AEO Fundamentals: Moving to Content for Machines/Rishi Mallik, Eric Nalbone - SEO, Author of Truth, FAQs ~ Spark of Ages Ep 46 - Spark of Ages

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Rajiv Parikh (00:00):
We talked about this notion of AIO, aeo, seo,
geo.
I've even heard it called LLMIO.
So many different ways ofsaying this topic.
So to keyword stuff, thisepisode.
I've said every one and I'llprobably keep using it, that way
.

Rishi Mallik (00:14):
I think that GEO is a repackaged term to make it
easier for people to understandwhat it actually is, but the
game has completely changed.

Eric Nalbone (00:23):
Our biggest perspective is that productive
participation in this AIrevolution or geo-optimization
craze is what we're after.

Stuff is changing every day, and so we have some
general tactics that have beenworking.
But it's interesting just howquickly it's moving and how
quickly even one tactic mightchange to another.

Again, it's one of those things where you can
frame it as a challenge, whereyou know traffic is down, or you
can frame it as an opportunity.
If revenue is up but traffic isdown, isn't that a win?

for everyone.
And that day when I called youwas my second council that month
where it was red alert for folkLike they were seeing 30%
declines in their actualclick-through traffic, and so to
maintain that humanauthenticity in many ways with
an AI model can be somewhateasier than trying to maintain
authenticity for an SEO page.

(01:13):
A lot of SEO was keywordcentric and a lot of GEO is
actually about the content.
We actually changed our head ofSEO's title to the head of GEO
or head of AI.

Is it GEO or AEO or AIO?
What do you call them now?
Right now, it of GEO or head ofAIO.
For this reason, is it GEO orAEO or AIO?
What do you call it now?

Right now it's GEO .
All right today it's GEO.

Welcome to the Spark of Ages podcast.
We've got another round table,this time centered on what
people are calling GEO, AEO orAIO.
In a world where search israpidly evolving beyond the
traditional blue link results,we're diving into the future of
how customers find businesses.
We'll explore the monumentalshift from SEO and discuss what

(01:57):
this new era of conversationalsearch means for marketing,
strategy and execution.
We've got two amazing gueststoday.
We have Rishi Mallik.
Rishi is currently the chiefgrowth officer at Workato, an
AI-driven integration platformthat automates workflows across
enterprises.
They are recognized as aGartner and Forrester leader in

(02:19):
the automation space.
He is an expert in growth,marketing and go-to-market
strategies, known for hisability to drive rapid
transformation, and is quicklybecoming an expert in
democratizing generative AIwithin organizations.
With over a decade at Workatoand prior experiences in mobile
and consumer internet companieslike Qik and Skype, Rishi has

(02:39):
been recognized as a 30 under 30leader.
So welcome Rishi, Eric.
Eric Nalbone Eric.
We love him.
Eric leads client strategy hereat position squared.
He's worked with top clients onstrategy development and
scaling growth initiatives,where he's guided advertising
campaigns, analytics andmeasurement best practices, as
well as branding and messagingguidance.

(03:01):
Before he joined us, Ericworked at companies like GE
Energy, eBay, Cabbage, whicheventually was acquired by
American Express and eventuallystarted his own company.
He also holds a degree ineconomics from Princeton
University.
So welcome, gentlemen, to theSpark of Ages.
Thank you, Thank you very much.
Thanks, Rajiv.
You may remember Rishi.

(03:21):
He was in episode seven back inJanuary 2024.
Longtime friend, we loveworking with Workato.
They're a fantastic place.
I just came back from the Worldat Workato event in Vegas, your
first big one in Vegas as partof your worldwide tour.

That's right, I'm still here.
Actually, we're just finishingup.
It was a great event.
We had over 1,500 people, andthanks for joining us there,
rajiv.
It's awesome and.

Eric has hired us at multiple firms that he's gone
to, so we really appreciate youcoming on board at Position
Squared.
You bring a level of depth, andknowledge and thinking that I
think will help transform ourclients, so I really appreciate
you coming on board.

Absolutely Well.
I'm glad to be here on thepodcast and glad to be here with
Position Squared.
You know, after a decade on theother side of the table, being
on the inside is a newexperience and certainly one
that we're really excited about.

You certainly get to see how sausage is made, so,
but really great to have you onboard.
And so today we talked aboutthis notion of AIO, aeo, seo,
geo.
I've even heard it called LLMIO so many different ways of
saying this topic.
So to keyword stuff, thisepisode I've said every one and
I'll probably keep using it thatway.

(04:31):
So my firm at Position Squared,I learned a lot SEO way back in
my days at AltaVista.
So back then I would go tothese events.
I was heading marketing andproduct management and folks
would ask me about all theseways that they can be seen in
the search engines, all theseways they can cheat, and we
would always say to them just dogreat, great content, it's not
that hard.

(04:51):
How hard can it be?
And of course, it's really hard.
Later on, when I started my ownfirm, we then embarked on SEO
as a practice and now, of course, it's changed dramatically.
So to start, eric, can youexplain to the audience exactly
what GEO or AIO is?
What makes it different fromSEO?
Why should they care?
So you know, there's companies,there's people with individual

(05:13):
digital presidents.
It's everyone these days thatcares about this.
So why does the shift matter?

Yeah.
So the shift matters becausehow people consume content
continues to evolve right, andsince search engines have been a
thing, since Google has been athing and we all decided that we
were going to prefer Googleover Bing or, you know, maybe
Bing over Google in someinstances.
But since search engines havebeen a thing, that has been the
front door of the internet andthat is changing and evolving to
the point where the front door,the entry point to the internet

(05:39):
, is moving.
For the first time in probably10, 15 years, we're starting to
see a shift to platforms likeChatGPT, to platforms like
Perplexity, or to seeingintegration of AI-driven
components into Google and intoMicrosoft.
That's an evolution that isn'tnew for anyone that's been
paying attention to the searchspace.

(05:59):
How people consume contentchanges as devices change.
It changes as platforms change,but what it means is that we
need to adapt to how we delivercontent changes as devices
change.
It changes as platforms change,but what it means is that we
need to adapt to how we delivercontent right, and so if the way
that people are going toconsume content changes, the way
we need to create it needs tochange and evolve, and all of us
have seen AI become just amassive part of our life, right?
I can barely go through thegrocery store without asking
ChatGPT to help me understandwhich brand of cheese is

(06:21):
preferred.
Right, let alone you know howwe should fix the cameras for
optimizing the podcast or makingsure that we've validated and
understood workflows in abusiness context.
And so if that's the way thatcontent consumption is going to
evolve, we need to move with it.
And if we move on the forefront, and if we move with intention
and we move with anticipation,we can participate productively.

(06:43):
And I think that that's ourbiggest perspective is that
productive participation in thisAI revolution or
geo-optimization craze is whatwe're after, right?
So how do we participateproductively?
How do we be a part of how theecosystem evolves and changes,
rather than simply be subject towhat happens to us?
There's a massive opportunityif we're willing to act, and act
quickly.

Totally nailed it.
I think it's just this wholeshift right.
So before, the primary way ofbeing found was through search
engines and, to some extent,through social media, was that's
how people would find us, so wecared about optimizing for that
.
Now, with ChatGPT or Claude, orPerplexity, deepseek, people
are now asking about us in thoseplaces and we need to be found

(07:27):
in it, and so that's differentand that's what we'll talk about
today.
So that's a great way of layingit up for us, eric.
So Rishi, you called me sometime ago and said Rajiv, this is
a hot thing.
There's companies today thathave lost a tremendous amount of
traffic, and it's something youcare about as part of driving
demand for Workato and helpinginform those who would look to

(07:48):
learn more about them and buyfrom them.
So how do you buy from yourfirm?
How do you specifically applyAEO or GEO principles, such as
optimizing for factual accuracy,entity optimization, structured
data?
How do you ensure that these AImodels accurately interpret and
summarize Workato's fairlysophisticated automation and AI

(08:08):
solutions for global enterprises?

Yeah, rajiv, it's a great question.
It's interesting, you know,because this stuff is changing
every day and so you know wehave some general tactics that
have been working, but it'sinteresting just how quickly
it's moving and how quickly evenone tactic might change to
another.
But what we found is, you know,I'm part of a number of these
CMO councils and that day when Icalled you was my second

(08:31):
council that month where it wasred alert for folk Like they
were seeing 30% declines intheir actual click through
traffic, and you know what washappening.
In my opinion, it was actuallykind of exciting.
You know, in the last decade wehaven't really seen a change in
search up until now, and it wasthe same old game of trying to
win the SEO battle.

(08:52):
But in terms of how we apply itis.
We were looking at OK, you know,people are going to Google now
they're going to Claude, they'regoing to Perplexity, they're
going to ChatGPT the ways thatthey were searching and the
content that these platformswere sort of bringing to the top
.
They're not necessarily waitingfor a click anymore.

(09:13):
They're sort of summarizingthis information.
But what we found, specificallywith Google and Gemini, is that
it does actually reference thelinks of the sources that it's
getting information from.
And when we looked at ourkeywords that we were trying to
rank for, we specifically lookedat what those links were that
it was referencing and we tookthe pattern matching there and

(09:35):
what we found is that articlesthat are FAQs, that are list
format, articles that are topfive, they actually were more
often being pulled in fromGemini.
So we took a tactic of lookingat all of our keywords and
looking at the articles thatwere being pulled and we
actually had content on thesubject.

(09:56):
That was the interesting part.
So to your point about itwasn't just good enough anymore
to have content, but it'sactually how the content was
written and what's going onbehind the scenes with these
LLMs.
We'll never know.
Every single time the responseis different, but there was a
pattern in the type of contentthat it liked, and so then the
team went on this battle,essentially for a few weeks, of

(10:18):
taking all of our keywords andactually using LLMs to provide
content that the LLMs liked.
It was very meta in that regard, but we essentially leveraged
LLMs, fed it our existingcontent and then gave it a model
content that we thought wasspecific to that keyword.
That was more effective.
So, for example, for a certainkeyword, we said, hey, rewrite

(10:40):
this content in this form of anFAQ.
And what it spit out?
We put that live and within aweek it was picking it up, and
so we actually saw that changein our organic traffic as well.
But we saw it started to turnaround once we put out these new
types of articles.

That's pretty cool , I mean it's so you quickly.
One of the things I do likeabout what you do at Workato is
you're not just optimizing forSEO in terms of like all the
keyword stuffing and sort offake backlinks.
You're actually creatingoriginal content.
You guys have videos.
You already had things in aform of Q&A and sounds like you
refined it even further and thentested and measured it even

(11:17):
further, so that's really cool.
Here's a question for both ofyou, as user trust in AI-sided
content is still forming and AImodels are tasked with
constructing the most usefulresponse.
So how can brands ensure theircontent, which is designed for
AI interpretation, authenticallyresonates with human users and
prevents the potential fromdoing more harm than good if

(11:39):
machine-predicted relevancedoesn't align with human
perception?
So what are the best practicesfor maintaining human centricity
in this AI world?

A lot of it has to do with how we write the
content.
So, as much as we're actuallyleveraging LLMs to produce
content that we feel the LLMswill pick up, but we do have a
review process at the end of theday.
So we still have our communitycouncils, we still have our
customer advisory boards and weleverage that to run the content

(12:10):
by them.
So if there was somethingwritten a certain way, you know
the LLMs are going to put outtheir interpretation of what we
put out there.
But we still needed to resonatewith customers and we still
leverage those channels.
I mean, the customer at the endof the day is the one that's
going to receive this, so youcan't completely take the human
out of the loop on this one.

I agree, and I think that part of what makes
all of the AI models socompelling and so interesting is
that they're conversing with usin a human-oriented fashion.
So in many ways, they're takingcontent and they're
reformatting in ways that wealready talk and we already
think.
And if you think about who youknow that are good communicators
and how they articulate ideasand how they present information

(12:54):
, they do it in a structured waywhere they have an idea, they
have support for ideas and theydraw conclusions based on the
information that they'vegathered.
And that's really what AImodels are trying to do.
And so to maintain that humanauthenticity in many ways with
an AI model can be somewhateasier than trying to maintain
authenticity for an SEO page.
And we talked for a long timeabout writing for search engines
and what the balance was ofwhen you're writing for a search

(13:15):
engine and you see a page thatjust scrolls and scrolls, and
scrolls and it's how many wayscan we say the same thing over
and over, focused on a verytight band of keywords?
Can we say the same thing overand over, focused on a very
tight band of keywords, and10,000 words later?
You know, you're not reallysure if humans have learned
anything, and you're not evenreally sure if humans will ever
really consume most of thatcontent.
In an AI world, we don't need10,000 words of scroll.

(13:36):
We need a couple of frameworksthat AI models can interpret
right, and when Rishi talksabout adapting things into FAQ
formats, that's because peopleask questions and they want
answers.
They don't want novels.
Novels are a dying art form andI happen to love them, you know
, as you can see, but peopledon't want novels when they're
trying to come up with answersto problems.
They want solutions and theywant them quickly, and so what

(13:58):
they need is they need thoseframeworks, and so it lends
itself well to structuredcommunication, and people who
can think through that andorganizations who can think
through that can create contentthat is human oriented in ways
that sometimes it's anopportunity that SEO has taken
away from us and we've gottenaway from that.
That AI offers us the chance tokind of get back to that core
communication skill.

Right, it's about that notion of snackable content
.
It's creating snippets.
It's really in ways that weconsume chunkable data.
That's how we, as humans, work,and so when you're trying to
write something for a searchengine, you may not hit the mark
.
Go ahead, rishi.

No, I love what Eric said.
It's funny.
I feel like the old way tomarket in B2B SaaS was to put
out a one pager, put out a whitepaper, and look at the intent
that came from that.
And it's funny.
You know, I'm the secondlargest buyer of technology at
the company not something tonecessarily be proud of, but in
terms of building out ourmarketing stack, we do interact

(14:54):
a ton with vendors and I'venever looked at a white paper or
a one pager.
In terms of how I buy and toEric's point it's you know, how
do I get to the point as quicklyas possible and that's
authentic to me in terms of howI buy and that's what's going to
resonate with prospects.

Yeah, and so you mentioned this earlier.
So, eric, maybe you can answerthis.
So Google searches, right, arethe ones that end without a
click.
These are AIO driven or AIoverviews, and they've been
shown to reduce clicks towebsites by 20 to 40%.
So now, how are brandsredefining the overall value
proposition and the tangible ROIof their online presence?

(15:31):
Right, everyone gets measured.
Right, we all have KPIs interms of that outcome of search
visibility, right?
So how do you work with that inthis world?
So you want to optimize for AIO, but then you're not
necessarily going to getreference traffic from that.

So I've done this with a number of clients where
we have this conversation, and Ilike to bring up what I call
like hey, this is your search isdead graph, right, where I open
up Search Console or I open upGoogle Analytics and I show them
.
Okay, if you think search isdead, let's just go take a look
and let's look through howtraffic is showing up on your
website, where it's going andwhat people are doing, and it

(16:04):
turns out that search veryclearly isn't dead because
people don't transact in many,many ways.
You're not transacting insidethese AI models, right, and so
your journey doesn't end in anAI model and your journey
doesn't end in chat GPT withmost of the things that we're
looking at.
Right, I'm not researching andhaving a conversation with chat
GPT just for kicks.

(16:25):
I'm having a conversation withChatGPT because the information
that I'm going to get out ofthat helps me solve a problem,
and then I go and I completethat transaction, or I solve the
problem with the vendor that'sgoing to help me do that, right,
and that can be from somethingas complicated as determining
what type of CRM I want to useor the optimal site map

(16:46):
organization for a website thatI'm developing.
Or it can be as simple as I'mwalking through the grocery
store and I don't know whatbrand of cheese to buy, right,
to just use the trivial examplethat I referenced earlier.
But either one of those, thetransaction doesn't happen
inside ChatGPT, and so when youlook at the traffic that you're
still getting, you're gettingtraffic that's answered a little
bit more questions, they'vesolved a little bit more of

(17:07):
their exploratory journey beforethey hit your website, but
they've consumed that contentand they've solved their
questions before they hit yourwebsite.
You, as opposed, to proceed withyour competitor in an ecosystem
that you don't manage directly.
And that's fine, right, as longas you stand firm on your value

(17:31):
propositions and you believe inyour service offering.
I'm actually agnostic as towhether you make the decision on
ChatGPT or you make thedecision while browsing my
website.
If you make the decision right,and when you talk about how we
measure, I think again, it's oneof those things where you can
frame it as a challenge, whereyou know traffic is down, or you
can frame it as an opportunity.
If revenue is up but traffic isdown, isn't that a win for
everyone, right?

(17:52):
And if this is a threat to youas a marketer instead of an
opportunity.
You're measuring the wrongthing, because if you're
measuring company outcomes, thenpeople consuming information
and making a decision, that thatpaints your company in a great
light and paints your offeringas a solution to problems that
people have.
That's the outcome that youshould measure.

so are you going to shift the way you measure it
now?
So you're going to say sobefore you would get data back
saying here's my organic result.
Right, yeah, now you may get aresult from chat, gpd, but
specifically ai overviews,they're just not good.
They're going to read it and bedone with it.
How are you going to measurethat?

Smart marketers have always understood what's
kind of an up funnel metric thatshould be indicative of down
funnel success and what isactually a success outcome in
and of itself.
And so if it turns out thatwhat we need to be measuring are
engagements and citationsinside AI models, then that's
what we'll measure, right.
And if clicks turn out to bethe wrong metric for a certain

(18:48):
type of content, then that'sabsolutely fine, right.
And so we've seen that where Ido want to measure click traffic
to specific portions ofwebsites for our clients,
because those are wheretransactions occur, and I'm
still going to measure clicks onthat traffic, there's other
sections entirely where I'vesaid I actually don't care about
the click traffic anymore.
I'm going to substitute andmeasure impression.

(19:10):
And you know, whereas we usedto go from impressions to clicks
because it's down funnel, nowit's not so down funnel anymore.
Impressions are the thing.
Then where do you get the click?
Where do you get the traffic?
Maybe this becomes directtraffic now as a measurement
instead of inbound trafficthat's tagged with Google, and
so all of the metrics that welove to look at.
You know, we want the mostimmediate feedback that we can
get, but we want things that arepredictive and indicative of

(19:33):
downstream success.
Right, we want things that aregoing to manifest in the P&L.
We want things that are goingto manifest with customers, not
things that live in GoogleSearch Console.
The only person that makesmoney off of Google Search
Console is Google.
Right, we make money whencustomers transact with us and
we keep making money whencustomers solve problems with us
and they have success with us.
What that measurement means islike let's get smarter about

(19:54):
sectioning out our website andunderstanding hey, this is
content that I'm fine with beingconsumed off of the website.
It's there because that's thedelivery mechanism to ChatGPT.
That's the delivery and how Iget this content to quad, so it
needs to be on the website.
But if it's being seen on chatGPT instead of our website, as
long as the decision is the same, that's great.

Rishi is very data-driven right, he lives for
data but at the same time, Ithink you say you love data,
decisions informed by data, butyou're not so data-driven that
you're solely focused on oneelement versus another.
So GEO shifts success metricsfrom keyword rankings to
mentions and citations withinAI-generated responses, and
that's actually hard to measurebecause every engine is going to

(20:36):
measure it differently, they'regoing to put weights on
different aspects of things andit's going to be personalized.
So how is Workato evolving yourown internal measurement and
iteration frameworks to trackthe quality and prominence of
your brand and content within AIoverviews and other generative
AI engines, in the absence ofthese standard tracking
mechanisms?

Yeah.
So we implemented the age-oldgrowth tactic of actually
allowing the user to self-reporthow they heard about us.
One thing that we did recentlyis to actually get to the root
of this problem and again it'snot a hundred percent, but it's
definitely directional which iswe still have our typical
tracking mechanisms or analyticplatforms that tell us the raw

(21:17):
click-through data.
Obviously, it attributes stuffto direct to Google in ways that
may or may not necessarily beas accurate, but then we matched
that up with.
We just had that simple fieldof how did you hear about us and
we sort of compare what usersare actually saying, how they
heard about us, versus what theanalytics are actually telling
us, and what we've found is alot of that.
Direct traffic is now comingfrom these platforms.

(21:39):
People are self-reporting thatthey're coming from chat, gpt,
that they're coming fromperplexity, that they're coming
from cloudGPT, that they'recoming from Perplexity, that
they're coming from Cloud andpiggybacking on something that
Eric said, which is when they'recoming from these platforms,
they are more informed.
So we actually looked ateveryone that was self-reported
coming from ChatGPT how thoseconverted down the funnel and
the conversion rates areactually higher.

(22:00):
So they're not just coming inbecause they heard about us, but
they're coming in with a muchmore informed decision of what
they're looking for.
That really helped.

You're hitting upon one of the things that
we've all been seeing more andmore in B2B decision making,
where it follows along withconsumer decision making that 70
to 80% of people are coming toyou even without talking to a
salesperson.
They're doing their ownresearch, they're figuring their
own things out, and so itsounds like you're seeing a
measurable difference in thatnumber, even with this general
trend.

Absolutely.

I'll ask this for both of you and Eric, maybe you
want to start.
It's like so there's thisnotion of visible in search
equals accessible to AI, right?
So now the content that youhave, these are the raw
materials for synthesized AIresponses and, rather than being

(22:50):
ranked and clicked, there's afundamental shift, right.
And so how do you redefine thislong-term investment for
content creation?
And I think we've talked aboutthis a lot of times that you
move away from single topicdata-backed pages to
comprehensive content hubs right, that can be synthesized.
So how do you think about thatin the content lifecycle?

Yeah, so exactly what terminology you use, we can
split hairs, but the way thatI've been helping people
understand our strategies arethat in an SEO-driven world, we
were really focused on keywords,and in an AI-driven world, I
want to be focused on topics,and so, from a very tactical
perspective, designing an SEOpage is about identifying one to

(23:24):
two target keywords and threeto five supporting keywords and
then making sure that we'vegrabbed them contextually and
talked about them in all thedifferent ways that search
engines interact with those kindof five to seven keywords, if
you will, and then if you wantto target another five to seven
keywords, you're going to wantanother page, and then if you
want another, you know by thetime you want to target 700

(23:44):
keywords, you need 100 pages.
I'd rather address a topic in anAI driven world, where each
topic has components of it, andwhat ChatGPT is trying to do is
build a framework and understandhow do you move through this
topic and what pieces arerelevant, what dots connect to
each other and how do I build anunderstanding of this problem

(24:05):
space so that I can formulate aunique answer to your question.
Right, it's not about I want torank order pages from one to a
million, or one to infinity, torank every page on the internet.
With respect to your query,this is I want to understand the
topic that you're exploring andI want to go grab bits and
pieces and surface them.
They're just fundamentallydifferent problems and so we
address them differently.

What's an example of that?

Yeah.
So an example of that would beI'm trying to understand what
CRM is best, right, okay.
Well, that's a complicatedquestion.
What kind of company am Iengaging with?
Is it a B2B company?
Is it a B2C company?
What kind of MarTechintegrations do I care about and
do I care about?

(24:47):
And in an SEO driven world,martech integration with the CRM
might be one page right, andthen that could be Salesforce
integration with Google, Then itcould be Salesforce integration
with LinkedIn, then it could beHubSpot integration with Google
, then it could be HubSpotintegration with LinkedIn, and
those can be four separate pages, versus a hub page or a topical
page that is more cognizant ofstitching together CRM and
advertising platforms.
Right, and now a page thattalks about stitching together

(25:09):
CRMs and advertising platforms.
That's not so specificallykeyword driven, helps an engine
understand that space right.
Here are the different blocksthat we are working with.
Here's how we believe they fittogether Essentially, by
publishing this content.
This is our recommendation to anAI engine that we would suggest
that you adopt this frameworkand that you use this framework
when you organize content foryour users.

(25:30):
Chatgpt isn't buildingframeworks from thin air.
It's looking for frameworks,it's looking for ways to connect
dots and if we can suggest thatthese are the dots and this is
how we've connected them.
Maybe ChatGPG draws a slightlydifferent picture than I do, but
you know we're building theconnect the dots picture right.

Rishi, maybe you can illuminate that right.
You have iPass, you have thenotion of integrations, you have
workflows and now you have thenotion of agents meant for the
enterprise.

Absolutely.
Yeah, it's.
You know, we've made thistransition With Roccato.
we've built out, you know, sortof the all of Gartner and
Forrester have named us for theleader now when it comes to the
integration space, so the leaderin iPass, and we've had a slew
of agentic features that havebeen a part of that.
And you know, with how peopleare buying today, we have this

(26:22):
very modern integration platform, but we really want to talk
about a lot of our agents today.
It doesn't take away from theintegration platform, but that
is what people are interested inat the moment and so sort of
almost a reframing of us beingsort of the new age agentic
platform for the enterprise thathappens to be built on sort of

(26:43):
the leading iPaaS, for theenterprise that happens to be
built on sort of the leadingiPass, and the way that buyers
typically bought in that spaceis changing.
And, to Eric's point, in termsof the framework of how people
buy, a lot of our content now isaround what the needs of an
integration platform arerequiring, sort of this agenda
capability, and that's one ofthe things that the LLMs are

(27:05):
starting to pick up on.
Which is it used to be?
Hey, workato is the leadingintegration platform, and now
it's talking about Workato beingan agentic platform for the
enterprise that happens to bebuilt on the most modern iPaaS,
and that change and that shift.
You know it's not one web pageis not going to get you there.
It's having a slew of web pagesthat repeats that sort of

(27:26):
framework and I think theinsight is a lot.

It's web pages.
They're starting to emphasizevideo content more and more.
I think you guys have a lot ofgreat video content.
It's multiple types of contentthat are stackable as well as
enable you to go more in depth,and what you're talking about is
a really special instance ofthis, because the way LLMs are
built right, they'll run andbuild a model, and so that's
long-term memory right of howthey're built.

(27:50):
So then you're coming at itfrom.
Well, we were considered aleading iPaaS company with
automation integrationcapabilities.
Now you're moving to theagentic enterprise, and so you
have the short-term memoryaspect of it, where it's going
to different search engines andupdating it.
So then there's this sort ofclash between the two, right?

(28:11):
So you have to manage thatshift to make sure that you're
seen in those engines, because,I would imagine, your users are
leading edge, your buyers areleading edge, and it's something
that we have to continuouslymonitor too.

It's a big piece of it because there's other
content that's going out there,not necessarily competitive, but
there's competing frameworksthat are potentially being put
out there not necessarilycompetitive, but there's
competing frameworks that arepotentially being put out there
in terms of how people buy.
So it's not a one and done sortof piece.
We actually changed our head ofSEO's title to the head of GEO
or head of AIO.

Is it GEO or AEO or AIO?
What do you call them now?


All right, today it's GEO A little bit more
general, I think, aio, I thinkyou call them now.

All right, today it's GEO Alittle bit more general.
I think AIO, I think Googlesort of coined the term, and so
we try to make it a little bitmore general.

It's a good call.
At the same time, search isstill the go-to for most people.
Are you seeing that dramaticshift or are you seeing it more
incremental?
I know Google reported itsrevenues recently and search

(29:10):
traffic was actually up.
Search revenues were actuallyup.
Now I know that's come at theexpense of pushing down organic
results.
Right, it's AIO results or AIgenerated, then it's sponsored
ads, then it's organic.
So maybe from a revenue pointof view they got a bump, but
they also seem to have a trafficbump, so it's not dead.
How do you think about it?

It's interesting.
We're seeing a dramatic shift.
So in the self side, we'reseeing a huge, dramatic shift.
About 20% of our traffic now isbeing classified as coming from
regenerative platforms.
But to Eric's point again, alot of these sort of piggyback
on each other If someone foundthat Ricotta was interesting for
something that they werelooking for as they were having

(29:42):
a conversation with ChatGPT andthen they open up a new tab and
search something related to thattopic in Google.
That's where these things sortof play off each other a little
bit more.
And right now, you know Rajiv,it is a little bit tough to
figure out what that userjourney is.
The thing that we are seeing isthere's a dramatic shift in how
people are self-reporting thatthey're finding us, and now are

(30:04):
they doing 100% of theirresearch on ChatGPT before they
get to us, or 100% of theirresearch on Cloud before they
get to us?
Or is it a combination ofpotentially using one of these
LLMs to doing a Google search,to potentially going back to
these LLMs?
That user journey, I thinkit'll be time until we get a
full insight of it, but we areseeing a dramatic shift in that

(30:27):
direct content coming to us.

Great, this is really helpful.
So here's both of you and I'mgoing to have Eric start.
So, as generative AI continuesto evolve its understanding of
how it synthesizes and itssynthesis of information, how
are forward-thinking brandsexperimenting with advanced
applications of schema, internallinking and cross-platform
signals so to not only claim butactively reinforce their
position as definitive author oftruth on specific topics?

(30:53):
So it's influencing AI toattribute and amplify their
contents preferentially in thiscrowded digital landscape.
And even Rishi you kind of hitupon it saying so many folks are
coming in with differentframeworks in today's market,
especially given how much yoursis changing.
So think about how are thesefirms experimenting and getting
ahead of each other using thesesomewhat technical ways of

(31:15):
thinking of things?

Yeah.
So I think that there's acouple of ways to think about it
, and the first is that modelswill use what they have
available to them, right, andthe things that are most helpful
to AI models today are thestructured elements of how pages
are built.
Those are the things that existthat were created, ultimately,
with search in mind.
Right, google didn't go createmarkup schemas, or markup

(31:37):
schemas were not exclusivelycreated by Google, but they
weren't adopted as an SEOframework by accident.
They were created with anintent to be used in search
engines.
As we've moved towards AI, right, and we think about AI as this
revolution, since, you know,openai or ChatGPT came out, but
it's been in the making for awhile.
As you think about, you know,the knowledge graph, you know

(31:58):
there's been moves in thisdirection since search has been
out there, right, and thetechnical elements of the page
really are easy for engines toconsume.
Right, schema that says what isthis page about, what is it
referencing, what topics areassociated with it.
All of those are gold for AIengines.
What changes is that they don'toperate off of the same kind of

(32:21):
ingested knowledge of the world, and so where Google has a map
of all of the internet and howit relates to each other in the
form of backlinks, right and itbuilds authority by knowing that
Forbes has a lot of authorityand when it writes about you, it
passes some of its authority toyou.
And when you write aboutsomeone, you lend some of your

(32:42):
authority to them, and it's thiswhole kind of distributed
scheme of authority flowing frompeople who have it to people
who need it.

Yeah, that's how Google has always been built,
right.
This is a page rank orauthoritative links from one to
the other.
Right, so Wikipedia, yeah.

So now flip it for AI, right.
And if ChatGPT doesn't havethis whole model of every link
that exists on the internetbecause they've crawled the
entire internet and built thiskind of holistic model of all
the backlinks that exist, whatis an AI model going to use as
its symbol of authority?
Well, it's going to start tounderstand expertise in a

(33:17):
different way.
It's going to understand thatbecause you relate to, you know,
if we think about a healthcarecompany, because you relate to
medical schools, through yourconnection with individual
doctors who have degrees from aninstitution which you know,
definitionally has authority,that type of authority signal,
which was helpful in SEO,becomes crucial in geo right,
and so we've always talked aboutbacklinks as this thing where
you want a lot of thingspointing at you.

(33:37):
Well, in a geo world, we'vetalked about the idea of
borrowing authority from people,proactively borrowing authority
by saying you know, rajiv, youhave authority by virtue of your
connection with the HBS network.
Okay, well, let's call that out, and that's the way that we're
going to borrow authority now,with an outbound link right
instead of an inbound link, andso you can do that, broadly

(33:59):
speaking, right, we want to citesources, we want to point at
high value sources that areoutside of our world, in order
to pull some of that authorityin and start to help AI models
understand that we've compiledauthority through a variety of
sources and so, in some ways,again, it's an opportunity,
right, whereas buildingbacklinks is a really, really

(34:19):
hard thing to do in SEO land.
It's rife with spam.
It's rife with, you know, allsorts of low quality players.
Borrowing authority whereyou've already accumulated it
for your organization or for anindividual is actually wonderful
in the geo world.
You know we're getting lost inthe alphabet soup here, but we
love this idea of author oftruth, right, and this question

(34:39):
of kind of who you shouldbelieve.
You know, why should youbelieve me?
Well, you should believe mebecause we have this resume of
emulated authority, right, andlet's let's show you what that
resume looks like and tell youabout it, and it's a lot easier
for me to tell you than it isfor you to research it and find
out all the reasons why I shouldbelieve it, or for you to just
trust that Google has done thatfor you.
So we love that shift and Ithink it's a great opportunity

(35:01):
for people who really do havethose credentials and are
willing to cite their sourcesand are willing to say that.
You know we've built a strongcase here it is you know we've
built a strong case here.

It is Rishi your thoughts.
I think it's spot on.
I think we've seen the exactsame shift.
You know, a lot of what'spicked up now has to do with,
you know, sort of some of theaccolades of the company.
And one of the things that wefound is, when we release
content, even if it's just onelink or one piece of content,
it's really depending upon howimpactful the actual content is

(35:34):
versus like how widespread thelink is or how many people are
referencing it ChatGPT or Cloudor any of these other platforms.
They give it a lot more weight,they give it a lot more heft
based on what's actually in thecontent versus oh, does it just
mention these keywords?
And if we're talking about aspecific topic around Riccardo's
agentic platform, for example,there are other agentic

(35:57):
platforms that are out there.
But when we talk about,potentially, some awards that
it's won or the number of usersthat are using the platform or
the actual accounts that areusing, or actual customers that
are using it and their brandrecognition, it definitely gives
it a lift and it's nice nowthat these platforms actually
have a way of sensing, like thetrue content versus gaming, the

(36:21):
system between by, like howwidespread your link is.

Right.
So nowadays, g2 may matter more, or Reddit seems to matter more
.
They have an agreement withChatGPT, right?
So people who are writing aboutyou, but not in a backlinky way
, I guess, is the way of puttingit.
So you're referencing authorityas Eric talks them up.
At the same time, people arereferencing you and it's
starting to put those thoughtstogether about whether you're an
authority or not.
So we're now going to shift toopinions about GEO versus SEO.

(36:48):
The rules of digital visibilityare being rewritten and the
battle lines are drawn.
For years, we've all beenplaying by the same playbook,
which is SEO.
We're about to throw down someopinions that'll force you to
choose a side about where youthink tomorrow's AI answers will
lead us.
So here we go.
Here's number one GEO is just arepackaged buzzword, heavy

(37:09):
version of good SEO.
The best way to get cited by anAI is to create a high quality,
authoritative piece of contentwith a great user experience,
which is what SEO has alwaysbeen about.

Partially agree.
I think that GEO is arepackaged term to make it
easier for people to understandwhat it actually is, but the
game is completely changed.
It's not the same as a lot ofSEO was keyword centric and a
lot of GEO is actually about thecontent.

I love it.
That's a great response,especially since you just gave
your guy the title GEO.
So go ahead, Eric.

Yeah, I agree.
In many ways, geo is arealization of the promise that
search has always had but failedto deliver.
And this idea of being able tofall back.
Geo is a realization of thepromise that search has always
had but failed to deliver right,and this idea of being able to
fall back on content as a sourceof truth.
We've forever heard Google'smantra of just create good
content, just create goodcontent, and sometimes that
works, sometimes that doesn'tright.
Google told us that backlinksdon't matter anymore, and

(38:06):
evidentiary proof shows thatthey do.
We've seen a lot of that kindof gamesmanship in the SEO world
and in many ways, ai kind ofhelps us take the step towards
realizing what people have hopedsearch could be in a way that
is really exciting for us.
So you know, that partialagreement, I think, is a good
place to land on that topic.

So here we go.
Here's the next one.
The shift to GEO will destroythe open web.
When users got their answersdirectly from an AI without ever
clicking on a source, allcontent creators, especially
news sites and bloggers, willlose the traffic and ad revenue
they need to survive.

Absolutely disagree.
I don't think the web is goinganywhere.
I think that there's absolutelyvalue in having your own
content in your own ecosystem.
I think the transactionscontinue to occur between an
individual or a company andanother company.
They're not going to occurbetween ChatGPT and another
company unless it's paying foryour monthly subscription.
You know and I think that theopen web is here to stay.

All right, rishi, completely disagree.
I think that it's actually madethe web a more exciting place.
I think it actually gives theunderdogs a voice now, even more
so, a lot more so than before.
You know, these huge companiesthat had huge marketing budgets
before could win the SEO battleinauthentically, and now, if
you're an underdog but you havethe right content or you have an

(39:25):
offering that resonates withfolk, you can stand toe to toe
now.

Okay, let me push you a little bit.
Resonates with folk.
You can stand toe to toe now.
Okay, Let me push you a littlebit.
So if I can now go and talk tothat wonderful voice that I have
on ChatGBT and get really goodcontent from some other source,
from various good qualitysources, how's that person going
to make money if that's allthey do is drive that content?

They don't need to make money on the first click.
Hopefully you can actuallyleverage these platforms to
inform decision-making orwhatever behavior through
conversation, throughconversation with that voice,
and then you know when they'reready to transact.
They'll come to you if you havean offering that's useful to
them.
That sounds great.

My team would say, especially if you're a content
creator, you're going to have togo to the next level on
Substack and pay someone asubscription, all right.
The next question SEOprofessionals who don't
emphasize AI content creationare going to go extinct.
You can't keep up with thevolume and speed of content
required to dominate AI-poweredsearch results without using AI

(40:25):
as your partner.
So this gets to a little bit oforiginality and using AI to
create content.

I agree.
If you're not playing in thesame game or if you're not
moving with the times and you'renot actually playing, a new
hand, that's being dealt.
I'm surprised that they're notalready extinct.

You got to keep up .
I agree with the asterisks,which is that we're not there
yet in terms of AI generatedcontent being perfect, and we
talk a lot about AIhallucinations and we talk a lot
about the incorrect conclusionsthat it draws.
So if you completely abandonedyour AI content creation or your
content creation to AI, you canget yourself in trouble In a
world where AI generates contentfor AI to consume.

(41:03):
You know that we just go in arecursive loop.
You know what's the sayingTurtles all the way down.
We don't want to get to thatworld, so there needs to be
oversight, like with anything.
I think that if you just set itloose on the world and you say,
okay, we'll go, there's actuallya site in the financial space
which is churning out likedozens of articles per minute,
that I was reading a Redditthread that is just seeing

(41:25):
skyrocketing traffic lately andyou know it's working for now.
And it's working for now and itreminds me of situations that
we've seen in SEO, where thingswork for a little while and then
someone realizes that, okay,that's actually not quite what
we're going for.
We're going to shut that down.
So I wouldn't say you canabandon your AI content creation
process to AI, but if you'renot using it to inform and help

(41:46):
accelerate your content creationprocess, you're missing the
boat Awesome.

Yeah, I think the other, you know, with AI created
content for AI to consume.
We run into originality, right,which is something that folks
have an issue with and that's awhole topic in and of itself.
But you're going to still haveto be original and if an AI
engine puts it out, you can'tjust go out and release a
million articles that are just,you know, just BS.

(42:10):
Okay, next one Optimizing forGEO is a fool's errand.
You can't game a generative AImodel the way you can an
algorithm.
The only reliable strategy isto focus on a strong brand and
loyal audience, because AI willeventually favor the most
reputable sources, not the mostoptimized one.

I disagree.
I think that we're alreadyseeing it with the models where
you can put out one piece ofdisagree.
I think that we're alreadyseeing it with the models where
you can put out one piece ofcontent.
It's not about quantity, it'sabout quality.
We've seen the experience timeand time again where we put out
one updated article, but thecontent was of such quality that
the generative platforms arepicking it up within days.
And it's less about gaming,it's more about the quality of

(42:52):
the content.

I agree with what Rishi explained earlier, where
you talk about taking existingcontent and asking for it to be
rewritten and reformatted in away that's acceptable or
attractive to generative models.
The same content formatteddifferently, will be picked up
and highlighted differently, andthat's not to say that you're
gaming it right.
And I think that that's wherepeople have always gotten into

(43:14):
trouble with search.
If you look at search as a gameto be won rather than a
continual iterative optimization, google's trying to deliver and
ChatGPT is trying to deliverthe best result for users.
We need to align with whattheir best practices are in
order to do that.
If we're doing that at ahighidelity way, I don't think
that's gaming the system, but Ithink it's moving where it'd be

(43:34):
risky skating to where the puckis going.

Yeah, Remember that company Demand Media, where
they would pay writers to writeabout various hot topics and
companies would pay for that.
And that lasts for a certainamount of time, but at some
point you have to do somethingof high quality, and so this
leads to the next one, which Ithink is kind of similar to the
last one.
The only content that willsurvive the AI revolution is
content with strong EEAT signal.
Generic, low-value content, nomatter how well optimized for AI

(44:00):
, will be devalued, proving thatquality and authority will
still trump all.
So EEAT means experience,expertise, authoritativeness and
trustworthiness.

I agree and I think that in a world where it's
much easier to create content,low value content just gets
shuffled to the bottom muchfaster.
I referenced a financial sitethat's churning out hundreds of
articles a day, driven by AI,and they call it out on their
site.
They say this is generated withAI, it's working for a little
while, but it's not going tolast for long, and that low

(44:31):
quality content will bedeprioritized and even AI
engines are going to start torealize that, gee, it's a bad
thing when I actually pick up oncontent that I just created.
Low quality is going to bepenalized and whether it's on
the longer cycles from the earlydays of SEO or it's on fast
cycles these days, we'll see lowquality content continue to be
penalized as well.
We should.
I don't think anyone wants toconsume low value content.

One of my friends, with his analytics company, was
generating thousands of pages amonth and he's like Rajiv,
we're knocking it out with allthis AI content.
And, yeah, you're right, itworks for a little while.

And Rajiv, you mentioned it earlier too which
is the voice of the authentic,user-generated content, is just
as important.
Now, with these models, so manyresponses that you might ask
ChatGPT or Cloud or Perplexity.
They'll give you sort of whatthey feel is factual, but then
they'll also say, you know,users have said X, or some

(45:25):
reviews have said Y.
I remember looking uprestaurant recommendations and
it was comparing cuisines, butthen it's also giving reviews
right, and so we have toremember that when it comes to
content creation, it is you knowfirst.
Low quality content is notgoing to survive.
It needs to be high qualitycontent.
But then having that contentbacked up by what your community
is saying, what users aresaying on other places of web

(45:48):
social, so on and so forth, isnow just as important to these
models as well.

I think that's a great way of putting it.
When you think about a movie,right, rotten Tomatoes may rate
something really well, and wherethe experts have a different
point of view than the useraudience, right, and so you have
to help, it has to help you getto a more personalized answer
to you, right?
I mean, that's the power ofNetflix, right?
It figures out what.
This may be a great movie interms of Rotten Tomatoes, but it

(46:14):
may not be your type of movie.
Same thing for your type ofrestaurant.
So the agents need to give youboth.
So this leads to this greatnext question, which I think is
it's about the specialist versusgeneralist world.
So GEO will lead to ahyper-specialized internet.
Since AI rewards depth andcomprehensive coverage on a
topic, the only way to be citedis to become an ultimate

(46:35):
authority on a specific niche,leaving no room for generalist
creators or publications.
Agree or disagree, Do I'll?

go niche-y or do generalists still do?
Okay, I agree, and I think thatwe'll start to see people get
rewarded for talking about whatthey know, and I don't think
that's a bad thing.
I think that I want to talk tothe people who are experts at
what they do, right?
I like talking to smart peopleabout things that they know a
lot about.
Whether I'm talking to a humanor whether I'm talking to a
company, I want the right answerand I want it in the level of
depth that is of interest and ofneed to me right now.

(47:04):
You know a hyper-specializedinternet.
It's absolutely going to dothat, right.
You can create your own GPTsinside ChatGPT, and I've got
half a dozen of them that dodifferent things for me and look
through content in differentways Absolutely
hyper-specialized internet, andI think that that's what we
should all be here for.

Is that the end of the Renaissance man?

I don't think it's the end of the Renaissance man,
but I don't know.
I'm a liberal arts guy with aliberal arts background, know, I
think that there will always bea place for the Renaissance man
.
But one of the things that yourealize as a Renaissance man is
to know when you don't knowsomething, and go ask the expert
, for you know the right answer.
So I think that that's whatthere will be is a place for

(47:43):
exploration and then a place forspecialization.
It's awesome.

Thank you, those are great.
Go ahead, rishi.

I think the generalist or the Renaissance
man will need to evolve.
It's really interesting withthese platforms.
I think it's actually teachingus how to be more niche in our
searches.
So, for example, if you ask anyof these platforms a very
general question, it willactually give you more
information probably than youwere looking for.
It won't just give you ageneral answer and walk away.

(48:10):
It'll start pointing out littlenuances for you to get into
your own rabbit hole, whichleads to this concept of prompt
engineering.
I mean, the most basic consumeris becoming an expert in prompt
engineering and they'rebecoming more advanced in how
they ask questions, which leadsto this rabbit hole of becoming,
you know, more niche in ourquestions and how we ask them.

(48:32):
So I think we're going to seethis shift of more and more
questions and people being moreadvanced in prompt engineering,
which is going to require thisshift in terms of the content
creator to be a little bit moreniche as well.
That's great insight.

Yeah, great insight on that.
It's like you can ask a generalquestion.
It'll ask you to dig intospecialized ways, like I can ask
it about a particular medicalcondition I have.
I'll actually feed it my testresults.
It'll give me an answerspecific to my situation, like
when I tore my Achilles andabout PT methods and maybe some
things I should supplement with.
But then afterwards, as I wentalong, it would keep asking me

(49:05):
what kind of PT plan do I want,what kind of things do I want?
It would keep prompting me withmore and I could change the
subject completely or I couldstay with it.
But it was smartly built forconversation.

You'll see, these days I feel like people are more
informed about very nichethings.
Just, I feel like in the lastsix months, the conversations
I'm having with folk, people aregetting into rabbit holes every
day.
Before, it was when you doveinto a rabbit hole.
It was a matter of looking at,you know, potentially a
documentary or watching a movieon the subject and reading a
book on the subject.
I mean now, within 30 seconds,you can become an expert on

(49:41):
herniated discs, or or, you know, implementing a new framework.
I mean, within 30 seconds, youbecome an expert on the topic
and I, at least in myconversations that I'm having
with people, I'm just surprisedabout the things that they know.

I think it's cool because then you end up getting
deeper with folks.
Either A you get exposed whereyou can only go so far, or, if
you're more open to it, you getinto a deeper conversation.
I think that's part of thathuman experience.
Wow that was really interestingto talk about all that's going
on with GEO.
Now we get the super funfeature of the show called the

(50:15):
Spark Tank, so I want to welcomeyou to our game.
Welcome to the Spark Tank.
Today, we're joined by RishiMalik, the Chief Growth Officer
at Workato, and Eric Nalbone,chief Strategy Lead here at
Position Squared.
We're about to test whether ourgrowth experts can spot the
difference between the tacticsthat actually happened and the
ones that sound too crazy to betrue.

(50:36):
Here's how it works.
I'll give you three statementsabout historical marketing
tactics from advertising's mostmanipulative eras.
Two of them are absolutely real, the kind of strategies that
major companies actually usedback then, when regulations were
a bit loose and ethics were abit flexible.
One is completely fake butdesigned to sound just

(50:57):
believable enough to mess withyour heads.
After each round, I'll countdown three, two, one, and you'll
both reveal your answerssimultaneously, ready to find
out who really knows theiradvertising history.

Let's go, let's do it Bring it.

Bring it.
Two truths and a lie.
Historical marketing tactics,enhanced edition.
All right, round one.
We're going to start with theprint to radio revolution of the
20s and 30s.
So number one Wheaties becamethe breakfast of champions by
sponsoring individual baseballplayers at bats with
announcements saying quote thathome run was sponsored by

(51:35):
Wheaties after each hit.
Number two Ovaltine createdDakota rings for little orphan
Annie listeners, but the secretmessages were actually
advertisements disguised as spycommunications.
Number three Listerine inventedthe term halitosis and funded
medical studies to prove it wasa serious social disease that

(51:57):
required their mouthwash to cure.
So you ready?
Three, two, one.
I see both of you at two.
I will be here at two.
Yes, okay, the answer is numberone the individual at-bat
sponsorship's a lie, although Ican see why you picked number

(52:19):
two.
Wheaties did become thebreakfast of champions and
sponsored sports broadcasts, butthey didn't sponsor individual
at-bats with real-timecommentary.

Ah, okay.

That was the trick .

That was the trick .
I didn't really think they didsomething around this, but I
didn't know it was the players.

It seems like it would make complete sense, right
, sponsored the at-bats or thesedays, basketball players run
around with logos on theirjerseys, or it seems whatever.
This touchdown pass sponsoredby whatever.
I feel like that would be,completely relevant in 2025.

And maybe with online gambling or gaming it
could happen.
Are Wheaties even still around.
Wheaties is still around.
It is still the breakfast ofchampions.

Maybe if they had sponsored, then I would have
known.

Okay, so Oval Teens Dakota Ring campaign was
absolutely real and incrediblysuccessful.
I think what may have thrownyou guys off was the part about
the spy communications.

I didn't think that Little Orphan Annie was so
popular.

Yeah, this is the 20s and 30s.

I acted in that in my middle school play.

What were you?

I was Burt Healy, the radio man.
Hey, hobo man that song, so Ican sing it for you if you want,
but no one wants that.

All right, I thought this was genius, that
Listerine did genuinely coinhalitosis and funded
questionable medical research tocreate demand for their product
.
One of the first manufacturedhealth scares in advertising.
I thought halitosis was a realthing.
Apparently it was invented.
Okay, round two, here we go.
This is radio's golden age,1940s and 50s.

(53:51):
I found these really hard, sohere we go.
Number one cigarette companiesrequired radio hosts to smoke
their brand on air and mentionhow it quote cleared their
throat for better broadcasting.
Number two soap companies ownthe scripts of daytime dramas so
completely that characterswould pause at a mid-emotional

(54:14):
breakdown to discuss laundrydetergent.
Three Ford Motor Companysponsored news programs but
contractually forbade mention ofany traffic accidents involving
Ford vehicles.
So remember, you got to pickwhich one.
The lie is Ready.
Three, two, one.

Do we have a consensus?

Oh, again, I think we're consistent again on two.

Both with two and it's actually three.
That's a lie.
Number three is the lie.
So Ford's new censorship isfake.
While car companies did sponsornews, they didn't have explicit
contracts censoring accidentreports.
But that wasn't Iran.

It's good to know, especially as we move in this
self-driving age.

Sounds a little bit like NFL broadcast, just
knowing that you're not supposedto criticize the NFL and the
league office, but nottechnically forbidding it.
Don't bite the hand that feedsyou, type thing.

Yeah, the announcers definitely don't go
too hard on some of the playcalls yeah.

Can we just talk about how the cigarette one was
true.
That's insane.

It's nuts right.
So cigarette companiesabsolutely required on-air
smoking with health claims.
It was standard practice.

I think that's weird, that that's on radio,
right, like, what does smokingsound like on the radio?
But you know, if it wastelevision I could see, but I'm
not quite sure what that soundslike.

Yeah, they're not clearing their throat.
They're like, oh, this isbetter, so much better for me,
yeah.
And then the second one, whichyou both guessed, and you can
tell me why you guessed it Soapcompanies did their own drama
scripts and worked productplacements into emotional scenes
, hence the name Soap Operas.

I thought Soap Operas were an element of
television.

Yeah, so it was originally a radio thing, so
started there.
Let's go to round three.
This is early television 1950s,1960s.
This is number one Cigarettecompanies hired doctors to
appear in white coats on live TVrecommending their brand for
throat comfort and betterdigestion.
Number two Breakfast cerealcompanies paid families to eat

(56:30):
their products on live morningshows with hidden cameras
capturing authentic familybreakfast conversations for
later broadcast.
Three toy companies createdentire Saturday morning cartoon
shows that were essentially30-minute commercials, with the
characters and accessories thatwere in their shows being the
products that were sponsoredthroughout each episode.

(56:52):
So number one cigarettecompanies hiring Dr Sapir and
Live Coats talking about kind ofa continuation of the previous
question.
Two is breakfast cerealscapturing authentic family
breakfast conversations.
And three was toy companiesmarketing their products.

What era?
Are we in?

We are in the 50s and 60s.
Okay, ready Three, two, one.

Do we have a consensus?
I think both of us haveanswered two on both questions.

At some point you're going to get one right.

I feel like this is the SATs, where you just fill
in bubble number C all the waydown the page.

Exactly.
And guess what?

You're both right, all right.

There, it is All right.

I love it.
I love it.
We got the win All right.
Why did you choose that?

It was the hidden conversations for me.
I wasn't sure if that part ofit, that recording the hidden
conversations seemed.

I asked you about the era because I figured the
cameras would be too big to hidein your kitchen in the 1950s or
60s.
That feels like a very bigbrother TV era, modern TV era.
Ah, I see, I like that.
See how you get the camera inthere.

Yeah, I think you're right, that would be hard
to do, remember, because backthen, up until recently, those
were incandescent lamps, right,those were big, bright and doing
the sound and all that.
Yeah, so definitely makes a lotof sense.
And you're right, doing anauthentic family conversation is
not like you could just reel itin.
So here's the other ones.
Doctors did really endorsecigarettes on TV in white coats.

(58:18):
This was incredibly commonuntil the health risks became
undeniable.
It's sad.
Toy companies absolutelycreated cartoon shows as
extended commercials from alater age.
You guys might remember He-Manfrom Mattel Toys, you might
remember Transformers, star Wars, right, 70s and 80s kind of
helped promote it, but this isactually 50s and 60s, so this

(58:40):
was quite a while ago.

I remember that was a big deal and I think
regulators stopped that nowbecause it was too invasive into
kids' brains in terms of whatto buy.

Absolutely yeah.
And they started regulatingfood too, like what can appear
and stuff like that.
I think there's regulationsaround that.

Kind of like Happy Meal with the Toys right.

Yeah.

So they did do hidden camera breakfasts, but
they were real actors, not asecret family filming.
So that makes more sense.

I feel like if you recorded my family breakfast
growing up, it would be like alot of grumpy kids complaining
about being up early, going toschool and like I don't think it
would have made for good radioor television.
My house it would have been amess.

My mom used to.
Just she would just put out thecereal and we'd go figure it
out.
I don't need to wake up withyou, kids, you guys can get to
school on your own, all right,round four this is Madison
Avenue Agency Tactics 1960s and70s.
Which mind manipulationtechnique is the lie Number one?

(59:44):
Coca-cola hires psychologiststo study how different bottle
shapes affected drinkingbehavior, then redesign their
bottles to encourage fasterconsumption.
Two airlines hired motivationresearchers to determine that
passengers felt safer whenstewardesses wore specific
shades of blue thatsubconsciously reminded them of

(01:00:06):
clear skies.
And three cigarette companiesfunded studies we love cigarette
companies on subliminaladvertising and briefly
experimented with flashing theirlogos for single frames during
movie theater previews.
So this is the whole subliminaladvertising game flashing a
frame during movie theaterpreviews.
So you guys ready?

(01:00:27):
Three, two, one, give me thelie.
Oh, we have dissonance here.

Rishi Mallik (01:00:35):
Oh, we changed it up All right.

Rajiv Parikh (01:00:36):
So I see, eric, you have one finger up, one One
and Rishi, you have three.
That's right.
And guess what the answer is?

Two.

Two.

Eric Nalbone (01:00:49):
We've all guessed two, two, two.
Finally, give us two.

He didn't go with the pattern.
The airline blue uniform wasnot real and I think the part of
it was the specific shades ofblue.
So Coca-Cola did hirepsychologists to study bottle
shapes and consumption patterns.
This was part of the motivationresearch craze.
So, as you probably know, righton the cans they show the
bottle, and so that bottlewasn't just nostalgia.

(01:01:14):
They really worked on it.

I would have thought the shade of blue thing.
You know, it became a Googlething right Like test what shade
of blue works on the links, andit's such a popular anecdote
from early search that I figuredwe've been thinking about the
shade of blue for a while butyou know, that's true, that's
true, and there's probably somethought that it would have been
safer to feel that way.
How did we decide that blue waslinks?
Why not?

some other color.
Every startup now too.
I mean well, including us.
Actually, we're constantlygoing blue.
Sorry, my brand team would notbe happy about that.
I think it's some shade of teal.

Can you give us the Pantone Rishi Pantone Rishi,
pantone 75.

Going back to the Coke bottle thing, though, coke
in the bottle is so much betterthan Coke in the can, it just
tastes magical in the bottleDown.

Here in Texas they have Mexican Coke which is much
more prominent than in otherregions of the US, with the real
cane sugar which I understandis being distributed more
broadly now.
But it really is delicious andI'm our dental clients are
probably doing cartwheels over,you know, real cane sugar being
back in Coke because it's greatfor them, but it really is
something, so we've enjoyed it.

That and getting rid of fluoride in water.
I mean, our dental clients aresuper thrilled.
But you know I do overpay forMexican Coke over regular Coke.

Not an overpay, if it's worth it.

That's right.
That's right, I feel better,and that's really what it's
about.
Cigarette companies did dosubliminal advertising research
and tried to do this notion ofputting a frame, even though the
effectiveness was questionableat the time.

So what you're telling us is, if you grew up to
be a black hat SEO in like themodern era, you were a cigarette
advertiser in prior ages, right, of course?

The blue one is airlines were careful about
appearance.
Specific sky blue safetypsychology study did not happen,
but it would have been a goodone.
All right, let's go to the lastone.
Number five digital dawn.
We're both tied at one.
This is the tiebreaker, so Idon't have to like multiply the
scores One of you get it or oneof you won't or we're just going
to have fun talking about thesecrazy stories.
Number one AOL created fake chatroom celebrities or paid

(01:03:17):
employees who would hang out inpopular chat rooms and casually
mention products they loved.
Number two Amazon's originalbusiness plan included a feature
where they would track whichbooks consumers lingered on but
didn't buy, then mail physicalcatalogs featuring those exact
titles.
Number three Yahoo paid collegestudents to create fake

(01:03:40):
personal websites with hiddenaffiliate links, making it look
like regular people wererecommending those products in
their online diaries.
So AOL, fake chatroomcelebrities.
Number two Amazon sendingphysical catalogs of books
you've searched for.
Number three Yahoo doing fakepersonal websites with hidden
affiliate links.

(01:04:01):
Ready, three, two, one.
I see three for Rishi, and I'mlooking at Eric's face.
What do you want?
Okay, the answer, and thewinner is Rishi.

Yes, I would have sworn that the affiliate one was
true.
You know, based on a lot of myexperience in affiliate
marketing, I could see affiliatemarketers doing anything.

But would David Filo do that?
Probably not, but I could seeit too.

I spent too much time working on it.

The catalog thing makes sense to me.
I had a friend that ran abusiness for auto parts and
catalogs were huge for them.
I get them in the mail all thetime and I feel bad I have to
throw them in the recycling.
But apparently it's a powerfultool, especially in certain
areas of the US.
So I could totally see thatbeing part of it.
Having participated in AOL chatrooms, I could definitely tell

(01:04:57):
there were definitely people inthere that were paid to do
something.
The last one I was a little bitunsure about, but the other two
seemed right, right.

So yeah, aol absolutely used fake chat room
personalities for marketing.
As part of their earlycommunity building strategy,
amazon did track browsingbehavior and experimented with
targeted physical mail campaignsbased on online behavior.
I think that's really clever.

The precursor to personalization.

Yeah, I mean we do throw away most catalogs unless
it's personal to us, which isreally freaking hard.
But Yahoo's fake diary websitescheme was fictional.
While affiliate marketing wasemerging in the late 90s, Yahoo
didn't orchestrate fake personalwebsites with hidden affiliate
links.

I know Good for them.

But I could see, you know, eric, I mean people
are trying all kinds of stuff.

I spent a lot of time, too, trying to build
affiliate programs out ofinternational countries.
You try and go build affiliateprograms in Eastern Europe and
Russia and you see some reallyinteresting things happen.
Build affiliate programs inEastern Europe and Russia and
you see some really interestingthings happen.
We had a lot of really greatexperiences and a lot of really
great affiliates, but you alsofound the spam that it's
notorious for.

I see it's because of that Eastern European
Russian experience.

I don't know what it is about the concentration of
spammers, but it's Nigeria forspecific types of scams, Eastern
Europe for other types of scams.
Scammers got the geographicmessage really well.

They're good at it .
All right, this was great guys.
Obviously, rishi's our winnertoday.
We love having our guest win.
We just want you to have fun.
All right, here we go.
The next section are somepersonal questions, so I'm going
to run through these.
Please give me, you know, ashort answer for each one.

(01:06:35):
We'll go through themrelatively quickly.
You guys have done such a greatjob covering so many topics, so
it's great to get more infoabout who you are and what you
care about, because people inthe world want to be like you.
All right, rishi, what'ssomething you used to be really
into that you now findcompletely baffling about your
past self?

I used to in high school.
I used to be really into mixingmusic, and not the cool kind,
so it was not a DJ not workingon controllers.
This is me sort of tinkering onmy computer.
What baffles me even more thanthat is my DJ name.
I was DJ Bhima, which I can'tbelieve.
I just said that live here.

DJ Bhima, were you like the Hulk?
I mean, was that how youenvisioned yourself?

I actually just really liked BMWs back in the
day oh that Bhima?

Okay, that Bhima.
I thought you meant the IndianBhima as the second Pandava.

That would have been cooler, you know, I
wouldn't have been asembarrassed about it.
And speaking of AOL, I thinkthat was also my screen name too
at some point.

so DJ Bhima.

Yeah, how'd you spell it B-E-A-M-A.

Oh, I like that.

Yo Bima, but it's long gone now that part of my
life is over.

I love it Sounds very New Hampshire, all right.
So, eric, what's a place you'venever been to but feel like
you'd somehow belong there, andwhat draws you to it?

I would go to the geographic center of Alaska as
far away from people as I couldget and you know I'll deal with
the 24 hours of darkness and the24 hours of daylight, but I
could be just beautiful and it'dbe probably serene when it
wasn't freezing cold.
You'd just be in the middle ofnowhere.
I'd pack up and go in aheartbeat so like denali, but
with your horses yeah, you know,if we can get them in there

(01:08:21):
might be the only way to get inthere on pack horses or
something like that Horses.

You have your sled dogs.

Yeah, we get the pictures of the dogs in the
background.

Hey Rishi, how would you answer that?
What's a place that you'redrawn to?

I would also go to a remote spot.
Mine would probably have alittle bit warmer weather than
Alaska.
Eric, probably an island andfunny story.
I actually went to los roquesin venezuela years ago and I
remember 300 islands off thecoast and I remember as an
excursion they dropped us off onan island that was completely

(01:08:58):
deserted and as the boat wasgoing away I was like, hey, if
they never came back, I don'tthink I would have any way to
contact anyone because there'snobody on this island, which was
kind of scary but alsorefreshing at the same time.
So I wouldn't mind going backthere.

It sounds amazing.
I don't know if I could do that.
I'd have to be around people,so I don't last very long by
myself.

But we're even the center of New York.

Maybe that's too many people, but I need to be
around some people.
It's just a thing for me, but Ido enjoy getting out there.
When I do go to Bangalore everyquarter, I do take the team out
to some jungle-like resort orsome really unique place as
opposed to the cities, so Iguess I'm in between you guys.
All right, rishi, what's apiece of conventional wisdom

(01:09:43):
that everyone around you acceptsbut you secretly think might be
wrong?

That breakfast is the most important meal and
maybe that's a bad habit for me.
I just can't do a big breakfast.
I'm actually, you know, justthinking about what I had today
and I think I just maybe hadsome Greek yogurt.
But yeah, the the big breakfastthing has has never done it for
me.
And then it's funny because now, apparently you know, all these
like workout plundits are likeintermittent fasting.

(01:10:09):
So I've been intermittentfasting all my life,
unintentionally.

I'm with you there .
It's a cup of coffee in themorning and you know I'm off to
the races and you might see mearound lunchtime, but breakfast
is usually not my jam.

Yeah, a cup of coffee is the most important.

Yeah, I mean, the whole intermittent fasting thing
is all about the notion that itgives your body more time to
take the same energy that itwould on processing food but
working on healing your bodyright.
So there's talks back and forthon it and I've had many debates
with my father because of hisheart condition about wisdom of
it.
It's an interesting one.
I think the jury's out on it.

(01:10:43):
I don't think there's any trueanswer, although I did recently
personally add breakfast, asmall breakfast, to my meal.

Was it a bowl of Wheaties?

Bowl of Wheaties where I would watch baseball and
announce particular players.
Mine is the Starbucks egg whiteand turkey bacon breakfast.
That's my morning breakfast.
All right, Eric, if you couldbe guaranteed to be really good
at one thing you're currentlyterrible at.
What would you choose?

I'd love to be great at playing football.
I've always loved football.
It's far and away my favoritesport.
I will watch probably everysingle NFL game that is
televised this year, and RedZone for those that are
televised at the same time.

Are you really going to watch every one?

I probably will watch every single game to the
extent that Red Zone helps memanage the ones that are on at
the same time.

But okay, but you're not going to watch all of
them.
Like that's 30 games, notindependently.

My wife would kill me if I tried to watch everyone
independently, but I would try.
But I would love to be great atplaying football.
You know, I was always a smallkid and I never really liked the
idea of getting hit very hard.
But if I was great at it, youknow, then maybe that would be a
different story.

Is it the strategy or the physicality?

The strategy.
I have no interest in pain, butI have no interest in a 300
pound man flattening me, but thestrategy of it, I think, is
just really fascinating.
You know it's such acomplicated game that I love.

I always dreamt of being a running back, but then
I realized that these guys are alot bigger.

Yeah, talk about pain.

I enjoyed running through people, but that was in
touch with Bob.
All right, rishi.
If you had to describe yourpersonality using only food
metaphors, what would you be andwhy?

It would make me hungry, sushi, sushi, since I
had it last night.
It's unassuming.
It's raw, it's layered and youget a kick of wasabi at the end
I like that.

I like that.

It's got good umami it does have umami you
have a roll.

It has that, that texture feel.
You can get all kinds offlavors in it.

It's very healthy for you the food metaphor
question in my head immediatelygoes to shrek and donkey talking
about the onions and the layers.
That's where I go when I thinkfood metaphors.

Onion and layers.

Yeah, you remember that scene from Shrek where you
know Donkey is explaining toShrek or vice versa about how
you know he's like an onion andyou got to get through the
layers.

Oh, there you go.
All right, eric, what's a pieceof technology from your
childhood that you genuinelymiss and what would you bring
back if you could?

Technology.
From my childhood I would go, Ithink, with.
Do you remember the old schoolgray Game Boys?
They were like the size of abrick.
I only really had or needed twogames.
I had Tetris and I had Mario,and as a kid I'd play Tetris
endlessly.
I'd write my high score down ona sticky note and leave it for
my dad stuck on the door when Iwent to bed, because he'd be

(01:13:26):
home late and my dad would tryand beat my score overnight and
I'd do it all again the next day.
But that Game Boy and I wereabsolutely inseparable.
It's still probably bouncingaround somewhere in my parents'
house, but I'd love to find it.

For me, my era was more playing Galaga, so every
now and then I get tempted tobuy one of those arcade, the old
fashioned arcade style versionsof them.

I feel like I'm worried that if I pick it up, it
just like wouldn't live up tothe memory.
So I may as well just likeleave it be.
You know, it's like one ofthose things that you remember
from your childhood.
That's maybe better than it was, because you're a little
smaller, you're a little morenaive, but I'll just leave that
alone.

I think I have that Tetris theme song just
forever imprinted in my brainthen it speeds up and it speeds
up and you're like okay, we'rein trouble.

We're in trouble now I love that all right, rishi
, tell me the second thing youlove, not the first thing, the
second thing.

Second thing I love I love walking around or
running around a new city andexploring it, and the first
thing I love is eating my wayaround the city.
My troubles usually revolvearound food.
You're a foodie, I am a foodie,but no, I think kind of going
for walks or runs just aroundthe city, getting lost in it,

(01:14:43):
finding new things, it'sdefinitely a passion.

Do you run more or walk more?

I used to run more until my doctor told me that my
knees are getting out.
I consider myself a young guy,but I guess not so much anymore.
It's just, I guess, running onthe pavement.
It's rough on your knees, but Ilove doing it.
So they have those cool newtreadmills now, though, that
have like the screen and you canrun through different cities.
Yeah, not exactly the same, butI remember using one of those,

(01:15:11):
and it was running around LosAngeles and it was very
realistic.
They even had the homelesspeople on the side of the street
and I was like OK, I'm in LA.

I'm back home.
I love those running.
Okay, I'm in LA, I'm back home.
I know this place you need alittle too realistic.
I love those running machinestoo.
I like that feeling of beingsomewhere as opposed to just
being on the treadmill, and Ilike when they vary the hills
and all that.
That's super fun.

I was a rower in a past life, so for me it's not a
treadmill, it's the ergometers,the rowing machines and
love-hate relationship likepeople have with treadmills.
Have you rode on the charles?
Yeah, I crashed on the charles.
Actually I, in the head of thecharles, I I crashed the
princeton men's eight into theharvard or no, into the cal
men's eight ahead of us, oh jeez.
And I single-handedly preventedthe cal coach from winning his

(01:15:58):
first race.
As the coach, old princeton guywho took the cal job and I
crashed ahead of the charlesthat year into their boat and
the stroke seat.
Who's a massive German guy inthe cow boat just shakes his
head and looks at me and he goesoh Princeton, what have you
done?
Sorry, dude.
So that was the last time I wason the Charles crashing.

Was that your last row on the Charles?
They never let you.

Yeah, that was the last time we rode on the
Charles.
That was my senior year ofcollege and we didn't go up to
Boston.
That year they came to us.
But yeah, that was the lasttime I was on the Charles.

All right.
Well, you know what we'll allhave to do at some point is you
know?
We all have a common friend,sean Jacobson, that likes to do
his marathon walks through SanFrancisco, so we'll have to do
one of those.
I've done one.
Instead of 26, we did 29 milesand it's a great experience.
Cold on the day that I did it,it was cold and rainy, but it
was an amazing experience.
Great way to hang out withpeople.

That sounds awesome.
29 miles, that's a heck of awalk.

Tough.
It was a tough one, but it wasenjoyable.
So, anyways, I thank you guys.
Today we covered a reallyinteresting topic that's super
top of mind today with AEO, geo,seo, and then I love how we can
take ourselves throughmarketing history and get
ourselves to chat a little moreabout who we're about and what
we're up to.
So thank you so much for beinghere today.

Thanks, rajiv had a blast.
Thanks for having us.

That was a great conversation.
I enjoy learning so much fromour own team member, eric, who
has so many incredibleexperiences, as well as Rishi,
who really thinks about thesesorts of things deeply as part
of driving results for hiscompany, as well as all the
different companies he advises.
So this session was all abouttalking about thinking about

(01:17:45):
this whole new move towards AEO,or AI engine optimization, and
I think it's important to putthat into context.
When you're designing yourcontent and how people will see
it, you want to think about itin these four ways.
First, when they come to yourwebsite or various touch points,
how does a human see it?
What is a human experience?
Second, think about it from asearch engine perspective.

(01:18:08):
How do you set up the technicalinfrastructure and the content
such that the search engines caneasily find you and run its
magic?
Then you want to think aboutthe different weights and
measures that an AI engine, likea cloud or a chat, gpt or
perplexity, will look at yoursite.
And then, fourth, startthinking about agents.

(01:18:28):
And so imagine agents who arelooking.
A buying agent or an RFP agentwill come to your site.
How does it need to communicatewith you?
So these notions of the MCP andhow do agents communicate with
other agents will be another wayto structure your content.
So if you think about it fromthat perspective, at least have

(01:18:49):
the strategic outpost for those,you'll be able to go far.
So now, thanks for listening.
If you enjoyed today's podcast,please take a moment to rate it
and comment.
It makes a huge difference tous.
You can find us on Apple,spotify, youtube and everywhere
podcasts can be found.
The show is produced by SunleepParikh and Anand Shah,
production assistance by TarynTalley, edited by Lorne Ballant.

(01:19:13):
I'm your host, rajiv Parikh,from Position Squared, an
AI-centric growth marketingcompany based in Silicon Valley.
Come visit us at position2.com.
This has been an FNFunnyproduction.
They do a fantastic job inputting these together.
We'll catch you next time.
And remember folks be evercurious.

All Episodes

AEO Fundamentals: Moving to Content for Machines/Rishi Mallik, Eric Nalbone - SEO, Author of Truth, FAQs ~ Spark of Ages Ep 46

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

CrimeLess: Hillbilly Heist

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}AEO Fundamentals: Moving to Content for Machines/Rishi Mallik, Eric Nalbone - SEO, Author of Truth, FAQs ~ Spark of Ages Ep 46