think that theagents are gonna want to use
like a Solana or like somethingfast to be able to transact and
uh fractionally buy.
And you're seeing it with APItoken usage as well, but I I
think it's even broader, right?
(25:47):
Like agent to agent commerce,which what I've read, it's gonna
be like 75% like eclipsing thecurrent e-commerce.
SPEAKER_02 (25:55):
The TAM projections are astronomical, yeah.
SPEAKER_01 (25:59):
Right.
And so, like, it's not even it'slike the agentic e-commerce.
Okay, that's like I I guess agood framework for I I would
just say the agent economy, likeagents to agents talking on your
behalf based on like what you'vedone.
And I was I was doing a littlebit of stuff on eBay where I was
like trying to get it to bid onstuff, and yeah, and then I
didn't put in some guardrailsand I bought a bunch of comic
(26:20):
books essentially.
My wife was like, Oh my gosh,uh, but but I mean, I wanted to
test it out, right?
Of like what's happening.
Um, and it was it was a reallygood experience uh to know
what's possible and like kind ofwhere things are going.
And I think there's a lot ofcatch up, but yeah, I I I mean,
I you're plugging in every APIand MPC to these things, like
(26:43):
they're gonna be connected in,like they're gonna they're gonna
know, like, okay, you put like Iwas I was I spoke at a
conference and it was likecameras, sensors, yeah, LLM
models are on all this stuff.
So, you know, okay, we'retalking um direct to consumer,
your refrigerator, the internetof things, knows like where your
milk's at, and then knows whatyou buy, and then it sends you
(27:06):
an alert hey, would you like meto order this milk?
You you know, it orders it, andthen people deliver it to your
house, right?
Like that that is gonna behappening very soon.
And and so I have been lookingat like, okay, where are we
going as an agency?
And um, I I bought another kindof governance AI, it's called
(27:28):
Modal Point.
Um, haven't been talking aboutit a lot, but it's it's really
like about data hygiene and likewhat is your data doing and
where is it going and likewhat's happening?
And the other piece that I'mkind of like looking at is
going, how are people gonna sellto these agents?
How are they gonna integratetheir business and services?
And I even interviewed a guyfrom Google right before COVID,
(27:50):
which was weird that it wasright before COVID.
But what he told me is he said,every business, and you know,
he's talking to like I wastalking to P B, he said every
business should have a shoppingcart on their website and be
able to do commerce throughtheir website as much as
possible because he saideverything's going that way, and
(28:13):
if you just think about how it'sall going to be interconnected,
it is 100%.
SPEAKER_02 (28:18):
And I the yeah, so I try and simplify this at least
so I have like language that Iunderstand in a as a career, as
a marketer, and so like I justthink of them as use cases and
channels still.
It's so easy to get like mireddown in the language and the
technology of AI, but like youget lost.
And so, to your point, likethere will be a bunch of use
(28:41):
cases of agenda commerce.
There will be like, is thisagent to agent?
Is this as simple as how do wemake e-commerce storefronts and
transactions work in a uh in alarge experience within a large
language model?
Is this it or is this an eightyou're a human's own agent
working and you're still buyingdirectly on e-commerce sites?
Like, I think we'll see allversions and permutations of,
(29:04):
but the nice thing is at leastas this matures, especially at
least from the large labs andthe LLMs that have you know the
majority of usage, protocols arestarting to be released, you
know.
Like Stripe is Stripe has donetheir integration with Chat GPT.
You're starting to get thefoundations and be able to see
at least a little bit into thefuture, and it allows you to
(29:26):
start building tools andstrategies and figure out what
to do.
SPEAKER_01 (29:29):
Well, I I want to I want to buy into Stripe and it's
like privately held, right?
Like I'm like Stripe's in themiddle of all of it, right?
And I I even think uh what a theuh what's his name?
Uh started square.
Um uh what is it?
Yeah, yeah.
Dorsey is I'm telling you, likethey they're right in the middle
(29:50):
of of of of and and I think ElonMusk knows what's going on too.
All the PayPal guys, right?
All the people guys were tryingto do this, right?
They were trying to do this, andthe technology just wasn't
caught up yet.
And to your point, thefragmented like growth is gonna
happen, like certain businessesand people are gonna get it, and
(30:12):
there's gonna be a huge tail onthis thing, but I mean it's all
eventually gonna end up there,and and and to your point, like
I just think that knowing thatthe the I mean, do you remember
when QR codes came out?
SPEAKER_02 (30:26):
Like, no one was using QR codes, they were
popular in Asia, but the NorthAmerican market for some reason
QR codes didn't work inmarketing, yeah.
SPEAKER_01 (30:36):
It didn't it didn't work, and then COVID happened,
and that use case came, andyou're like you can't touch
anything, and now everybody'sgot a QR code, everybody knows
how to use it.
Every menu, yes, yeah, yeah,everything, right?
And and so you know, maybethere's gonna be some kind of
catalyst like that, but I I seea gentic, whatever, you know,
(30:57):
people have different names too,right?
Like autonomous, and likethey're everybody's using their
own branded, like connectedwhatever, uh connected worker,
like and so but I I seecryptocurrency as kind of this
like frontier technology, uh,and payments.
Let's just call it payments,really.
Yeah, absolutely is is what Ithink it is, uh, more than
anything else.
(31:17):
Like, there's a couple coolthings, but you different use
cases, but but I think paymentsis the big thing, and then you
got agentic, and and then andthen you got well like decision
infrastructure essentially.
Like, how do you which is data,yeah, which is data, but big
data, data lakes, like how doyou hook it all together?
And all these things are justcoming together, and I I I feel
(31:40):
like there's there's just somuch in that space in any
direction, it just is going inall these different directions,
and it could turn into all thesedifferent use cases.
And and I think that that wherethose points meet is where uh
where everything's goingeventually.
Uh, and as those things kind ofcollide into each other and
(32:01):
people develop new things.
Like, here's the coolest thingfor anybody that hasn't done any
programming, you don't need toknow how to program, you just
need to know how to know whatyou built and kind of
understand, like maybe theblueprint piece of it.
But like, I got a 3D printerright here.
I I mean, I can create whateverin my head and then it can make
it like the like thepossibilities of where things
(32:27):
are going, and then like I mean,you I could just I mean, there's
technology that I could justdrop ship that to somebody super
easy, or you could probably have3D printers that would make
stuff, and then someone has abunch of them working and they
drop ship them for you.
Like, I mean, that's and thenyou talk about organs, they're
doing it with organs, like thewhere everything's going is just
(32:48):
um mind blowing, and marketingis right in the middle of it.
And marketers, I went to thisconference, Andrew.
That that is all agenciesowners, and what they were
telling everybody is you need tolearn to be the AI expert, and
clients are gonna outsource thatto you, and like and that's what
(33:08):
the data is showing, andeverybody was like, I did not go
to college for this.
SPEAKER_02 (33:14):
Like I think it was I think it was Mark Cuban the
other day on TV PN, but likethis is the new category, right?
Is is AI integration.
We had digital transformation,we we had to move into the cloud
at one point.
But like, how do you build newdigitally native businesses?
And those businesses that arenot digitally native, how do you
(33:34):
make that integration seamless?
Like, that is a massivecategory.
And I think that brands andbusinesses don't have enough
internal expertise yet.
Like that will change over time,and then there will, but right
now there's a big chasm.
And I at least for marketersspecifically, I think there's
one luxury is you don't need youdon't need to be all the way at
(33:55):
the bleeding edge because markthe marketing doesn't become
relevant until the consumerbehavior like substantiates it.
So we actually don't need it's aone, we should understand the
systems and talk about itbecause it's important, but like
we sort of get the fast followbecause we don't need to beat
our customers to a channel or toa use case, we just have to be
incredibly responsive andnimble.
SPEAKER_01 (34:16):
And and the clients have a lot of regulation and
internal stuff that they have todeal with, so they want to push
it to us right now because theycan't do it yet, right?
Like, literally, they're tryingto outsource it because they
can't bring it in-house yet,like they haven't all figured it
out.
So now is like this kind ofgolden sweet spot, whatever, for
(34:40):
anybody that's listening that'sin marketing to take advantage
of it and become that expert.
And I mean, there was peoplewhen when chat GPT really came
came on the uh scene, evenpeople in our organization that
didn't know how they felt aboutit and kind of was like, I'm not
sure about this, and people areconcerned.
And I'm like, this is not likejust a fact, this is a trend,
(35:04):
this is a movement, and we gottaget on the span wagon and in
front of this and become AIfirst, AI native as quickly as
we can, because that experiencecompounds.
Like, I see it, and even peopleon my team, I got power users, I
got people that are dabbling,and the power users are
compounding their quick, right?
(35:26):
Yeah, and then like if you're ifyou're getting into it, yeah,
you're gonna compound too, buthow much how much far ahead is
this person or this companygonna be ahead of you before you
jump on it, you know?
SPEAKER_02 (35:40):
100%.
I mean, I I think, and also theit's hard to build really great
marketing when you also havethrashing like at the
foundational level.
I mean, as as we have theability to advertise across AIs,
as we have the ability, ascommerce really like gets its
footing and becomes a real thingthat shoppers rely on, like
(36:01):
there's gonna be a bunch ofchanges.
So I think it actually it's veryhard.
You have to sort of thread theneedle by like this matters.
Brands, marketers, businessesneed to start like understanding
this is how they're chopper andhow businesses buy and search
and use the internet, but at thesame time, there is such a thing
as being too far out in front ofyour skis because things are
(36:21):
gonna change on a dime.
SPEAKER_01 (36:24):
So it's called parsnip, right?
SPEAKER_02 (36:27):
Is that like the vegetable?
Yeah, yeah.
SPEAKER_01 (36:29):
Like everybody that so that's interesting.
Okay, I want to talk about thatfor a second before we get into
it.
Is I feel like I'm seeing, andwe're we're actually launching a
campaign um the next month, nextmonth, we're launching a
campaign next month.
But I'm seeing so many AIcompanies find a a thing, like
(36:52):
what whatever it is, uh bear oflike like so.
In this space, there's bears,there's carrots, there's yeah,
they're they're like any kind oflike uh thing that everybody
knows what it is and canremember it, but then they like
put their spin on it so you youyou can uh subconsciously, I I
forget what it's called, butyou're you're creating a peg to
(37:13):
something that they know about,and so it's helping you with uh
memory recall, but it'ssomething different enough that
it's not AI is gonna know thedifference in whatever it is.
Does that make sense?
Uh 100.
SPEAKER_02 (37:28):
Naming is hard.
Like, I mean, I this it's it'slike one of those silly things
that like entrepreneurs likeit's actually a it's actually a
hurdle.
Um, and I think it'schallenging.
I think it would be to ourbenefit if people liked
parsnips, the vegetable more.
It's a pretty inconsequentialvegetable, so we have to make it
into a more consequentialsoftware business.
But uh, no, but like namings,naming is hard.
SPEAKER_01 (37:50):
So just to finish out kind of what I was saying is
um so we re we used to be callede web results.
We've been around 26 years, ouragency's been around a long
time.
We did web development.
I don't talk about it a lot onthe podcast.
Our audience set of who we workwith and who listens, completely
different.
EWR sponsors podcast, probablyshould do a commercial, but
(38:11):
essentially, um the brand therewas an e in front of everything,
right?
With the internet, right?
Yeah, so e web results, e webstyle, e-web results were really
in vogue, but you know, 10 yearsafter no one uses the e, it was
kind of like we need to updateit.
So we did the KFC thing, that'sall I could think of.
(38:31):
The KFC Kentucky Fried Chickenwent to KFC.
We shortened it to EWR digital,but nobody can remember EWR,
they can't remember it.
SPEAKER_02 (38:41):
Interesting.
SPEAKER_01 (38:43):
Okay, right, and so I'm trying to peg it to
something that people remember.
Yeah, so we're launching acampaign, we're doing a big
rebrand.
EWR uh is an airport tickersymbol.
Anything that's government, youhave no like marketing rights
to, there's huge volume.
Anybody that goes through EWR isusually going international, um,
(39:05):
right?
So our business users aremultinational companies.
We work with a lot of energycompanies, they know of EWR, and
people at business conferenceswill be like, EWR, like are you
out of New York or Newark orlike yeah, like whatever.
So I was just like, Okay, thisis the cloak.
I'm not gonna change the nameagain because a rebrand, and
(39:26):
I've done a number of podcastsabout this.
I would not encourage anybody todo a rebrand if you don't have
to.
It is tons of work, painful,like there's so much legacy
stuff out there.
You're talking about datahygiene, clean, it's bad.
And and and even these thememory sets on the search
engines and now AI, like it'sgonna be difficult.
SPEAKER_02 (39:45):
AI learn faster though, so that's to your
benefit.
As in, it takes a lot longer toclimb the you know the blue
links in traditional search,it's faster.
SPEAKER_01 (39:53):
Yeah, fair enough.
So, so we're gonna be doing thiswhole campaign on like kind of
take yourself to the skies, likelike as a hub, uh, for whatever.
Like, we're gonna do like thisbig play on it.
Yeah, and and then I get andthen my team, we were kind of
debating it.
And uh, sorry, I'm going on atangent, everybody, but we were
doing that.
We're it's interesting, I think.
Hopefully, you'll find itinteresting.
(40:13):
So basically, they're like, No,we're gonna get confused at the
airport.
And I was like, is that a badthing necessarily?
Because AI, I can tell you itlistens to us and it knows like
who we're around and they knowyour behaviors.
So if you're a businesstraveling person and you're
looking for a marketing company,it will probably know, hey, I
(40:37):
don't know which one to showthis to.
But if you're just like aregular user and you're not
looking for a marketing agency,it's gonna know to serve EWR, so
there's gonna be some confusion.
We're gonna get a little bit oflift, I think, temporarily from
EWR.
SPEAKER_03 (40:52):
Pun intended.
SPEAKER_01 (40:53):
Yeah, well, yeah, pun intended, I guess.
Yeah, I gotta uh there's a lotof fun things actually.
I'm thinking from a marketingstandpoint, we could do, but but
same thing, like I just I seethis consistently uh with new
companies as they're tagging itto something of a word or a
topic or an object that peopleare familiar with, and then you
(41:16):
add AI or whatever you're doingon it, and then and now you're
like, Oh, I remember that, andyou know, like it's totally
different than that thing.
So I I I think I'm I'm that wasa long-winded way to
congratulate you on uh thenaming of it is difficult, and
you did something that peoplealready have in their brain.
SPEAKER_02 (41:35):
That's the help.
And we have we have an emoji wecan use because there's a
carrot, so it we can shorthandit to the emoji keyboard if we
need to.
SPEAKER_01 (41:43):
Yeah, so tell tell me a little bit about kind of
how you started and then umwhere you see the opportunity
now because as as we starttalking about at the beginning
of this, AI um visibility maynot be a category, it's just
kind of a standardinfrastructure as we move
forward.
SPEAKER_02 (42:00):
Yeah, I absolutely think that there will a peer
play AI visibility platform willnot be the type of tools we're
talking about.
And I don't know the timehorizon is this six months, this
12 months.
I think you have to understandAI visibility.
I think you need to optimize forit, but that dovetails into the
use cases as a marketer that youactually need to perform, right?
(42:21):
Yeah, like a big piece of thatis content, right?
One of the one of the core waysto go combat traditional SEO and
improve GEO visibility isthrough developing great
content.
Like Parsnip is not a contentcompany, right?
There are some really great AIvisibility tools that in my
mind, they're these are contentcompanies, right?
There will be when I think ofParsnip, we have an AI
(42:45):
visibility tool that doesexactly what brands need.
You figure out how you indextoday, you go through, you
figure out how you rank againstcompetitors, and then you spit
out recommendations on how toget started.
That's cool.
I think the core competency,though, is for us is agenda
commerce.
So figuring out if we have thisreally long customer journey
(43:05):
where you have to build content,you have to update parts of like
you know, website schema andproduct feeds, and you're gonna
be able to advertise across AI.
You will have software toolsthat help, I think, and have
competencies in each of those.
For us, it's figuring out if youunderstand your visibility, how
do you move, how do youintegrate that with the ability
(43:26):
to sell on the surface?
So that and we're e-commercefolks, you know, historically.
That's that's our background.
We built kind of software at theintersection of social media
marketing and e-commerce for thelast decade.
So in many ways, it's the same,you know, just the usage that
we're helping marketers respondto is not people on social
media, it's people using largelanguage models.
SPEAKER_01 (43:48):
Yeah.
And I what you started talkingabout where I keyed in on it
specifically, and I haven't donea podcast since the new Google
Core update release, whichContent AI generated content, a
bunch of uh directory sites gottaken down.
Um there are things that youneed to do when you're
(44:11):
generating content and and toolslike Parsnet um help you uh make
sure you're checking all theboxes, like make sure you're if
you are using AI to helpgenerate the content, um there
there are fundamentals andthings that you need to be doing
to make sure that you don't gettaken down and usually attaching
(44:33):
it to a human, there's kind ofeats that you need to focus on,
expertise, authority, trust,experience.
Um, I mean, Google came out andI thought they waved the light
flag and said, hey, if it'suseful content, it's fine.
I feel like AI generatedgenerated content was pretty
targeted uh in this last update.
Um, but there's things that youcan do to mitigate it, there's
(44:55):
things that you need to bethinking about, and it's
constantly evolving uh to yourto your point, Andrew.
And and the consulting andstrategy piece, I think is big
of what people are looking forand in tools providing that that
that knowledge to say, what do Ineed to be focused on?
What do I need to be learningat?
Because I'm not going to be theexpert in this.
(45:16):
I'm trying to do my job and allthis other stuff.
And so I I think these kind oftools are really, really helpful
to give you almost like achecklist or or synthesize it
down or abstract it out so thatyou kind of have a direction of
like these are the buoys that Ineed to start to hit or go in
that direction.
SPEAKER_02 (45:35):
But it's because most folks, like we like we nerd
out and talk about these thingsall day, but most folks go,
okay, I understand this, youknow, I understand people are
using chat GPT every day andhaving conversations I care
about for my business, but theydon't know where to start.
So when I think of when I thinkof what Parsnip looks like
today, it makes it as easy aspossible to understand where you
(45:56):
stand and how you get started.
What are the 30 things youshould start doing?
I don't think that's the fullpicture.
And I all and so I think we'regonna see the evolution really
fast.
Like we made the decision not togo deep down the content route.
I think there will be greatcontent companies that support
GEO and AI visibility.
And I'm also a little, these areput these are my words, you
(46:18):
know, this wasn't part of theGoogle announcement, but like
why AI content was targeted.
I don't think it was targetedbecause AI can't help build
great useful content.
I think the problem was some ofthe early GEO strategies are
kind of like when we used to putclear text on websites for
keywords in old school search.
You know, you'd have yourwebsite and then you'd have your
(46:39):
keyword 30 times in clear font,you know, so that you know
Google pick.
I think that we there's a therewas a little bit of we're gonna
have use AI to generatederivative content, answer the
same thing 50 different waysthat was not useful.
So, like yeah, yeah, it's spamfilter.
SPEAKER_01 (46:57):
Yeah, yeah, it's just a spam filter.
It's like it's not genuinelyuseful content.
Yeah, yeah.
Hey, so uh I uh I was talking tosomebody and I I use the term AI
slop and they got a theyactually got offended.
It was like a big like, and thenlike I'm like, this is the word
of the year for like uh whatwhat what what uh what um
(47:20):
dictionary was it?
Was it Webster or something?
That AI slop was the word of theyear.
Oh, where they do word of theyear, yes.
Yeah, it was AI slop.
I was like, I was like, sorry,like I didn't think that this
was offensive, but um, you gotto be careful AI slop.
SPEAKER_02 (47:36):
So um it's it's about usefulness and to your
point, like e like AI.
So by the way, there's great AIcontent out there, there's also
really horrible AI content outthere, and so I think using it's
like using the right tool tosolve the right problem, like AI
will be the answer in a lot ofthose contexts, but it doesn't
mean that you flood the system.
(47:57):
It's also not the way theinternet can work.
If agents are havingconversations with themselves
proactively on behalf of theshopper to infinity, like nobody
wins ultimately.
Like that's a that's a badformat.
SPEAKER_01 (48:10):
So, so Andrew, we gotta we're getting close to
time here.
I gotta wrap up.
Um, I would love to hear likewhat is the biggest takeaway
that with companies you'veworked with consistently on the
e-commerce side that they'vejust failed to do, or there's a
gap, a blind spot in that theycould help improve uh their
(48:31):
their whatever they need to do,ranking, visibility, access to
the right customers, kind of inthat vein.
And then share a little bit moreabout uh where you're posting
online, like kind of where yourconversations are happening, so
people can follow you and uhplug whatever you would like to
plug.
SPEAKER_02 (48:47):
Absolutely.
Well, thank you.
Thank you for having me as well.
Been a fun conversation.
Um, I think the biggest, thebiggest issue most folks have is
it's kind of like analysis, it'sparalysis through analysis, is
everyone is having these typesof conversations, like the one
we had today, and assume theyhave to like hop over the chasm.
Specifically with commerce,there are baby steps, and there
(49:09):
are things that if you market orsell online, things you
absolutely need to be doing thatare not overly challenging.
It's just more likely than not,this isn't a marketing
discipline that your team is upto speed on.
So it there are ways to takebite-sized chunks, is the
biggest feedback.
There are ways to make sure youhave a correct tracking pixel
for Chat GPT, or integrate aproduct feed, or make sure that
(49:33):
your product catalog isorganized with a schema that
LLMs can read.
Like very simple stuff thatevery e-commerce professional or
digital marketer, theyunderstand the tactics, but I
think there's a disconnectbetween all the things they're
really good at and they've donetheir whole careers and AI.
And a lot there's not enoughconversation that's connecting
(49:54):
them.
So they go, well, AI is this bigquagmire, we need to figure it
out.
And then they understand taggingand product feeds and tracking
and content marketing.
And so figure out the way totake a few small steps and start
working in the direction.
SPEAKER_01 (50:08):
Love it, love it.
SPEAKER_02 (50:09):
Um then obviously we, you know, part snip is we
are we launched the business inNovember, uh, but we do a lot on
LinkedIn uh and our website arethe easiest places, but we've
tried to make it's a it's a freetool to try.
We've tried to kind of put ourmoney where our mouth is in
terms of make it really easy toget started and just start to
figure out what the roadmapshould look like.
SPEAKER_01 (50:30):
Very cool.
Well, Andrew, thank you so muchfor coming on.
I think this is a greatdiscussion.
Um, I think that you know, youget a bunch of PhDs in the room
and they're talking about abunch of AI stuff uh on both
sides, and nothing gets done.
And you gotta you gotta havesomebody that comes in and
translates it.
And hopefully we we did that foryou today, uh, for people
(50:51):
listening, like understandingwhere things are going, maybe
some action steps and kind ofgiving you that general
direction to take the next stepand to move forward.
We do do coaching and classes ifyou want to reach out, uh find
out more information about that.
Uh, bestseopodcast.com.
Um until the next time, guys.
(51:13):
My name is Matt Bertram.
Uh, if you want to grow yourbusiness with the largest, most
powerful tool on the planet,which maybe is AI, not the
internet anymore, uh, reach outto EWR for more revenue in your
business.
Thank you so much, Andrew, forcoming on.
Thank you, everybody, forlistening.
Talk to you later.
Bye bye for now.
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.
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.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com