October 7, 2025 • 35 mins
Guy Yalif is Chief Evangelist at Webflow and a veteran B2B marketing leader with 20+ years across Twitter, Yahoo, BrightRoll, and as co-founder/CEO of Intellimize (acquired by Webflow). He champions AI-driven optimization for sites and content, bringing a rare blend of aerospace-engineer rigor and operator experience from four successful exits to help teams win the shift from SEO to AEO.
Discussed in this episode
- Why AEO is an evolution of SEO (and what truly changes)
- The shift from keywords to “clusters of questions” as the new topic model
- Webflow’s four-part AEO framework: content, technical, authority, measurement
- Tactics that moved the needle: adding FAQs + schema; prioritizing freshness
- Why PR/brand and plain-text mentions matter more to AI engines
- How to measure AEO: presence in questions, share of voice, and sentiment
- Where to start: two moves any founder can ship this week
- Risks of ignoring AEO and the early-adopter advantage
Episode highlights
00:21 — “Your SEO resources are your AEO resources. This is an evolution, not a reset.”
01:15 — Webflow’s AEO promise: answer engines are a massive arbitrage—akin to early SEO/SEM/mobile.
03:00 — Why “ranking for keywords” is obsolete; topics = clusters of questions across the funnel.
07:49 — The 4-part AEO framework: content, technical (schema & structure), authority, measurement.
10:11 — Case study: Add ~6 FAQs + inline schema to product pages → half of new citations came from 6 pages; +24% organic in 2 weeks.
15:23 — If you only do two things: (1) answer questions comprehensively, (2) add schema metadata.
21:46 — Webflow data: AI-search traffic converts ~6x better than non-branded organic; unbranded share grew from 0% → 42% in a year.
24:02 — How buyers actually use LLMs in-flow; why your website still matters (to humans and machines).
29:58 — The learning curve is back: why AEO is resetting the playing field and rewarding curiosity.
This episode is brought to you by our sponsor:
ZoomInfo is the GTM Intelligence Platform built for sales, marketing, and RevOps.
By unifying data, workflows, and insights into a single system, ZoomInfo helps revenue teams find and engage the right buyers, launch go-to-market plays faster, and drive predictable growth.
With industry-leading accuracy and depth of data, it gives your team the intelligence advantage to win in competitive markets.
It’s trusted by the fastest-growing companies and has become the category leader in GTM Intelligence.
Learn more at zoominfo.com.
Follow Guy Yalif:
- LinkedIn: https://www.linkedin.com/in/gyalif/
- X (Twitter): https://x.com/gyalifX (formerly Twitter)
- Webflow author page: https://webflow.com/people/guy-yalif
The GTMnow Podcast
The GTMnow Podcast is a weekly podcast featuring interviews with the top 1% GTM executives, VCs, and founders. Conversations reveal the unshared details behind how they have grown companies, and the go-to-market strategies responsible for shaping that growth.
Visit gtmnow.com for more episodes and other interesting content.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_02 (00:00):
AI is advancing at such a rapid rate.
SPEAKER_00 (00:02):
Do you mean with LLMs crawling the web almost
every query, you can have animpact this afternoon.
You can use it to outsource yourthinking.
You can use it to think moredeeply.
And door number two seems to beranking a whole lot better.
SPEAKER_01 (00:17):
And what else do people need to really think
about at the macro scale of AEO?
SPEAKER_00 (00:21):
Somebody was like, hey, if this Fortune 500 CMO
fired his or her entire SEOteam, we're doing a huge
mistake.
Your SEO resources are your AEOresources.
Your SEO agency is your AEOagency because this is an
evolution.
SPEAKER_02 (01:07):
That's where this conversation begins.
Because today, answer enginesfrom ChatGPD to perplexity and
more are reshaping how buyersfind products.
And for founders and operators,this is one of the biggest
arbitrage opportunities sincethe early days of SEO.
Gaia Leaf, chief evangelist atWebflow, has seen this movie
before.
And because he has, he's takingaction and has the data to show.
SPEAKER_00 (01:29):
In two weeks, half of the incremental citations
Webflow got were from those sixpages.
And organic traffic to thosepages is up 24%.
SPEAKER_02 (01:37):
In this conversation, you'll learn
Webflow's four-part AEOframework.
Why answering questions, notkeywords, is the new growth
edge, and how early adopters canturn this uncertainty into
leverage.
SPEAKER_00 (01:48):
Or, as Guy puts it, the people that went earlier,
they found relative bargains.
SPEAKER_02 (01:53):
In other words, the kind of upside we haven't seen
since the early days of search,SEM, and mobile.
All right, let's get into it.
Guy, welcome to the podcast.
SPEAKER_00 (02:02):
Thanks, Sophie.
Thanks for having me.
Excited to talk with you again.
SPEAKER_02 (02:05):
It is great to have you here.
And nobody better to learn aboutAEO from right now at this
inflection point.
And you've lived through the bigSEO boom.
You've felt it.
What does this inflection pointfeel like compared to that?
SPEAKER_00 (02:22):
They're very similar, actually.
Um like the early days ofsearch, there are a bunch of
different players.
Things are changing all thetime, like they did back then.
I mean, it makes the news nowthat Google updated their
algorithm because it happens afew times a year, because it's a
settled space.
And here, you know, things arechanging very rapidly.
(02:44):
The rules of what matter areupdating frequently enough that
everybody feels behind, andthere's real opportunity.
Like any new medium, those thattake action early, they'll find
bargains, basically, effectivebargains, you know, return
relative to the effort they putin.
SPEAKER_02 (03:01):
Mm-hmm.
That's really interesting.
And if you were explaining thiskind of new world, even though
it feels a little bit moresimilar, but to a founder, what
are the key differences?
SPEAKER_00 (03:11):
I think there are some that are saying, look, SEO
is dead, and there's this awhole new world, respectfully.
I 180 degrees, black and whitecould not disagree more.
I think AEO is an evolution ofwhat good SEO always should have
been.
You know, genuinely valuableoriginal content.
It's getting rewarded.
(03:32):
The the LLMs, they get to seemore context about what's good.
When we all learn to speakGoogle, our average query was
four words long.
The average query in an LLM is23 words long.
So they get more context, theyhave a conversation.
So if I were explaining it to afounder, I would say, look, take
SEO, minus one thing plus acouple.
The thing that's minus is theconcept of ranking for keywords
(03:55):
doesn't exist anymore.
Because you got paragraphs now.
What's the plus?
Plus is topics used to be abasket of keywords, a cluster of
keywords, and you tried to makesure you had them in your
content.
Now a topic is a group ofquestions that your prospects
may be asking in the funnelclustered together.
(04:15):
So shift from counting keywordsto answering questions, that'd
be one.
Two, help the LLM understand thestructure and meaning of your
content.
We've been doing that with SEOforever, right?
We have the metadata.
There's a bit more.
Do this thing called schema.
And the last would be we focusedon backlinks for years.
They still matter.
They don't not matter, butincrementally more important
(04:38):
than an SEO is repeated mentionsin plain text on other sites
saying, oh, GTM Now is awesome.
Oh, this is one of the toppodcasts that matters more than
it did with search.
SPEAKER_02 (04:49):
Mm-hmm.
Mm-hmm.
The pendulum is almost shiftingback to that third party.
It's bringing PR and all thesedifferent things back that, you
know, had had the moment of notnot mattering less by any means,
but maybe less emphasis on them.
And now are getting highlightedmore, which is really
interesting.
SPEAKER_00 (05:04):
VR's back, brand is back.
Those things matter more thanthey did before, yes.
SPEAKER_02 (05:09):
Yeah.
And those things matter.
So we know definitively what'simportant.
But what what about overall justknowing?
Like, how do you know?
Does anyone actually know what'sworking with AEO?
Or are these best guesses?
How are we thinking about asingle source of truth around
AEO?
SPEAKER_00 (05:28):
So you said definitively no, and I'm glad
you asked the question becauseanyone who says they
definitively know is lying.
Like no one knows.
But in the same vein that no oneknows for SEO, right?
We all we all try a bunch ofthings.
We observe the results of anexperiment.
Oh, I added more keywords, fewerkeywords, longer metadata,
shorter metadata.
And they saw what works.
(05:49):
But no one actually knows what'shappening in the Google
algorithm.
Now, with the LLMs, not even theLLM providers themselves know
what's actually happening inthere.
So when we talk about theserules of thumb, they are very
well educated, very wellresearched, data-driven guesses.
They're the current bestthinking uh about what one
(06:10):
should do to go manage the riskand opportunity of LLNs, which
is really to take back controlof our brand narrative and drive
more traffic to our websites togo generate revenue.
SPEAKER_02 (06:22):
Absolutely.
And AI is advancing at such arapid rate.
Yeah, I never heard of it.
What what do you think thatchanges about the SEO adoption
to AEO?
So when you say they're alleducated best guesses, are those
guesses changing at a pace thatis faster than SEO?
(06:43):
Or would you say they'reevolving at the same pace as
they did?
SPEAKER_00 (06:46):
My humble opinion, it's faster, exactly for the
reason you described.
The LLMs are improving theirability to understand meaning
and structure and content, whaton like a monthly basis every
couple of months?
A hat tip to Google and Yahooand Bing and everyone, but like
it didn't go that fast.
I don't recall from back in theday that it did.
Um mercifully, we have skippedover a whole generation of spam
(07:11):
where if you look at the stuffthat ranks top in Google, 85% of
it is either entirely humangenerated, that's about 65%, and
another 20% is like a little bitof AI.
So the notion that I can createenough infinite content with an
LLM and I'm just gonna go spamout zillions of pages, it's not
working.
Thankfully.
SPEAKER_02 (07:32):
Thankfully.
Less noise.
Yes.
Less noise is always good.
And when you're thinking aboutoverall frameworks or advice to
people that are leaning intoAEO, do you have any kind of
recommendations?
I know at Webflow, you're doinga ton of work in the space and
know those best practices andhave implemented them yourself.
We'd love to hear a little bitmore around that work.
(07:53):
And I guess your best educatedguess, not necessarily what is
technically working, but whatare your best educated guesses
right now?
SPEAKER_00 (08:00):
Um Vivian, who leads our SEO and AEO team, has done
some amazing work.
And uh, we went and looked forexamples, case studies, spent
lots of time, many hourslooking.
There aren't any.
My best guess about why is thatpeople are doing the work.
They aren't yet, it's not yetrepeatable enough that they're
sharing it externally.
(08:21):
So we're sharing a lot of whatwe're doing externally.
All the research we've done,talking with other experts in
the space, looking at the data,humbly, we would suggest
thinking about the transitionfrom SEO to AEO in four
categories.
The first is content, where yougo from counting keywords to
answering clusters of questionsto making content super relevant
(08:45):
with personalization.
So this is all about like thecontent you own.
The second is technical, whereyou go from on-page SEO to
helping LLMs understand thestructure and meaning of your
content by giving them moremetadata.
Like we all did it withFacebook's Open Graph.
SPEAKER_03 (09:01):
Yeah.
SPEAKER_00 (09:02):
Now there's this thing that many know about, but
uh many also don't call itschema, that is just another
form of metadata that Google,Bing, Yahoo, Yandex, and others
came up with.
And it's like, oh, here's howyou specify an events page, a
product page, a bio page.
88% of sites don't have schema,but 73% of Google's top results
(09:23):
do.
So, like, tons of opportunitythere.
Coming back to technical, andyou want that on a super fast
site global.
Third area is authority, wherewe went from backlinks that
still matter, they don't notmatter, to repeated positive,
plain text mentions about yourbrand pointing to a visually
(09:44):
stunning, emotionally evocativeexperience because that's a
signal about authority too.
And the last is measurement,where we go from keyword
ranking, which no longer exists,to did I show up in the
questions I care about?
To what's my share of voicerelative to my competitors and
what's the sentiment?
Because over here in search, theGoogle is using our words.
(10:06):
Here the LLMs are reformulatingeverything.
So the notion of sentiment beingimportant exists.
So those four buckets uhcontent, technical, authority,
and measurement.
SPEAKER_02 (10:16):
Incredible.
And from implementing those, yousaid the team's been doing a
great job.
What kind of results have youseen?
SPEAKER_00 (10:22):
Vivian and team have done a bunch.
They partner with our SEO andAEO firm Graphite, and a couple
of things come to mind.
Um, she recently took our tophalf dozen product pages and
tried to put some structure onthem so the LLMs could
understand the meaning andstructure of them.
What did she do?
She put FAQs at the bottom ofthe page, so about half a dozen
(10:44):
FAQs.
She then put schema, thatmetadata, there in line with
those FAQs.
In two weeks, half of theincremental citations Webflow
got were from those six pages.
Six pages out of thousands thatwe have.
Yeah.
And organic traffic to thosepages is up 24% in two weeks.
So that was one thing she did,and it's great, and that's the
(11:06):
beginning.
We're going to do more.
The second is that freshnessmatters.
So the content machine is ashungry as ever for good original
content.
Um 85, maybe 95% of the contentGB chat GPT sites was updated in
the last 10 months.
So she worked with anotherpartner of ours, AirOps, to
(11:27):
increase the speed with which werefreshed content.
She did that and uh found 20%,20, 40, 40% more traffic coming
from AI search because we wereupdating the content more
regularly.
Uh, and and we see that contentwith freshness indicator, like
(11:49):
last updated date, gets 1.8xmore citations.
So giving structure and updatingmore regularly.
The two experiments that haveworked really well for us.
SPEAKER_02 (11:58):
Wow.
That's a huge difference justfrom updating content.
Yeah.
What and maybe this is gettinginto the tactics.
I need to give her a call, butwhat constitutes updating?
How much work would you say?
There's there's a lot ofcompanies essentially that have
a ton of content, but it's notnecessarily optimized for AI
engines and they want to refreshthem.
(12:19):
How much work is it to refreshexisting content?
SPEAKER_00 (12:23):
The direct answer is I don't know.
SPEAKER_02 (12:25):
Yeah.
SPEAKER_00 (12:26):
Because I didn't, I have Vivian and Devontae.
Totally.
I do believe first, it's worthrunning the experiment.
Somebody said, hey, wait aminute.
If you need fresh content, whatif you just change the last
updated?
Is that enough?
I hope it's not.
SPEAKER_02 (12:40):
It's not.
At least from the SEO side, Iknow it's not.
I I don't know if AI enginesdiffer, but from the SEO side,
you need something in the mainbody.
You need like a little bit of amain point and the date.
So I would imagine it's thesame.
But I'm I don't know.
I gotta call Vivian.
SPEAKER_00 (12:56):
One, I would bet without data.
All of those stuff was datadriven.
This is hypothesis.
And the other hypothesis wouldbe that um AirOps, the tool we
partner with, makes it easy todo that in a workflow and with
LLMs, but that can't be it.
Right.
If you just pump out an LLMupdated page, you need human
(13:16):
review, you need humaneditorial.
I've got friends who startedcompanies, maybe GTM Fund
invested in one, where it'slike, I can detect if this was
entirely LLM created usingstatistical pattern matching on
the content.
And so uh we touch with thehuman every piece of content in
a material way before it goesout the door.
SPEAKER_01 (13:36):
Interesting.
SPEAKER_00 (13:37):
So both hypotheses rather than data driven, but
you've got the SEO data there.
SPEAKER_02 (13:42):
Yeah.
And it sounds like the commontheme of you can't rely on AI to
get from zero to a hundred.
It'll take you part of the waythere, but you do need that
process for actuallyimplementing the human touch.
SPEAKER_00 (13:54):
Totally.
And I feel like one of the oneof the underspoken but
foundational things in LLMs aslike kindergartners through to
CGOs learn about them and howthey change everything, is you
can use it to outsource yourthinking.
You can use it to think moredeeply.
And door number two seems to beranking a whole lot better with
(14:18):
answer engines than door numberone.
And by the way, it helps makeall of us more employable.
So I feel like it's something wearen't talking about enough, but
it's it looks the same on thesurface you used in LLM.
Yeah.
But what's going on inside yourhead could be completely
different.
SPEAKER_02 (14:33):
And it's tempting to go the quick route, but always
play the long game and actuallyinvest in the depth.
A quick pause to tell you abouta company you need to know.
ZoomInfo is the go-to-marketintelligence platform built for
sales, marketing, and reps.
By unifying data, workflows, andinsights into a single system,
ZoomInfo helps revenue teamsfind and engage the right
buyers, launch go-to-marketplacefaster, and drive predictable
(14:55):
growth.
With industry leading accuracyand depth of data, it gives your
team the intelligence advantageto win in competitive markets.
It's trusted by the fastestgrowing companies and has become
the category leader ingo-to-market intelligence.
Learn more at zoominfo.com.
You mentioned two points thatyou did that had a huge impact,
and one of them was updating andrefreshing pages for, let's say,
(15:18):
founders or early stagecompanies that are really
looking in to implement AEOtactics, but there's a lot out
there.
You listed off a lot ofdifferent really interesting
points that people could diveinto.
What are the two big ones thatyou'd recommend?
Are those the two, or would yourecommend two others?
SPEAKER_00 (15:35):
If you only did two things, I would suggest you
shift, you answer questions.
So shift from keywords toquestions and answer them
comprehensively.
And the second is add schemametadata to your site so that
you can help the LLMs understandthe meaning of your site more
effectively.
Those are the easily the top twothings I would suggest doing.
They're the right place tostart.
SPEAKER_02 (15:56):
Keywords, metadata, answers.
Brilliant.
Yes.
And answers.
Sorry.
No, no.
This is how much has been grayedin my brain.
SPEAKER_00 (16:05):
You too have many years in search.
SPEAKER_02 (16:07):
I I do, yeah.
It's a bit of a rewiring.
So it is a process.
Okay.
SPEAKER_00 (16:12):
Super helpful.
But your point about therewiring, it's not like throw it
all out.
It's adjustments.
It's an evolution rather than awholesale revolution.
SPEAKER_01 (16:21):
Mm-hmm.
Mm-hmm.
Okay.
And what else do people need toreally think about at the macro
scale of AEO?
SPEAKER_00 (16:29):
I heard, I hope it's not true.
It's not validated at all, butat one of these conferences,
somebody was like, hey, if thisFortune 500 CMO fired his or her
entire SEO team and is like,we're doing AEO.
I think it's a huge mistake.
Your SEO resources are your AEOresources.
Your SEO agency is your AEOagency because this is an
evolution.
So SEO is dead and this is acomplete revolution, I think, is
(16:52):
one of the easiest ones to seesome great articles about.
And uh I I I I don't think thedata supports it.
Um and the second is just likewe did with SEO, you can start
small and grow.
In the example we talked aboutwith Webflow, Vivian chose half
a dozen pages, saw real results.
Obviously, she's then gonna gocontinue, but she didn't do
(17:14):
everything all at once day one.
Mm-hmm.
SPEAKER_02 (17:17):
Mm-hmm.
No, that makes sense.
Okay.
Super helpful advice.
I love it.
It'll be interesting to see howit plays out now over time.
SPEAKER_00 (17:26):
To your point, we are all learning together,
right?
This this mental model we haveof these four categories.
We also made up a rubric of fivedifferent levels of them, but
it's idle keeper justice, as weall know.
SPEAKER_02 (17:39):
And how do you learn?
How do you learn about AEO andthen AI overall?
SPEAKER_00 (17:47):
A bunch of time, really.
It's it is it is it is a bunchof time experimenting, focusing
on data, because it's it's thereare uh a lot of things being
said.
Uh and when like I'll read anarticle about something and I'll
dig into the source for that andthe source and the source and
the source.
And sometimes I'm like, wait,that article, it contradicts the
(18:09):
original underlying data, but itwas like a telephone game of
referencing each other uh andthen staying close to the work
we're doing, experimentingdirectly uh with both.
I happen to like two podcasts.
One is this one, which I listento religiously, and the other
one is um at Marketing AIInstitute's uh uh podcast, which
(18:34):
uh Paul and Mike, they producegreat content too.
SPEAKER_02 (18:37):
Okay.
I love it.
I think that's where at with thefast evolution people are
learning the most on the fly andour individuals, our networks
like podcasts and channels andso forth, where it's just a
little bit more live and andflexible.
I know that's where personally Ilearned the most too is podcasts
from people individually, howthey're learning.
We're seeing a lot more contentout there that is dynamic,
(18:59):
that's actually tactical,actionable advice on AEI.
And I know you're producing alot of content around that too,
actually.
So for for anyone kind ofcurious about their own AEO
strategy, what kind of resourcesare out there?
Because I know that you are areworking on a ton.
SPEAKER_00 (19:17):
We wouldn't I would invite people to go to
webflow.com slash resourcesslash AEO.
There, you will not find apitchful webflow.
What you will find is a place togo enter your domain name, hit
enter, a couple of minuteslater, you will get an email
saying, we ran in the backgrounda 600-line prompt.
(19:37):
It's about to become 1700 in V2that goes and looks at your
site, goes and looks off yoursite, and suggests, hey, here's
what we see in those fourcategories: content, tech,
authority, and measurement.
Here are the gaps we see.
Here are suggestions for whatyou might do to level up.
And it bottom lines it, like youwere doing earlier.
If you only did two things, godo these.
(19:58):
And then below that is detailaround these four categories and
the things we're seeing that canhelp people evolve.
SPEAKER_02 (20:07):
Amazing.
And no cost?
No cost.
No cost.
Okay.
Well, then we'll drop it in theshow notes.
That's amazing.
Yeah, what a great resource,especially after you just broke
down those four points andthey're super, super important
for people optimizing.
And what about the people thataren't optimizing?
What do you think happens ifpeople aren't actually following
an AEO strategy right now andinto the future?
SPEAKER_00 (20:28):
In my humble opinion, is it's it's missed
opportunity.
I think AEO is truly threat andopportunity for us.
Over time, we will lose controlof our brand narrative because
our carefully differentiated andcrafted words are being
reformulated.
We'll lose traffic.
Bain found 15 to 25% decrease intraffic, which I think is closer
(20:49):
to reality than the headlinegrabbing, oh, Spot lost 80% of
their traffic.
I don't think that's the normand was cherry-picked for them
as an example.
And the third is we all soundthe same.
We're all using the same LLM.
So how many people are, youknow, the leader, the best, the
first, the fastest?
Uh and I think um uh actuallyI've lost the train of thought
(21:10):
entirely.
Where was that?
SPEAKER_02 (21:11):
Okay.
Um what happens if people Oharen't doing it.
SPEAKER_00 (21:15):
Thank you.
SPEAKER_02 (21:15):
Don't yeah.
SPEAKER_00 (21:16):
So then so the risk is we all sound the same.
If we then go invest some inAEO, which is with your existing
resources, same hours, samepeople, shifting strategy a
little bit.
You will be earlier to market,and chances are you will find
opportunity.
Just like we did in the earlydays of search, in the early
days of SCM, in the early daysof mobile, in the early days of
(21:38):
native, the people that wentearlier, they found relative
bargains.
They got more return relative tothe amount of effort they were
putting in before everybody wasthere.
SPEAKER_02 (21:48):
Early adopters, advantage.
Yes.
And is there a certainpercentage that you'd recommend
teams think about ofreallocating resources from SEO
to AEO under the same team?
SPEAKER_00 (21:59):
I think it depends on how where they are seeing
traffic coming from and how faralong they are in their SEO
journey.
So I I I do suggest teams golook at traffic to their site.
At Webflow, 8% of our self-servesignups come from Answer engines
today, AI search.
Uh by the way, it convertsbetter too, because they're
further down the funnel.
It converts six times, not sixpercent, six times better than
(22:22):
non-branded organic, which iseye-opening.
But so I suggest those to youlook at how much traffic you're
getting from Answer engines.
Look if it's increasing.
For us, in unbranded, it wentfrom zero a year ago to 42%
today.
unknown (22:38):
Okay.
SPEAKER_03 (22:38):
Yeah.
SPEAKER_00 (22:39):
It's it's like it's and so that's worth investing
time in.
That's a clear signal to goinvest in AEO strategies.
Um, that having been said, 25%of the top pages out there don't
even have basic SEO metadata.
They have broken links.
That's also worth tackling ifyou don't have those
fundamentals in place.
SPEAKER_02 (22:57):
A very good call out.
And perhaps they can be done intandem, but maybe we're skipping
steps by jumping ahead whenreally there's a lot of SEO
leverage to be had too.
But 6X, that's a wild statistic.
Like talk to me about what youmean around that traffic is
converting 6x better.
SPEAKER_00 (23:16):
I that the traffic that comes from answer engines,
you look at touch to purchase,because here we're talking about
self-search sign-up.
Yeah.
That is 24x 24% conversion rate.
Non-branded organic search is 4%conversion rate for that metric.
Two of the prominent SEO firms,HREFs and SEMrush, found in
(23:39):
their own data 4x and 23x.
So our 6x is in the range.
SPEAKER_01 (23:44):
That's really interesting.
SPEAKER_00 (23:46):
It's uh and why we think because folks are further
down the funnel.
So yeah, less traffic, but it'smore qualified because they have
self-qualified.
SPEAKER_02 (23:56):
So what do you think behaviorally they're doing?
Because I'm thinking to youknow, my procurement processes
and yours and the way everybodyindividually operates.
Are they researching on theirown and then turning to LLMs?
Are we not starting with LLMs,or do you think people are
starting with LLMs?
SPEAKER_00 (24:14):
Forrester surveyed B2B marketers.
95% of them said LLMs are goingto be part of my buying process
this year.
SPEAKER_03 (24:21):
Yeah.
SPEAKER_00 (24:21):
Answer engines will be.
So it's clearly something thatis important.
Did they literally start there?
Did they start with search?
Did they start with friends?
I don't know.
Anecdotally, I think word ofmouth is easily continues to be
the most trusted.
That's quantitative.
It is the most trusted form ofrecommendation.
By the way, your website'snumber two.
That was eye-opening.
SPEAKER_02 (24:42):
Um uh and so your website, so your website isn't
dead.
SPEAKER_00 (24:50):
Your website's not dead at all.
In fact, I firmly believe thatpeople are going to want to hear
your story told your wayforever.
That is not going away.
To your point about how we'reengaging with LLMs, often the
first look is there, and peopleare learning.
Well, what matters in thiscategory?
Who are the players?
Some of the tofu mofu thingsthat we've all done, but then
(25:11):
they are looking to see yourstory told your way.
By the way, that website needsto then speak to humans in
visually stunning, emotionallyinvocative, engaging ways, and
to machines.
And folks are talking about thislike it's new, but it's not.
We've had search metadataforever.
Now we have a bit more metadatafor the LLMs.
And so we've been doing this fora while.
(25:33):
I think folks are, this isobviously more in the zeitgeist
now.
You need websites that speak toboth.
SPEAKER_02 (25:39):
And you've obviously built in this space and solved
in this space before.
You know it so intimately.
One of the quotes that I justabsolutely love, and I think
about all the time around thework that you've done, the work
that I've done in the past, thework that a lot of people are
doing right now is people imputethe quality of your product or
service from the quality of yourwebsite.
So even if there are thirdparties like LLMs, it is your
(26:02):
story.
You're still owning your storyand design and visually
appealing sites and emotionallyinvocative, like you said, are
huge, huge part of that.
SPEAKER_00 (26:11):
I am very, very much with you, both quantitatively
and just subjectively as anindividual.
Uh and I think the LLMs aregoing to use it as a signal of
quality.
That's belief.
Why do I believe that?
Because Google looks at bouncerate in search, right?
Did somebody leave your site andcome back immediately?
It would be surprising if the uhanswer engines weren't doing
(26:34):
that.
And you see perplexity as abrowser, ChatGPT is rumored to
be creating one.
Why might they do that?
I I believe two reasons.
One, to not pay the gajilliondollars, Google's paying Apple
to be the default and Safari,fine.
Perhaps more valuably, they getto see where you're going.
They get to see what's actuallyvaluable content.
(26:54):
And so you having a compellinghuman-focused story on your
website will continue to mattera lot.
It may actually get you toappear in the answer engines
more often because they can tellit's valuable content.
SPEAKER_02 (27:08):
So the quality metric matters a lot more now,
it sounds like it is definitelya signal.
Yeah.
Someone told me a very tacticalnugget around AEO.
And it was that just as yousaid, your website matters so
much.
Yes.
So that pairing it with theformats that are performing well
is really impactful.
So publishing a listicle fromyour own website of the top
(27:31):
companies in your area actuallyperforms really well.
So for a venture firm, forexample, that would be, you
know, top early stage venturefirms and listing out the top
15.
And you could obviously positionyourself first or wherever you
want on the list.
It feels a bit funny, but theywere saying that this has been
uh been a huge incrementaltactic and one of the many check
(27:53):
boxes on AEO.
But it feels a little funny.
I know we personally haven'tdone it, but there's a lot of uh
a lot of interesting thingsaround that.
SPEAKER_00 (28:01):
Without a doubt, it feels weird to name your
competitors and compliment themon your site.
SPEAKER_02 (28:05):
Or just have yourself listed first on your
own listicle, even that alsofeels weird.
SPEAKER_00 (28:10):
And I think the underlying principle is that in
the content, separate from themetadata we've been talking
about, you can also signalstructure to LLMs, a listical
bullets to use like in summary,to put FAQs at the bottom, to
use one H1 on the page, and thenlogically structured H2, H3s,
(28:30):
H4s.
I can tell you how many sitesyou're like, H1, then H3, then
H2, that and you're like, holdon.
There, they're you're you'resignaling structure to have
accessibility info.
They use that, they use thatactively.
And so Listical is a great one.
The other underlying principlethat we see is in search, it
took a while to have an impactbecause you needed to create
(28:53):
authority, and that was based onbacklinks, and that took time.
With LLMs crawling the webalmost every query, you can have
an impact this afternoon.
You can go see the impact somuch more quickly, and it's even
multiplied when you and yourcompetitors all do that
listicle, which feels superweird.
But you can collectively go dothat and have an impact this
(29:15):
afternoon.
By the way, the flip side, thepainful side of that coin is you
change your positioning, you'vechanged it on your website
today.
It may or may not have an impactbecause lots of other people
still have your old positioningthere.
And that can have an impact.
Or I'm sure none of thelisteners here have this
problem, but if you have someoutdated content on your site,
(29:37):
again, nobody listening to thishas any of that.
SPEAKER_02 (29:38):
Of course not.
SPEAKER_00 (29:39):
But if they did, that also will have an impact.
And so the notion of taking downcontent, which is anathema in
search, is a thing in AEO.
SPEAKER_02 (29:48):
Mm-hmm.
Fascinating.
Well, guy, thank you for theexpertise around AEO.
This has been fantastic.
I think you said it well aroundnobody really has the answers,
but you yourself are somebody.
That's been in search for 17years.
You're at one of the companiesleading in the space.
So we appreciate your thoughtleadership on it and better
informing the insider scoop.
SPEAKER_00 (30:11):
Thanks for the chance to talk about it.
And I'm curious.
It was great to learn from youhere.
And I'm eager to continue theconversation and learn from the
listeners as we all rundifferent experiments and learn
in this rapidly evolving space.
SPEAKER_02 (30:21):
Yeah.
I think that you just hit on areally interesting point there,
too, is it almost resets theplaying field.
AI resets the playing fieldwhere suddenly everyone is
learning something at the sametime.
And so it really democratizesexperience in a way.
How does that feel as somebodythat has, you know, built and
(30:45):
successfully sold a company andoperated at the highest level?
Like, what does that feel liketo suddenly have something that
is so new?
SPEAKER_00 (30:56):
Joyful.
It's awesome to have a learningcurve like this again.
unknown (31:07):
Right?
SPEAKER_00 (31:07):
Yeah.
To like be like, oh my God, thiswhole new thing is new.
That's part of what I enjoyedabout the founder CEO journey is
the heterogeneity, the diversityof what you needed to focus on
regularly.
So you're constantly learningnew things.
Now the whole world has alearning curve like this.
That's cool.
When was the last time we hadthat?
In school.
Like it's awesome.
Or when we tried a new position,or it doesn't happen all the
(31:29):
time.
This one's gonna last for years.
SPEAKER_02 (31:33):
I think the last one was the internet.
Do you think there was adifferent one?
SPEAKER_00 (31:37):
I I don't believe there's ever been one at this
speed.
I think mobile was a huge one,social media was a huge one.
The none of them went this fast.
SPEAKER_02 (31:45):
No, none.
It's amazing.
SPEAKER_00 (31:50):
Exactly to that point.
15, 20 years ago, SEO was anissue.
Every CMO and every practitionerwould want to talk about it.
Today, it's mature.
What CMOs will generally want totalk about it?
Those who are in trafficarbitrage businesses or those
for whom that channel is notperforming.
AEO is now where SEO was before,where every CMO I talk to and
(32:14):
every practitioner I talk to, Iwonder if you have the same
experience, absolutely wants totalk about this because we've
all got this learning curveagain.
It's fun.
SPEAKER_02 (32:20):
Oh, 100%.
It's the conversation of everysingle room.
And that's the beautiful thing,is it's the best conversation
point.
I wonder when they'll hit thatthreshold of people not wanting
to talk about AI as much, butright now it is just absolutely
captivating any conversation inthe room.
And you yourself, you talkedabout how you learn about AI.
What about overall?
Have there been any impactfulbooks on your career?
SPEAKER_00 (32:47):
I particularly liked a book a long time ago called Uh
Getting Real or Get Real.
It's by 37 Signals.
Do you know it?
SPEAKER_02 (32:55):
I don't.
Yeah.
SPEAKER_00 (32:56):
It is about building products, but I think it's about
how to run teams.
And it is a lot about rapiditeration, quick learning, lots
of communication on small teams.
And so the team that createdBackpack and a few other
products created this.
And it's not like a long book.
It's a bunch of a paragraph on apage, and then the next thought,
(33:18):
the next thought, the nextthought.
And I thought it was veryactionable.
I also happen to be a fan ofAtomic Habits.
A classic class.
SPEAKER_02 (33:25):
So, so good.
SPEAKER_00 (33:26):
Yes.
SPEAKER_02 (33:26):
Yeah, I love it.
You cannot go wrong with thatbook.
James Clare just did anincredible, incredible job.
Absolutely incredible job.
SPEAKER_00 (33:31):
Yeah.
And comparing us with peers,like what we do at your annual
general meeting.
Yes.
It's great to simply talk withfolks dealing with similar
issues.
SPEAKER_02 (33:41):
There's no better learning than just that in-depth
conversation.
And that's actually part of whatwe're trying to replicate here
with the podcast is how do youdemocratize access to a lot of
the insight and kind of the bestminds in tech and the
conversations that go on and theknowledge.
So appreciate you sharing that.
And you've had some some wildstories lately.
I know you were just at F1 andyou had, you know, a phone
(34:02):
incident.
I'd love to hear a little bitmore about that.
SPEAKER_00 (34:07):
I fell way down the F1 rabbit hole five, six years
ago as a former aerospaceengineer, because you know,
aerospace engineer to CMO islike this super popular career
for linear super linear.
SPEAKER_02 (34:18):
Yeah.
SPEAKER_00 (34:18):
Um was at a race this weekend.
And at the end of the race, theygenerously let you uh go around
the track in this flatbed truck,and you're like, this is holy
ground.
Oh my god, they were just racingon this.
And the guy next to me, italmost at the end was like this.
And I was holding my phonetightly, it went flying onto the
ground.
And an hour and change later,with lots of help from many
(34:39):
great people, my wife in the USincluded, doing find my phone
and all of that.
Thank you, Monica.
Uh the phone looked demolisheduh and took it to a repair shop.
They said just the screen isokay.
And then I got to watch thevideo the 17 minutes after it
got hit the ground.
It got run over five times.
unknown (34:59):
Oh wow.
SPEAKER_00 (35:01):
Here it is.
This is two days later, and it'sjust fine with the new screen.
Like very impressive.
SPEAKER_02 (35:06):
We might need an Apple ad here.
This rate.
Yeah.
Incredible.
You can see where it was runover right here.
SPEAKER_00 (35:13):
Yes.
SPEAKER_02 (35:13):
But otherwise, you never know.
SPEAKER_00 (35:15):
Looked like the surface of the moon.
It was all like craterseverywhere, but new screen, and
the rest of it was fine.
And I got the videos from theday off of it.
SPEAKER_02 (35:22):
That's hilarious.
What a memory.
SPEAKER_00 (35:24):
It was a great memory and war story.
SPEAKER_02 (35:26):
Yeah, I'm sure.
Well, I'm glad.
AI is advancing at such a rapid rate.
SPEAKER_00 (00:02):
Do you mean with LLMs crawling the web almost
every query, you can have animpact this afternoon.
You can use it to outsource yourthinking.
You can use it to think moredeeply.
And door number two seems to beranking a whole lot better.
SPEAKER_01 (00:17):
And what else do people need to really think
about at the macro scale of AEO?
SPEAKER_00 (00:21):
Somebody was like, hey, if this Fortune 500 CMO
fired his or her entire SEOteam, we're doing a huge
mistake.
Your SEO resources are your AEOresources.
Your SEO agency is your AEOagency because this is an
evolution.
SPEAKER_02 (01:07):
That's where this conversation begins.
Because today, answer enginesfrom ChatGPD to perplexity and
more are reshaping how buyersfind products.
And for founders and operators,this is one of the biggest
arbitrage opportunities sincethe early days of SEO.
Gaia Leaf, chief evangelist atWebflow, has seen this movie
before.
And because he has, he's takingaction and has the data to show.
SPEAKER_00 (01:29):
In two weeks, half of the incremental citations
Webflow got were from those sixpages.
And organic traffic to thosepages is up 24%.
SPEAKER_02 (01:37):
In this conversation, you'll learn
Webflow's four-part AEOframework.
Why answering questions, notkeywords, is the new growth
edge, and how early adopters canturn this uncertainty into
leverage.
SPEAKER_00 (01:48):
Or, as Guy puts it, the people that went earlier,
they found relative bargains.
SPEAKER_02 (01:53):
In other words, the kind of upside we haven't seen
since the early days of search,SEM, and mobile.
All right, let's get into it.
Guy, welcome to the podcast.
SPEAKER_00 (02:02):
Thanks, Sophie.
Thanks for having me.
Excited to talk with you again.
SPEAKER_02 (02:05):
It is great to have you here.
And nobody better to learn aboutAEO from right now at this
inflection point.
And you've lived through the bigSEO boom.
You've felt it.
What does this inflection pointfeel like compared to that?
SPEAKER_00 (02:22):
They're very similar, actually.
Um like the early days ofsearch, there are a bunch of
different players.
Things are changing all thetime, like they did back then.
I mean, it makes the news nowthat Google updated their
algorithm because it happens afew times a year, because it's a
settled space.
And here, you know, things arechanging very rapidly.
(02:44):
The rules of what matter areupdating frequently enough that
everybody feels behind, andthere's real opportunity.
Like any new medium, those thattake action early, they'll find
bargains, basically, effectivebargains, you know, return
relative to the effort they putin.
SPEAKER_02 (03:01):
Mm-hmm.
That's really interesting.
And if you were explaining thiskind of new world, even though
it feels a little bit moresimilar, but to a founder, what
are the key differences?
SPEAKER_00 (03:11):
I think there are some that are saying, look, SEO
is dead, and there's this awhole new world, respectfully.
I 180 degrees, black and whitecould not disagree more.
I think AEO is an evolution ofwhat good SEO always should have
been.
You know, genuinely valuableoriginal content.
It's getting rewarded.
(03:32):
The the LLMs, they get to seemore context about what's good.
When we all learn to speakGoogle, our average query was
four words long.
The average query in an LLM is23 words long.
So they get more context, theyhave a conversation.
So if I were explaining it to afounder, I would say, look, take
SEO, minus one thing plus acouple.
The thing that's minus is theconcept of ranking for keywords
(03:55):
doesn't exist anymore.
Because you got paragraphs now.
What's the plus?
Plus is topics used to be abasket of keywords, a cluster of
keywords, and you tried to makesure you had them in your
content.
Now a topic is a group ofquestions that your prospects
may be asking in the funnelclustered together.
(04:15):
So shift from counting keywordsto answering questions, that'd
be one.
Two, help the LLM understand thestructure and meaning of your
content.
We've been doing that with SEOforever, right?
We have the metadata.
There's a bit more.
Do this thing called schema.
And the last would be we focusedon backlinks for years.
They still matter.
They don't not matter, butincrementally more important
(04:38):
than an SEO is repeated mentionsin plain text on other sites
saying, oh, GTM Now is awesome.
Oh, this is one of the toppodcasts that matters more than
it did with search.
SPEAKER_02 (04:49):
Mm-hmm.
Mm-hmm.
The pendulum is almost shiftingback to that third party.
It's bringing PR and all thesedifferent things back that, you
know, had had the moment of notnot mattering less by any means,
but maybe less emphasis on them.
And now are getting highlightedmore, which is really
interesting.
SPEAKER_00 (05:04):
VR's back, brand is back.
Those things matter more thanthey did before, yes.
SPEAKER_02 (05:09):
Yeah.
And those things matter.
So we know definitively what'simportant.
But what what about overall justknowing?
Like, how do you know?
Does anyone actually know what'sworking with AEO?
Or are these best guesses?
How are we thinking about asingle source of truth around
AEO?
SPEAKER_00 (05:28):
So you said definitively no, and I'm glad
you asked the question becauseanyone who says they
definitively know is lying.
Like no one knows.
But in the same vein that no oneknows for SEO, right?
We all we all try a bunch ofthings.
We observe the results of anexperiment.
Oh, I added more keywords, fewerkeywords, longer metadata,
shorter metadata.
And they saw what works.
(05:49):
But no one actually knows what'shappening in the Google
algorithm.
Now, with the LLMs, not even theLLM providers themselves know
what's actually happening inthere.
So when we talk about theserules of thumb, they are very
well educated, very wellresearched, data-driven guesses.
They're the current bestthinking uh about what one
(06:10):
should do to go manage the riskand opportunity of LLNs, which
is really to take back controlof our brand narrative and drive
more traffic to our websites togo generate revenue.
SPEAKER_02 (06:22):
Absolutely.
And AI is advancing at such arapid rate.
Yeah, I never heard of it.
What what do you think thatchanges about the SEO adoption
to AEO?
So when you say they're alleducated best guesses, are those
guesses changing at a pace thatis faster than SEO?
(06:43):
Or would you say they'reevolving at the same pace as
they did?
SPEAKER_00 (06:46):
My humble opinion, it's faster, exactly for the
reason you described.
The LLMs are improving theirability to understand meaning
and structure and content, whaton like a monthly basis every
couple of months?
A hat tip to Google and Yahooand Bing and everyone, but like
it didn't go that fast.
I don't recall from back in theday that it did.
Um mercifully, we have skippedover a whole generation of spam
(07:11):
where if you look at the stuffthat ranks top in Google, 85% of
it is either entirely humangenerated, that's about 65%, and
another 20% is like a little bitof AI.
So the notion that I can createenough infinite content with an
LLM and I'm just gonna go spamout zillions of pages, it's not
working.
Thankfully.
SPEAKER_02 (07:32):
Thankfully.
Less noise.
Yes.
Less noise is always good.
And when you're thinking aboutoverall frameworks or advice to
people that are leaning intoAEO, do you have any kind of
recommendations?
I know at Webflow, you're doinga ton of work in the space and
know those best practices andhave implemented them yourself.
We'd love to hear a little bitmore around that work.
(07:53):
And I guess your best educatedguess, not necessarily what is
technically working, but whatare your best educated guesses
right now?
SPEAKER_00 (08:00):
Um Vivian, who leads our SEO and AEO team, has done
some amazing work.
And uh, we went and looked forexamples, case studies, spent
lots of time, many hourslooking.
There aren't any.
My best guess about why is thatpeople are doing the work.
They aren't yet, it's not yetrepeatable enough that they're
sharing it externally.
(08:21):
So we're sharing a lot of whatwe're doing externally.
All the research we've done,talking with other experts in
the space, looking at the data,humbly, we would suggest
thinking about the transitionfrom SEO to AEO in four
categories.
The first is content, where yougo from counting keywords to
answering clusters of questionsto making content super relevant
(08:45):
with personalization.
So this is all about like thecontent you own.
The second is technical, whereyou go from on-page SEO to
helping LLMs understand thestructure and meaning of your
content by giving them moremetadata.
Like we all did it withFacebook's Open Graph.
SPEAKER_03 (09:01):
Yeah.
SPEAKER_00 (09:02):
Now there's this thing that many know about, but
uh many also don't call itschema, that is just another
form of metadata that Google,Bing, Yahoo, Yandex, and others
came up with.
And it's like, oh, here's howyou specify an events page, a
product page, a bio page.
88% of sites don't have schema,but 73% of Google's top results
(09:23):
do.
So, like, tons of opportunitythere.
Coming back to technical, andyou want that on a super fast
site global.
Third area is authority, wherewe went from backlinks that
still matter, they don't notmatter, to repeated positive,
plain text mentions about yourbrand pointing to a visually
(09:44):
stunning, emotionally evocativeexperience because that's a
signal about authority too.
And the last is measurement,where we go from keyword
ranking, which no longer exists,to did I show up in the
questions I care about?
To what's my share of voicerelative to my competitors and
what's the sentiment?
Because over here in search, theGoogle is using our words.
(10:06):
Here the LLMs are reformulatingeverything.
So the notion of sentiment beingimportant exists.
So those four buckets uhcontent, technical, authority,
and measurement.
SPEAKER_02 (10:16):
Incredible.
And from implementing those, yousaid the team's been doing a
great job.
What kind of results have youseen?
SPEAKER_00 (10:22):
Vivian and team have done a bunch.
They partner with our SEO andAEO firm Graphite, and a couple
of things come to mind.
Um, she recently took our tophalf dozen product pages and
tried to put some structure onthem so the LLMs could
understand the meaning andstructure of them.
What did she do?
She put FAQs at the bottom ofthe page, so about half a dozen
(10:44):
FAQs.
She then put schema, thatmetadata, there in line with
those FAQs.
In two weeks, half of theincremental citations Webflow
got were from those six pages.
Six pages out of thousands thatwe have.
Yeah.
And organic traffic to thosepages is up 24% in two weeks.
So that was one thing she did,and it's great, and that's the
(11:06):
beginning.
We're going to do more.
The second is that freshnessmatters.
So the content machine is ashungry as ever for good original
content.
Um 85, maybe 95% of the contentGB chat GPT sites was updated in
the last 10 months.
So she worked with anotherpartner of ours, AirOps, to
(11:27):
increase the speed with which werefreshed content.
She did that and uh found 20%,20, 40, 40% more traffic coming
from AI search because we wereupdating the content more
regularly.
Uh, and and we see that contentwith freshness indicator, like
(11:49):
last updated date, gets 1.8xmore citations.
So giving structure and updatingmore regularly.
The two experiments that haveworked really well for us.
SPEAKER_02 (11:58):
Wow.
That's a huge difference justfrom updating content.
Yeah.
What and maybe this is gettinginto the tactics.
I need to give her a call, butwhat constitutes updating?
How much work would you say?
There's there's a lot ofcompanies essentially that have
a ton of content, but it's notnecessarily optimized for AI
engines and they want to refreshthem.
(12:19):
How much work is it to refreshexisting content?
SPEAKER_00 (12:23):
The direct answer is I don't know.
SPEAKER_02 (12:25):
Yeah.
SPEAKER_00 (12:26):
Because I didn't, I have Vivian and Devontae.
Totally.
I do believe first, it's worthrunning the experiment.
Somebody said, hey, wait aminute.
If you need fresh content, whatif you just change the last
updated?
Is that enough?
I hope it's not.
SPEAKER_02 (12:40):
It's not.
At least from the SEO side, Iknow it's not.
I I don't know if AI enginesdiffer, but from the SEO side,
you need something in the mainbody.
You need like a little bit of amain point and the date.
So I would imagine it's thesame.
But I'm I don't know.
I gotta call Vivian.
SPEAKER_00 (12:56):
One, I would bet without data.
All of those stuff was datadriven.
This is hypothesis.
And the other hypothesis wouldbe that um AirOps, the tool we
partner with, makes it easy todo that in a workflow and with
LLMs, but that can't be it.
Right.
If you just pump out an LLMupdated page, you need human
(13:16):
review, you need humaneditorial.
I've got friends who startedcompanies, maybe GTM Fund
invested in one, where it'slike, I can detect if this was
entirely LLM created usingstatistical pattern matching on
the content.
And so uh we touch with thehuman every piece of content in
a material way before it goesout the door.
SPEAKER_01 (13:36):
Interesting.
SPEAKER_00 (13:37):
So both hypotheses rather than data driven, but
you've got the SEO data there.
SPEAKER_02 (13:42):
Yeah.
And it sounds like the commontheme of you can't rely on AI to
get from zero to a hundred.
It'll take you part of the waythere, but you do need that
process for actuallyimplementing the human touch.
SPEAKER_00 (13:54):
Totally.
And I feel like one of the oneof the underspoken but
foundational things in LLMs aslike kindergartners through to
CGOs learn about them and howthey change everything, is you
can use it to outsource yourthinking.
You can use it to think moredeeply.
And door number two seems to beranking a whole lot better with
(14:18):
answer engines than door numberone.
And by the way, it helps makeall of us more employable.
So I feel like it's something wearen't talking about enough, but
it's it looks the same on thesurface you used in LLM.
Yeah.
But what's going on inside yourhead could be completely
different.
SPEAKER_02 (14:33):
And it's tempting to go the quick route, but always
play the long game and actuallyinvest in the depth.
A quick pause to tell you abouta company you need to know.
ZoomInfo is the go-to-marketintelligence platform built for
sales, marketing, and reps.
By unifying data, workflows, andinsights into a single system,
ZoomInfo helps revenue teamsfind and engage the right
buyers, launch go-to-marketplacefaster, and drive predictable
(14:55):
growth.
With industry leading accuracyand depth of data, it gives your
team the intelligence advantageto win in competitive markets.
It's trusted by the fastestgrowing companies and has become
the category leader ingo-to-market intelligence.
Learn more at zoominfo.com.
You mentioned two points thatyou did that had a huge impact,
and one of them was updating andrefreshing pages for, let's say,
(15:18):
founders or early stagecompanies that are really
looking in to implement AEOtactics, but there's a lot out
there.
You listed off a lot ofdifferent really interesting
points that people could diveinto.
What are the two big ones thatyou'd recommend?
Are those the two, or would yourecommend two others?
SPEAKER_00 (15:35):
If you only did two things, I would suggest you
shift, you answer questions.
So shift from keywords toquestions and answer them
comprehensively.
And the second is add schemametadata to your site so that
you can help the LLMs understandthe meaning of your site more
effectively.
Those are the easily the top twothings I would suggest doing.
They're the right place tostart.
SPEAKER_02 (15:56):
Keywords, metadata, answers.
Brilliant.
Yes.
And answers.
Sorry.
No, no.
This is how much has been grayedin my brain.
SPEAKER_00 (16:05):
You too have many years in search.
SPEAKER_02 (16:07):
I I do, yeah.
It's a bit of a rewiring.
So it is a process.
Okay.
SPEAKER_00 (16:12):
Super helpful.
But your point about therewiring, it's not like throw it
all out.
It's adjustments.
It's an evolution rather than awholesale revolution.
SPEAKER_01 (16:21):
Mm-hmm.
Mm-hmm.
Okay.
And what else do people need toreally think about at the macro
scale of AEO?
SPEAKER_00 (16:29):
I heard, I hope it's not true.
It's not validated at all, butat one of these conferences,
somebody was like, hey, if thisFortune 500 CMO fired his or her
entire SEO team and is like,we're doing AEO.
I think it's a huge mistake.
Your SEO resources are your AEOresources.
Your SEO agency is your AEOagency because this is an
evolution.
So SEO is dead and this is acomplete revolution, I think, is
(16:52):
one of the easiest ones to seesome great articles about.
And uh I I I I don't think thedata supports it.
Um and the second is just likewe did with SEO, you can start
small and grow.
In the example we talked aboutwith Webflow, Vivian chose half
a dozen pages, saw real results.
Obviously, she's then gonna gocontinue, but she didn't do
(17:14):
everything all at once day one.
Mm-hmm.
SPEAKER_02 (17:17):
Mm-hmm.
No, that makes sense.
Okay.
Super helpful advice.
I love it.
It'll be interesting to see howit plays out now over time.
SPEAKER_00 (17:26):
To your point, we are all learning together,
right?
This this mental model we haveof these four categories.
We also made up a rubric of fivedifferent levels of them, but
it's idle keeper justice, as weall know.
SPEAKER_02 (17:39):
And how do you learn?
How do you learn about AEO andthen AI overall?
SPEAKER_00 (17:47):
A bunch of time, really.
It's it is it is it is a bunchof time experimenting, focusing
on data, because it's it's thereare uh a lot of things being
said.
Uh and when like I'll read anarticle about something and I'll
dig into the source for that andthe source and the source and
the source.
And sometimes I'm like, wait,that article, it contradicts the
(18:09):
original underlying data, but itwas like a telephone game of
referencing each other uh andthen staying close to the work
we're doing, experimentingdirectly uh with both.
I happen to like two podcasts.
One is this one, which I listento religiously, and the other
one is um at Marketing AIInstitute's uh uh podcast, which
(18:34):
uh Paul and Mike, they producegreat content too.
SPEAKER_02 (18:37):
Okay.
I love it.
I think that's where at with thefast evolution people are
learning the most on the fly andour individuals, our networks
like podcasts and channels andso forth, where it's just a
little bit more live and andflexible.
I know that's where personally Ilearned the most too is podcasts
from people individually, howthey're learning.
We're seeing a lot more contentout there that is dynamic,
(18:59):
that's actually tactical,actionable advice on AEI.
And I know you're producing alot of content around that too,
actually.
So for for anyone kind ofcurious about their own AEO
strategy, what kind of resourcesare out there?
Because I know that you are areworking on a ton.
SPEAKER_00 (19:17):
We wouldn't I would invite people to go to
webflow.com slash resourcesslash AEO.
There, you will not find apitchful webflow.
What you will find is a place togo enter your domain name, hit
enter, a couple of minuteslater, you will get an email
saying, we ran in the backgrounda 600-line prompt.
(19:37):
It's about to become 1700 in V2that goes and looks at your
site, goes and looks off yoursite, and suggests, hey, here's
what we see in those fourcategories: content, tech,
authority, and measurement.
Here are the gaps we see.
Here are suggestions for whatyou might do to level up.
And it bottom lines it, like youwere doing earlier.
If you only did two things, godo these.
(19:58):
And then below that is detailaround these four categories and
the things we're seeing that canhelp people evolve.
SPEAKER_02 (20:07):
Amazing.
And no cost?
No cost.
No cost.
Okay.
Well, then we'll drop it in theshow notes.
That's amazing.
Yeah, what a great resource,especially after you just broke
down those four points andthey're super, super important
for people optimizing.
And what about the people thataren't optimizing?
What do you think happens ifpeople aren't actually following
an AEO strategy right now andinto the future?
SPEAKER_00 (20:28):
In my humble opinion, is it's it's missed
opportunity.
I think AEO is truly threat andopportunity for us.
Over time, we will lose controlof our brand narrative because
our carefully differentiated andcrafted words are being
reformulated.
We'll lose traffic.
Bain found 15 to 25% decrease intraffic, which I think is closer
(20:49):
to reality than the headlinegrabbing, oh, Spot lost 80% of
their traffic.
I don't think that's the normand was cherry-picked for them
as an example.
And the third is we all soundthe same.
We're all using the same LLM.
So how many people are, youknow, the leader, the best, the
first, the fastest?
Uh and I think um uh actuallyI've lost the train of thought
(21:10):
entirely.
Where was that?
SPEAKER_02 (21:11):
Okay.
Um what happens if people Oharen't doing it.
SPEAKER_00 (21:15):
Thank you.
SPEAKER_02 (21:15):
Don't yeah.
SPEAKER_00 (21:16):
So then so the risk is we all sound the same.
If we then go invest some inAEO, which is with your existing
resources, same hours, samepeople, shifting strategy a
little bit.
You will be earlier to market,and chances are you will find
opportunity.
Just like we did in the earlydays of search, in the early
days of SCM, in the early daysof mobile, in the early days of
(21:38):
native, the people that wentearlier, they found relative
bargains.
They got more return relative tothe amount of effort they were
putting in before everybody wasthere.
SPEAKER_02 (21:48):
Early adopters, advantage.
Yes.
And is there a certainpercentage that you'd recommend
teams think about ofreallocating resources from SEO
to AEO under the same team?
SPEAKER_00 (21:59):
I think it depends on how where they are seeing
traffic coming from and how faralong they are in their SEO
journey.
So I I I do suggest teams golook at traffic to their site.
At Webflow, 8% of our self-servesignups come from Answer engines
today, AI search.
Uh by the way, it convertsbetter too, because they're
further down the funnel.
It converts six times, not sixpercent, six times better than
(22:22):
non-branded organic, which iseye-opening.
But so I suggest those to youlook at how much traffic you're
getting from Answer engines.
Look if it's increasing.
For us, in unbranded, it wentfrom zero a year ago to 42%
today.
unknown (22:38):
Okay.
SPEAKER_03 (22:38):
Yeah.
SPEAKER_00 (22:39):
It's it's like it's and so that's worth investing
time in.
That's a clear signal to goinvest in AEO strategies.
Um, that having been said, 25%of the top pages out there don't
even have basic SEO metadata.
They have broken links.
That's also worth tackling ifyou don't have those
fundamentals in place.
SPEAKER_02 (22:57):
A very good call out.
And perhaps they can be done intandem, but maybe we're skipping
steps by jumping ahead whenreally there's a lot of SEO
leverage to be had too.
But 6X, that's a wild statistic.
Like talk to me about what youmean around that traffic is
converting 6x better.
SPEAKER_00 (23:16):
I that the traffic that comes from answer engines,
you look at touch to purchase,because here we're talking about
self-search sign-up.
Yeah.
That is 24x 24% conversion rate.
Non-branded organic search is 4%conversion rate for that metric.
Two of the prominent SEO firms,HREFs and SEMrush, found in
(23:39):
their own data 4x and 23x.
So our 6x is in the range.
SPEAKER_01 (23:44):
That's really interesting.
SPEAKER_00 (23:46):
It's uh and why we think because folks are further
down the funnel.
So yeah, less traffic, but it'smore qualified because they have
self-qualified.
SPEAKER_02 (23:56):
So what do you think behaviorally they're doing?
Because I'm thinking to youknow, my procurement processes
and yours and the way everybodyindividually operates.
Are they researching on theirown and then turning to LLMs?
Are we not starting with LLMs,or do you think people are
starting with LLMs?
SPEAKER_00 (24:14):
Forrester surveyed B2B marketers.
95% of them said LLMs are goingto be part of my buying process
this year.
SPEAKER_03 (24:21):
Yeah.
SPEAKER_00 (24:21):
Answer engines will be.
So it's clearly something thatis important.
Did they literally start there?
Did they start with search?
Did they start with friends?
I don't know.
Anecdotally, I think word ofmouth is easily continues to be
the most trusted.
That's quantitative.
It is the most trusted form ofrecommendation.
By the way, your website'snumber two.
That was eye-opening.
SPEAKER_02 (24:42):
Um uh and so your website, so your website isn't
dead.
SPEAKER_00 (24:50):
Your website's not dead at all.
In fact, I firmly believe thatpeople are going to want to hear
your story told your wayforever.
That is not going away.
To your point about how we'reengaging with LLMs, often the
first look is there, and peopleare learning.
Well, what matters in thiscategory?
Who are the players?
Some of the tofu mofu thingsthat we've all done, but then
(25:11):
they are looking to see yourstory told your way.
By the way, that website needsto then speak to humans in
visually stunning, emotionallyinvocative, engaging ways, and
to machines.
And folks are talking about thislike it's new, but it's not.
We've had search metadataforever.
Now we have a bit more metadatafor the LLMs.
And so we've been doing this fora while.
(25:33):
I think folks are, this isobviously more in the zeitgeist
now.
You need websites that speak toboth.
SPEAKER_02 (25:39):
And you've obviously built in this space and solved
in this space before.
You know it so intimately.
One of the quotes that I justabsolutely love, and I think
about all the time around thework that you've done, the work
that I've done in the past, thework that a lot of people are
doing right now is people imputethe quality of your product or
service from the quality of yourwebsite.
So even if there are thirdparties like LLMs, it is your
(26:02):
story.
You're still owning your storyand design and visually
appealing sites and emotionallyinvocative, like you said, are
huge, huge part of that.
SPEAKER_00 (26:11):
I am very, very much with you, both quantitatively
and just subjectively as anindividual.
Uh and I think the LLMs aregoing to use it as a signal of
quality.
That's belief.
Why do I believe that?
Because Google looks at bouncerate in search, right?
Did somebody leave your site andcome back immediately?
It would be surprising if the uhanswer engines weren't doing
(26:34):
that.
And you see perplexity as abrowser, ChatGPT is rumored to
be creating one.
Why might they do that?
I I believe two reasons.
One, to not pay the gajilliondollars, Google's paying Apple
to be the default and Safari,fine.
Perhaps more valuably, they getto see where you're going.
They get to see what's actuallyvaluable content.
(26:54):
And so you having a compellinghuman-focused story on your
website will continue to mattera lot.
It may actually get you toappear in the answer engines
more often because they can tellit's valuable content.
SPEAKER_02 (27:08):
So the quality metric matters a lot more now,
it sounds like it is definitelya signal.
Yeah.
Someone told me a very tacticalnugget around AEO.
And it was that just as yousaid, your website matters so
much.
Yes.
So that pairing it with theformats that are performing well
is really impactful.
So publishing a listicle fromyour own website of the top
(27:31):
companies in your area actuallyperforms really well.
So for a venture firm, forexample, that would be, you
know, top early stage venturefirms and listing out the top
15.
And you could obviously positionyourself first or wherever you
want on the list.
It feels a bit funny, but theywere saying that this has been
uh been a huge incrementaltactic and one of the many check
(27:53):
boxes on AEO.
But it feels a little funny.
I know we personally haven'tdone it, but there's a lot of uh
a lot of interesting thingsaround that.
SPEAKER_00 (28:01):
Without a doubt, it feels weird to name your
competitors and compliment themon your site.
SPEAKER_02 (28:05):
Or just have yourself listed first on your
own listicle, even that alsofeels weird.
SPEAKER_00 (28:10):
And I think the underlying principle is that in
the content, separate from themetadata we've been talking
about, you can also signalstructure to LLMs, a listical
bullets to use like in summary,to put FAQs at the bottom, to
use one H1 on the page, and thenlogically structured H2, H3s,
(28:30):
H4s.
I can tell you how many sitesyou're like, H1, then H3, then
H2, that and you're like, holdon.
There, they're you're you'resignaling structure to have
accessibility info.
They use that, they use thatactively.
And so Listical is a great one.
The other underlying principlethat we see is in search, it
took a while to have an impactbecause you needed to create
(28:53):
authority, and that was based onbacklinks, and that took time.
With LLMs crawling the webalmost every query, you can have
an impact this afternoon.
You can go see the impact somuch more quickly, and it's even
multiplied when you and yourcompetitors all do that
listicle, which feels superweird.
But you can collectively go dothat and have an impact this
(29:15):
afternoon.
By the way, the flip side, thepainful side of that coin is you
change your positioning, you'vechanged it on your website
today.
It may or may not have an impactbecause lots of other people
still have your old positioningthere.
And that can have an impact.
Or I'm sure none of thelisteners here have this
problem, but if you have someoutdated content on your site,
(29:37):
again, nobody listening to thishas any of that.
SPEAKER_02 (29:38):
Of course not.
SPEAKER_00 (29:39):
But if they did, that also will have an impact.
And so the notion of taking downcontent, which is anathema in
search, is a thing in AEO.
SPEAKER_02 (29:48):
Mm-hmm.
Fascinating.
Well, guy, thank you for theexpertise around AEO.
This has been fantastic.
I think you said it well aroundnobody really has the answers,
but you yourself are somebody.
That's been in search for 17years.
You're at one of the companiesleading in the space.
So we appreciate your thoughtleadership on it and better
informing the insider scoop.
SPEAKER_00 (30:11):
Thanks for the chance to talk about it.
And I'm curious.
It was great to learn from youhere.
And I'm eager to continue theconversation and learn from the
listeners as we all rundifferent experiments and learn
in this rapidly evolving space.
SPEAKER_02 (30:21):
Yeah.
I think that you just hit on areally interesting point there,
too, is it almost resets theplaying field.
AI resets the playing fieldwhere suddenly everyone is
learning something at the sametime.
And so it really democratizesexperience in a way.
How does that feel as somebodythat has, you know, built and
(30:45):
successfully sold a company andoperated at the highest level?
Like, what does that feel liketo suddenly have something that
is so new?
SPEAKER_00 (30:56):
Joyful.
It's awesome to have a learningcurve like this again.
unknown (31:07):
Right?
SPEAKER_00 (31:07):
Yeah.
To like be like, oh my God, thiswhole new thing is new.
That's part of what I enjoyedabout the founder CEO journey is
the heterogeneity, the diversityof what you needed to focus on
regularly.
So you're constantly learningnew things.
Now the whole world has alearning curve like this.
That's cool.
When was the last time we hadthat?
In school.
Like it's awesome.
Or when we tried a new position,or it doesn't happen all the
(31:29):
time.
This one's gonna last for years.
SPEAKER_02 (31:33):
I think the last one was the internet.
Do you think there was adifferent one?
SPEAKER_00 (31:37):
I I don't believe there's ever been one at this
speed.
I think mobile was a huge one,social media was a huge one.
The none of them went this fast.
SPEAKER_02 (31:45):
No, none.
It's amazing.
SPEAKER_00 (31:50):
Exactly to that point.
15, 20 years ago, SEO was anissue.
Every CMO and every practitionerwould want to talk about it.
Today, it's mature.
What CMOs will generally want totalk about it?
Those who are in trafficarbitrage businesses or those
for whom that channel is notperforming.
AEO is now where SEO was before,where every CMO I talk to and
(32:14):
every practitioner I talk to, Iwonder if you have the same
experience, absolutely wants totalk about this because we've
all got this learning curveagain.
It's fun.
SPEAKER_02 (32:20):
Oh, 100%.
It's the conversation of everysingle room.
And that's the beautiful thing,is it's the best conversation
point.
I wonder when they'll hit thatthreshold of people not wanting
to talk about AI as much, butright now it is just absolutely
captivating any conversation inthe room.
And you yourself, you talkedabout how you learn about AI.
What about overall?
Have there been any impactfulbooks on your career?
SPEAKER_00 (32:47):
I particularly liked a book a long time ago called Uh
Getting Real or Get Real.
It's by 37 Signals.
Do you know it?
SPEAKER_02 (32:55):
I don't.
Yeah.
SPEAKER_00 (32:56):
It is about building products, but I think it's about
how to run teams.
And it is a lot about rapiditeration, quick learning, lots
of communication on small teams.
And so the team that createdBackpack and a few other
products created this.
And it's not like a long book.
It's a bunch of a paragraph on apage, and then the next thought,
(33:18):
the next thought, the nextthought.
And I thought it was veryactionable.
I also happen to be a fan ofAtomic Habits.
A classic class.
SPEAKER_02 (33:25):
So, so good.
SPEAKER_00 (33:26):
Yes.
SPEAKER_02 (33:26):
Yeah, I love it.
You cannot go wrong with thatbook.
James Clare just did anincredible, incredible job.
Absolutely incredible job.
SPEAKER_00 (33:31):
Yeah.
And comparing us with peers,like what we do at your annual
general meeting.
Yes.
It's great to simply talk withfolks dealing with similar
issues.
SPEAKER_02 (33:41):
There's no better learning than just that in-depth
conversation.
And that's actually part of whatwe're trying to replicate here
with the podcast is how do youdemocratize access to a lot of
the insight and kind of the bestminds in tech and the
conversations that go on and theknowledge.
So appreciate you sharing that.
And you've had some some wildstories lately.
I know you were just at F1 andyou had, you know, a phone
(34:02):
incident.
I'd love to hear a little bitmore about that.
SPEAKER_00 (34:07):
I fell way down the F1 rabbit hole five, six years
ago as a former aerospaceengineer, because you know,
aerospace engineer to CMO islike this super popular career
for linear super linear.
SPEAKER_02 (34:18):
Yeah.
SPEAKER_00 (34:18):
Um was at a race this weekend.
And at the end of the race, theygenerously let you uh go around
the track in this flatbed truck,and you're like, this is holy
ground.
Oh my god, they were just racingon this.
And the guy next to me, italmost at the end was like this.
And I was holding my phonetightly, it went flying onto the
ground.
And an hour and change later,with lots of help from many
(34:39):
great people, my wife in the USincluded, doing find my phone
and all of that.
Thank you, Monica.
Uh the phone looked demolisheduh and took it to a repair shop.
They said just the screen isokay.
And then I got to watch thevideo the 17 minutes after it
got hit the ground.
It got run over five times.
unknown (34:59):
Oh wow.
SPEAKER_00 (35:01):
Here it is.
This is two days later, and it'sjust fine with the new screen.
Like very impressive.
SPEAKER_02 (35:06):
We might need an Apple ad here.
This rate.
Yeah.
Incredible.
You can see where it was runover right here.
SPEAKER_00 (35:13):
Yes.
SPEAKER_02 (35:13):
But otherwise, you never know.
SPEAKER_00 (35:15):
Looked like the surface of the moon.
It was all like craterseverywhere, but new screen, and
the rest of it was fine.
And I got the videos from theday off of it.
SPEAKER_02 (35:22):
That's hilarious.
What a memory.
SPEAKER_00 (35:24):
It was a great memory and war story.
SPEAKER_02 (35:26):
Yeah, I'm sure.
Well, I'm glad.
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.
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
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!