July 6, 2025 • 19 mins
Artificial Intelligence Optimization (AIO) is fundamentally changing online visibility, with 13% of Google queries already showing AI overviews rather than traditional search results. We explore what still works, what doesn't, and how to get your content recognized and cited by AI systems in this new search landscape.
• Technical excellence remains essential – fast, accessible sites provide clear signals for AI tools
• Topic clustering and comprehensive content demonstrate authority to AI systems
• Clear writing that directly answers questions is critical as "LLMs are voracious readers but picky eaters"
• Schema markup functions as your site's "organized resume" for AI, helping it understand relationships
• FAQs and how-to formats serve as "LLM fuel" due to their modular, digestible structure
• Keyword stuffing and content created for algorithms rather than users is now "dead on arrival"
• Focus on intent clusters instead of keywords to build content ecosystems around user needs
• Target specific, scenario-based queries that match how people actually ask questions to AI
• Structure content with question-based headings followed by direct, citation-ready answers
• Use AI tools to reverse-engineer what gets cited and run visibility tests with your key terms
Start experimenting today with these AIO techniques – the brands that win will be those who understand how LLMs think and structure content to consistently earn a place in AI answer boxes.
Thanks for tuning in to Local SEO Unlocked! If you enjoyed today’s episode, don’t forget to subscribe, leave a review, and share it with others who want to master Local SEO. Stay connected with us weekly for more insights on SEO! Until next time, keep optimizing and stay ahead in local search!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome to the Deep Dive.
We're here to take that pile ofsources you've got and turn it
into clear, actionable insightsfor you.
Today, we're digging intosomething that's really changing
the game for online visibilityartificial intelligence
optimization you might hear itcalled AIO.
We've got some fascinatingmaterial lined up, including
insights from Don Phelps, andreally our mission for you today
(00:22):
is to cut through all the noise.
We wanna figure out what stillworks, what really doesn't
anymore and what you actuallyneed to do to win in this new AI
search era.
Speaker 2 (00:31):
Yeah, it's a really critical conversation to be
having right now.
I mean think about this as ofthe first quarter of 2025, over
13% of Google queries.
They're already showing AIoverviews.
Speaker 1 (00:42):
Wow, 13% already.
Speaker 2 (00:43):
Exactly so.
We're clearly moving away fromjust lists of links, those
traditional results, towardswhat some of our sources are
calling curated conversations.
Speaker 1 (00:52):
Curated conversations .
I like that.
So if you're looking to, youknow, really get your content
not just found but actuallyrecognized and maybe even cited
by AI, this deep dive isdefinitely for you.
Now, I think it's easy toassume AIO is just, you know,
seo with a bit of AI sprinkledon top.
Speaker 2 (01:08):
Right.
Speaker 1 (01:08):
But the sources we've looked at, they're pointing to
something much bigger a realstrategic mindset shift.
Speaker 2 (01:14):
That's spot on.
And what's really interestingis, yeah, AIO absolutely rewards
a lot of the SEO best practices.
We already know Things liketechnical health, good content,
but those things they're now, asone source puts it, table
stakes, not differentiators.
They're just the baseline.
You need to even play.
Speaker 1 (01:32):
Table stakes.
Okay, so what's the coredifference then?
Speaker 2 (01:35):
Well, the core difference is this In
traditional SEO, you're mostlytrying to get found in that list
of blue links.
Right In AIO, you'reessentially auditioning for a
quote.
You want the AI to pick yourcontent to feature in its answer
.
Speaker 1 (01:49):
Auditioning for a quote.
That really changes theperspective, doesn't it?
It's not just about ranking onthe page, Not anymore.
It's about getting your stuffrecognized by the AI system so
well that it actually shows upin the answer when people ask
questions.
Speaker 2 (02:00):
Precisely.
And that brings us to okay,what does still work, what are
those foundational things, thenon-negotiables that still
matter?
Speaker 1 (02:08):
Right the table stakes.
Speaker 2 (02:10):
Exactly.
Our sources point to about fivekey areas.
First up, technical excellenceyour site still needs to be fast
loading, mobile, optimized,easy for search engines.
And now AI is to crawl.
It's critical.
Speaker 1 (02:23):
Why is that still so important for AI?
Specifically?
Speaker 2 (02:26):
Well, because AI tools aren't just reading text,
they're parsing structure,understanding context.
Clean, fast sites have lessnoise, more signal for them.
They just, you know, theyprefer sites that are fast and
functional.
Makes their job easier.
Speaker 1 (02:40):
Okay, makes sense.
Remove the friction for the AI.
What's second?
Speaker 2 (02:43):
Second is topic clustering.
If you've already invested timein building out pillar pages,
topic clusters, strong internallinking, you're actually in a
good spot.
Speaker 1 (02:52):
Ah, so that work wasn't wasted.
Speaker 2 (02:54):
Not at all AI tools, especially these LLMs large
language models.
They really prefercomprehensive resources that
show topical depth.
Think of it like this the AIwants one hub, not a mess of
scattered thoughts.
It helps them see you as anauthority.
Speaker 1 (03:09):
Yeah, I can see how an AI would rather pull from one
solid source than try to piecethings together from all over
the place.
Absolutely.
Speaker 2 (03:15):
Third is clear writing that answers questions.
This sounds basic, but it's key.
Llms are described as voraciousreaders, but also picky eaters.
Content that's genuinelywritten well, aiming to teach or
explain something clearly to ahuman.
That works just as well for AI.
Speaker 1 (03:31):
So good writing is good writing, whether for a
person or an AI.
Speaker 2 (03:35):
Pretty much.
If a human can easily get theanswer, chances are an LLM can
too.
Fourth schema markup Structureddata.
Speaker 1 (03:42):
Ah, schema Always important, but maybe more so now
.
Speaker 2 (03:46):
Definitely more so.
It's foundational.
Think of it like your site'shighly organized resume for the
AI Schema helps LLMs understandwhat your content is about, not
just the words it contains.
One source called it one of theclearest signals LLMs rely on.
It helps them map outrelationships between things.
Speaker 1 (04:05):
So you're basically giving the AI a clear index
telling it this is a recipe,this is a product, this is a
how-to guide.
Speaker 2 (04:12):
That's a perfect way to put it.
And finally, number five FAQs,how-to, guides and lists these
formats.
They're like LLM fuel.
Speaker 1 (04:20):
Why they're specifically.
Speaker 2 (04:22):
Because they're naturally modular.
They're digestible.
Ai tools tend to lean heavilyon these formats when they need
to summarize information or pullout specific points or steps to
cite.
So if you're already creatinganswer-ready content like that,
you're basically alreadyspeaking the language of AIO.
Speaker 1 (04:35):
Okay, so those are the foundations that still hold
strong.
Let's flip it then.
What about the tactics thatmaybe they worked a few years
ago, but now are just wellineffective, or maybe even
invisible to AI?
Speaker 2 (04:46):
Right, and this isn't about you know, throwing out
your entire playbook, but it'sreally crucial to know which
Flays just aren't going to workanymore.
First big one Keyword stuffingwithout intent.
Llms are smart.
They aren't fooled by just likea pile of terms duct taped to a
blog post, as one sourcecolorfully put it.
Okay, If your content onlyexists to try and rank for a
(05:07):
term and not to actually helpsomeone, the AI is not buying it
.
Speaker 1 (05:11):
So that whole era of writing thin content just to hit
a keyword density number that'sreally over now.
Oh, absolutely.
Speaker 2 (05:19):
Which leads to the second point content for the
algorithm, not the user.
Remember those like 500 wordfluff pieces targeting best
email platform 2024?
Speaker 1 (05:28):
Yeah, we've all seen them.
Speaker 2 (05:29):
Our sources are pretty blunt and they're DOA and
AIO dead on arrival.
Ai search tools are activelylooking for the most complete,
helpful and well-structuredcontent, not just keyword filler
.
Speaker 1 (05:39):
That feels like a significant shift, even from
just a couple of years back inSEO.
Speaker 2 (05:43):
It really is.
Third thing that's lesseffective chasing head terms,
you know, trying to rank forthose huge broad terms like CRM
or SEO the old holy grail,Exactly.
But AI tools now, they muchprefer long tail contextual
queries.
Something like best CRM forsolo founders with no sales team
.
Speaker 1 (06:03):
Yeah.
Speaker 2 (06:04):
The broader the term, the harder it is for the AI to
give a truly useful, specificanswer.
So it leans towards specificity.
Speaker 1 (06:12):
So it's less about casting that super wide net and
much more about precisiontargeting.
Speaker 2 (06:16):
Precisely Number four blind backlink chasing.
Now, backlinks aren't totallyirrelevant, let's be clear.
Speaker 1 (06:22):
Okay.
Speaker 2 (06:23):
But AI search seems to be more interested in what
you say, how clearly you say it,and whether other credible
sources actually reference andvalidate it.
Authority isn't just a numbersgame of how many links you have
anymore.
It's more about demonstratedexpertise and genuine trust
signals.
Speaker 1 (06:39):
That's a big change.
It puts way more emphasis backon the actual quality and
relevance of the content itself.
Speaker 2 (06:45):
It really does.
And the fifth point here isignoring tone and context.
This is interesting.
Lms don't just read the words,they analyze how things are said
.
If your brand's tone comesacross as like robotic or the
messaging feels off for theuser's actual question, you
might not just get skipped over.
Speaker 1 (07:01):
You could get misquoted.
Speaker 2 (07:02):
Yeah, or maybe cited in a context that doesn't
actually help your brand.
The AI might misunderstand ormisrepresent your point if the
tone isn't right.
Speaker 1 (07:10):
OK, so if those are the shifts and the things to
avoid, what does it actuallytake to show up effectively, to
show up well, in this AIdiscovery layer?
What's the new playbook looklike?
Speaker 2 (07:22):
Yeah, this is the crucial part, isn't it?
How do we adapt for what onesource called user empathy at
scale?
The first rule in this newplaybook seems to be thinking
intent clusters, not justkeyword clusters.
Speaker 1 (07:32):
Intent clusters.
Okay, what does that meanexactly?
Speaker 2 (07:35):
It means forgetting about just stuffing keywords.
Aio needs a much deeper diveinto why someone is asking a
question, what's theirunderlying goal, and then you
build whole content ecosystemsaround those intents.
So for A-B testing you don'tjust have one page.
You'd have say what is A-Btesting?
Then maybe how to run an A-Btest in Klaviyo, a-b testing
(07:55):
ideas for e-commerce, maybe evenwhat to do when your A-B test
fails.
Speaker 1 (08:02):
Each piece tackles a specific angle of that core
intent.
That makes sense.
You're providing comprehensivecoverage without making one page
impossibly long, and you'resignaling to the AI hey, we
really know this topic insideand out.
Speaker 2 (08:12):
Exactly Builds authority.
Second tactic target long tailscenario-based queries.
Generic is out, context isdefinitely in.
So instead of just emailmarketing tips, you'd aim for
something much more specific,like how do I improve open rates
in Gmail versus Outlook,specifically for B2B emails?
Speaker 1 (08:31):
Wow, that's ultra specific.
Speaker 2 (08:33):
It is and it sounds exactly like something a real
person would actually type orask an AI.
Right, You're meeting a veryprecise need.
Speaker 1 (08:39):
Yeah, you're anticipating how someone would
genuinely ask for help in a realsituation.
Speaker 2 (08:43):
Precisely.
Third, use headings asquestions and answer them
clearly.
Think modular LLM skim.
They look for structure whenyour H2 heading directly matches
a likely user query like how doI increase my email open rate
and outlook?
Speaker 1 (08:57):
Instead of something vague like improve your email
strategy.
Speaker 2 (09:00):
Right and then the paragraph right below it gives a
clear, concise answer.
You make it incredibly easy forthe AI to just grab that
section and cite you.
Speaker 1 (09:07):
Get straight to the point for the AI, but that also
makes it clear for the humanreader too.
Win-win.
Speaker 2 (09:12):
Absolutely.
Fourth point structure contentfor AI tools.
This is where formatting reallymatters.
Things like bullet points,numbered lists, step-by-step
instructions, good internallinking LLMs apparently eat this
stuff up.
Also include clear definitions,good examples, maybe subtle
clarifiers.
These things make your contentvery quotable for an AI.
Speaker 1 (09:35):
So it's not just the writing, it's how you present
the information, making it superdigestible, AI friendly.
Speaker 2 (09:40):
That's a huge part of it.
And fifth, maybe the mostimportant account for the human
behind the query, when youreally optimize for the user's
actual mindset, you know, arethey short on time, are they
budget conscious, are theyskeptical?
When you address that oursources suggest, you don't just
rank, you resonate, you buildactual trust.
Speaker 1 (09:58):
Okay, resonate, I like that.
Now you mentioned AI tools.
Think differently.
This is where it gets reallyinteresting, right, how search
intent itself is shifting.
Speaker 2 (10:05):
Yeah, this is a big one.
Ai doesn't just answer faster,it approaches the type of answer
differently.
We used to talk abouttraditional search like a buffet
of links.
Right, you pick what looks good.
Ai search is more like gettingone plate, one answer curated
for you.
One plate, okay.
So how does that change intent?
(10:26):
Well, the classic intent typesinformational, transactional,
navigational.
They're still relevant, yeah,but our sources are really
emphasizing two additional typesthat seem to carry a lot more
weight for LLMs instructionaland scenario-based.
Speaker 1 (10:39):
Instructional and scenario-based.
Why are those so important now?
Speaker 2 (10:42):
Because they signal really high intent.
Speaker 1 (10:44):
Yeah.
Speaker 2 (10:44):
And they provide rich context.
Someone asking an instructionalhow do I?
Or scenario-based what's thebest?
X for Y situation query they'reusually ready to do something
or make a decision.
That's where the money ispotentially.
Speaker 1 (10:56):
And there's data backing this up, isn't there
Something from SEMrush?
Speaker 2 (11:00):
Yeah, exactly Our sources mentioned.
Semrush found that whenGoogle's AI overviews, search,
gpt or active standardinformational queries, plummeted
from like 81% down to about 39%and get this, 70% of queries
suddenly fell into this unknownbucket because they didn't fit
the old models.
Neatly, 70%.
But that unknown bucket, that'snot a problem, it's the
(11:22):
opportunity.
It's full of those neurons,context-rich, help-me-decide
type questions that AI isperfectly suited to answer.
Speaker 1 (11:29):
Okay, so this changes how we need to approach
optimization, which leads us tothe AIO process, itself
described as speed dating withrobots.
Speaker 2 (11:38):
Huh yeah, that's a memorable way to put it, because
you really might only get onequick shot to make an impression
and get cited.
So the process needs to besharp.
Step one do smart queryresearch.
Forget just looking at keywordvolume.
You need to understand how realpeople actually ask questions
in AI chat interfaces.
Speaker 1 (11:53):
How do you do that effectively?
Speaker 2 (11:54):
Well, one great pro tip from the sources is actually
ask to chat, GPT or similartool.
Ask it.
How would a beginner ask aboutyour topic?
You get instant insight intonatural language queries.
Speaker 1 (12:04):
Oh, that's clever.
Use the AI to understand howpeople use the AI.
Speaker 2 (12:08):
Exactly Step two.
Map intent like a human Reallyfigure out.
Is this person trying to learnsomething new?
Compare options fix a problem.
Are they ready to buy?
The examples given were great.
What is schema markup?
That's clearly informational.
Schema markup versus meta tagsthat's consideration.
How to add schema in Shopifydefinitely instructional.
(12:28):
Best plugin for schema markup.
They're likely purchase ready,knowing that intent shapes
everything.
Speaker 1 (12:34):
Okay, mapping the real goal.
Speaker 2 (12:35):
Right Step three craft clear, modular answers.
This is critical.
Write in a way that an AI canbasically copy paste your answer
without needing to rewrite it.
That means one question, oneclear answer, short paragraphs.
Use those bullets and numberedsteps, and absolutely no
throat-clearing intros or cleverdetours.
Get straight to the point.
You're not telling a story,you're solving the user's
(12:56):
immediate need.
Speaker 1 (12:57):
Direct concise structured yes.
Speaker 2 (13:00):
Step four Monitor and improve.
Aio isn't something you set uponce and walk away from.
You have to track it.
Are you getting cited in LLMresponses?
What's the tone like when youare?
Where are your competitorsshowing up?
Speaker 1 (13:13):
Active monitoring is key.
Speaker 2 (13:14):
Absolutely.
Step five optimize the tech.
We mentioned technicalexcellence earlier Page speed,
mobile crawlability, schema theyall still matter, but the focus
shifts slightly.
It's about how easily machinescan interpret your page
specifically in the context of aquery.
And step six structure like amachine learner would want.
Speaker 1 (13:34):
Structure like a machine learner?
Okay, yeah.
Speaker 2 (13:35):
LLMs don't browse visually like humans.
They parse data to make itsuper easy for them.
Use headings that are fullquestions.
Include definitions andexamples.
Clearly Use specific schematypes.
How-to QI page article thatexplicitly tell the AI the
purpose of your content.
Break answers into logicalchunks.
It's all about making itmachine readable.
The sources stress this wholeprocess requires discipline.
Speaker 1 (13:55):
Discipline.
Okay, that makes sense.
Now, doing all this, itprobably requires different
tools than we might have usedfor traditional SEO.
Right, this isn't a game fortools built back in 2015.
What are some of theinteresting players emerging in
this AIO space?
Speaker 2 (14:10):
Yeah, the tool landscape is definitely evolving
.
Our sources mentioned a fewinteresting ones.
There's Peak.
Speaker 1 (14:14):
Peak.
Speaker 2 (14:15):
Yeah, P-E-C.
Apparently, it helps you seeexactly where and how your brand
is getting mentioned in AIanswers across ChatGPT, Gemini,
Perplexity, those kinds ofplatforms.
It basically turns that AIblind spot into a measurable
scoreboard.
Speaker 1 (14:29):
Okay, visibility tracking for AI answers.
That sounds useful.
Speaker 2 (14:36):
For sure.
Then there's Athena.
This one seems focused oncompetitive intelligence within
AI search, so seeing who'sreally owning which types of
queries in the AI results.
Speaker 1 (14:42):
Competitor analysis for AI.
Oh God.
Speaker 2 (14:44):
Goodie.
This tool is designed to helpmarketers actually optimize
their content for AI results.
Competitor analysis for AI ohgood Goodie.
This tool is designed to helpmarketers actually optimize
their content for AI engines.
It looks at things like tone,structure, scheming, use and
gives recommendations based onhow AI seem to be behaving.
Speaker 1 (14:55):
So more on the content optimization side.
Speaker 2 (14:58):
Exactly.
And then, of course, you haveestablished players evolving.
The Semrush AI toolkit wasmentioned.
It's extending theirtraditional SEO tools into this
new territory, things liketracking your AI visibility,
analyzing the sentiment aroundyour brand in AI mentions.
Speaker 1 (15:14):
Sentiment Like is the AI describing your brand as
expensive or easy to use?
Speaker 2 (15:20):
Precisely that kind of thing, Plus suggesting topics
tailored specifically for thoseAI-style long-tail queries we
talked about.
Speaker 1 (15:26):
Okay, so there are tools emerging.
It sounds like the key takeawayhere is, like one source said,
the best tool is the one thatcloses your current gap.
Whatever you need most rightnow.
That seems right, but the worstmove is just ignoring this
whole AI layer completely.
Speaker 2 (15:40):
Yeah, ignoring it is definitely the only wrong move.
This AI discovery layer it'salready shaping how people
perceive brands and makepurchase decisions.
Action is really needed now.
Speaker 1 (15:50):
So let's bring it home.
What are the concrete strategicmoves people listening can
start making, like today or thisweek, to actually begin winning
in AIO?
Speaker 2 (16:00):
Okay, actionable steps.
First, shift your strategy tointent clusters.
Stop thinking article byarticle.
Start thinking ecosystem.
Build those networks of contentaround core user needs, Like
the example pillar topic what ismarketing attribution Supported
by best tools for B2Battribution?
How to explain attribution toyour CFO?
(16:21):
Common attribution mistakes.
Doing that trains the AI to seeyou as the comprehensive,
trustworthy source.
Speaker 1 (16:29):
Building that web of expertise makes sense.
Speaker 2 (16:31):
Second, use AI tools to reverse engineer AI behavior.
Be proactive.
Go into ChatGPT or Geminiyourself.
Type in your key terms, yourcustomer questions, see who gets
cited.
Look at the structure of thecontent that gets pulled.
Then go back and improve yourown stuff based on that recon.
Speaker 1 (16:46):
Use the AIs to learn from the AIs.
Speaker 2 (16:48):
Exactly.
Third, get specific, get real.
Generic just doesn't cut itanymore.
Don't write how to increasewebsite traffic.
Write how to drive traffic to aShopify store without using
paid ads.
Loms seem to prefer answersthat sound like they were
written for a person, not for aplaybook.
That specificity signalsrelevance.
Speaker 1 (17:06):
It feels more authentic too, like you're
answering a real question.
Speaker 2 (17:09):
It does.
Fourth, and this is a quick win, run a five-minute AI
visibility test.
Seriously, just take fiveminutes.
Type your brand name, your mainproduct, your top competitors,
maybe your top three customerquestions into chat.
Gpt, gemini perplexity.
See what comes up.
Are you mentioned?
What's the tone?
Who is showing up?
The sources are clear.
If you're not showing up, thework starts now.
Speaker 1 (17:32):
That's a really simple eye-opening exercise
anyone can do right away.
Speaker 2 (17:41):
Totally.
And fifth pilot measure repeatDon't feel like you have to
overhaul everything overnight.
Start small, pick maybe fivekey pages.
Restructure them using this AIOprocess.
Add the right schema, rewritethe headings as questions, use
bullet points, then monitor ifthey start getting cited more in
AI tools.
If it works, scale it up.
If not, adjust your approach.
The key is, as the source said,but don't wait.
Speaker 1 (17:59):
Don't wait, start experimenting now.
Pilot measure repeat Okay, sowrapping this all up.
What does this really mean foryou, the listener?
It sounds like AIO isn't justsome side quest or passing fad.
Speaker 2 (18:11):
Not at all.
It really is the next evolutionof SEO.
It's where things are headed.
Speaker 1 (18:15):
And the brands that are going to succeed.
Speaker 2 (18:17):
The brands that win will be the ones who take the
time to understand how theseLLMs actually think, what real
users are truly asking, and thenfigure out how to structure
their content so it consistentlyearns its place in that AI
answer box.
Speaker 1 (18:32):
Not just landing somewhere in the link pile,
exactly Getting into the curatedanswer.
So you don't need to throw outeverything you've learned about
SEO and online visibility.
Speaker 2 (18:40):
No, definitely not.
The foundations are stillimportant.
Speaker 1 (18:42):
But you absolutely have to evolve how you apply
that knowledge in this newcontext and the payoff if you
get it right.
Our sources suggest you don'tjust show up.
Speaker 2 (18:52):
You get quoted, you get trusted, you get chosen.
It's about becoming the AI'spreferred source, which
ultimately means becoming theuser's preferred answer.
Speaker 1 (19:02):
That's powerful, get quoted, trusted, chosen.
Speaker 2 (19:05):
Which leaves, I think , one important question for
everyone listening to considerAre you ready to make AI work
for your brand and not the otherway around?
Are you ready to adapt?
Welcome to the Deep Dive.
We're here to take that pile ofsources you've got and turn it
into clear, actionable insightsfor you.
Today, we're digging intosomething that's really changing
the game for online visibilityartificial intelligence
optimization you might hear itcalled AIO.
We've got some fascinatingmaterial lined up, including
insights from Don Phelps, andreally our mission for you today
(00:22):
is to cut through all the noise.
We wanna figure out what stillworks, what really doesn't
anymore and what you actuallyneed to do to win in this new AI
search era.
Speaker 2 (00:31):
Yeah, it's a really critical conversation to be
having right now.
I mean think about this as ofthe first quarter of 2025, over
13% of Google queries.
They're already showing AIoverviews.
Speaker 1 (00:42):
Wow, 13% already.
Speaker 2 (00:43):
Exactly so.
We're clearly moving away fromjust lists of links, those
traditional results, towardswhat some of our sources are
calling curated conversations.
Speaker 1 (00:52):
Curated conversations .
I like that.
So if you're looking to, youknow, really get your content
not just found but actuallyrecognized and maybe even cited
by AI, this deep dive isdefinitely for you.
Now, I think it's easy toassume AIO is just, you know,
seo with a bit of AI sprinkledon top.
Speaker 2 (01:08):
Right.
Speaker 1 (01:08):
But the sources we've looked at, they're pointing to
something much bigger a realstrategic mindset shift.
Speaker 2 (01:14):
That's spot on.
And what's really interestingis, yeah, AIO absolutely rewards
a lot of the SEO best practices.
We already know Things liketechnical health, good content,
but those things they're now, asone source puts it, table
stakes, not differentiators.
They're just the baseline.
You need to even play.
Speaker 1 (01:32):
Table stakes.
Okay, so what's the coredifference then?
Speaker 2 (01:35):
Well, the core difference is this In
traditional SEO, you're mostlytrying to get found in that list
of blue links.
Right In AIO, you'reessentially auditioning for a
quote.
You want the AI to pick yourcontent to feature in its answer
.
Speaker 1 (01:49):
Auditioning for a quote.
That really changes theperspective, doesn't it?
It's not just about ranking onthe page, Not anymore.
It's about getting your stuffrecognized by the AI system so
well that it actually shows upin the answer when people ask
questions.
Speaker 2 (02:00):
Precisely.
And that brings us to okay,what does still work, what are
those foundational things, thenon-negotiables that still
matter?
Speaker 1 (02:08):
Right the table stakes.
Speaker 2 (02:10):
Exactly.
Our sources point to about fivekey areas.
First up, technical excellenceyour site still needs to be fast
loading, mobile, optimized,easy for search engines.
And now AI is to crawl.
It's critical.
Speaker 1 (02:23):
Why is that still so important for AI?
Specifically?
Speaker 2 (02:26):
Well, because AI tools aren't just reading text,
they're parsing structure,understanding context.
Clean, fast sites have lessnoise, more signal for them.
They just, you know, theyprefer sites that are fast and
functional.
Makes their job easier.
Speaker 1 (02:40):
Okay, makes sense.
Remove the friction for the AI.
What's second?
Speaker 2 (02:43):
Second is topic clustering.
If you've already invested timein building out pillar pages,
topic clusters, strong internallinking, you're actually in a
good spot.
Speaker 1 (02:52):
Ah, so that work wasn't wasted.
Speaker 2 (02:54):
Not at all AI tools, especially these LLMs large
language models.
They really prefercomprehensive resources that
show topical depth.
Think of it like this the AIwants one hub, not a mess of
scattered thoughts.
It helps them see you as anauthority.
Speaker 1 (03:09):
Yeah, I can see how an AI would rather pull from one
solid source than try to piecethings together from all over
the place.
Absolutely.
Speaker 2 (03:15):
Third is clear writing that answers questions.
This sounds basic, but it's key.
Llms are described as voraciousreaders, but also picky eaters.
Content that's genuinelywritten well, aiming to teach or
explain something clearly to ahuman.
That works just as well for AI.
Speaker 1 (03:31):
So good writing is good writing, whether for a
person or an AI.
Speaker 2 (03:35):
Pretty much.
If a human can easily get theanswer, chances are an LLM can
too.
Fourth schema markup Structureddata.
Speaker 1 (03:42):
Ah, schema Always important, but maybe more so now
.
Speaker 2 (03:46):
Definitely more so.
It's foundational.
Think of it like your site'shighly organized resume for the
AI Schema helps LLMs understandwhat your content is about, not
just the words it contains.
One source called it one of theclearest signals LLMs rely on.
It helps them map outrelationships between things.
Speaker 1 (04:05):
So you're basically giving the AI a clear index
telling it this is a recipe,this is a product, this is a
how-to guide.
Speaker 2 (04:12):
That's a perfect way to put it.
And finally, number five FAQs,how-to, guides and lists these
formats.
They're like LLM fuel.
Speaker 1 (04:20):
Why they're specifically.
Speaker 2 (04:22):
Because they're naturally modular.
They're digestible.
Ai tools tend to lean heavilyon these formats when they need
to summarize information or pullout specific points or steps to
cite.
So if you're already creatinganswer-ready content like that,
you're basically alreadyspeaking the language of AIO.
Speaker 1 (04:35):
Okay, so those are the foundations that still hold
strong.
Let's flip it then.
What about the tactics thatmaybe they worked a few years
ago, but now are just wellineffective, or maybe even
invisible to AI?
Speaker 2 (04:46):
Right, and this isn't about you know, throwing out
your entire playbook, but it'sreally crucial to know which
Flays just aren't going to workanymore.
First big one Keyword stuffingwithout intent.
Llms are smart.
They aren't fooled by just likea pile of terms duct taped to a
blog post, as one sourcecolorfully put it.
Okay, If your content onlyexists to try and rank for a
(05:07):
term and not to actually helpsomeone, the AI is not buying it
.
Speaker 1 (05:11):
So that whole era of writing thin content just to hit
a keyword density number that'sreally over now.
Oh, absolutely.
Speaker 2 (05:19):
Which leads to the second point content for the
algorithm, not the user.
Remember those like 500 wordfluff pieces targeting best
email platform 2024?
Speaker 1 (05:28):
Yeah, we've all seen them.
Speaker 2 (05:29):
Our sources are pretty blunt and they're DOA and
AIO dead on arrival.
Ai search tools are activelylooking for the most complete,
helpful and well-structuredcontent, not just keyword filler
.
Speaker 1 (05:39):
That feels like a significant shift, even from
just a couple of years back inSEO.
Speaker 2 (05:43):
It really is.
Third thing that's lesseffective chasing head terms,
you know, trying to rank forthose huge broad terms like CRM
or SEO the old holy grail,Exactly.
But AI tools now, they muchprefer long tail contextual
queries.
Something like best CRM forsolo founders with no sales team
.
Speaker 1 (06:03):
Yeah.
Speaker 2 (06:04):
The broader the term, the harder it is for the AI to
give a truly useful, specificanswer.
So it leans towards specificity.
Speaker 1 (06:12):
So it's less about casting that super wide net and
much more about precisiontargeting.
Speaker 2 (06:16):
Precisely Number four blind backlink chasing.
Now, backlinks aren't totallyirrelevant, let's be clear.
Speaker 1 (06:22):
Okay.
Speaker 2 (06:23):
But AI search seems to be more interested in what
you say, how clearly you say it,and whether other credible
sources actually reference andvalidate it.
Authority isn't just a numbersgame of how many links you have
anymore.
It's more about demonstratedexpertise and genuine trust
signals.
Speaker 1 (06:39):
That's a big change.
It puts way more emphasis backon the actual quality and
relevance of the content itself.
Speaker 2 (06:45):
It really does.
And the fifth point here isignoring tone and context.
This is interesting.
Lms don't just read the words,they analyze how things are said
.
If your brand's tone comesacross as like robotic or the
messaging feels off for theuser's actual question, you
might not just get skipped over.
Speaker 1 (07:01):
You could get misquoted.
Speaker 2 (07:02):
Yeah, or maybe cited in a context that doesn't
actually help your brand.
The AI might misunderstand ormisrepresent your point if the
tone isn't right.
Speaker 1 (07:10):
OK, so if those are the shifts and the things to
avoid, what does it actuallytake to show up effectively, to
show up well, in this AIdiscovery layer?
What's the new playbook looklike?
Speaker 2 (07:22):
Yeah, this is the crucial part, isn't it?
How do we adapt for what onesource called user empathy at
scale?
The first rule in this newplaybook seems to be thinking
intent clusters, not justkeyword clusters.
Speaker 1 (07:32):
Intent clusters.
Okay, what does that meanexactly?
Speaker 2 (07:35):
It means forgetting about just stuffing keywords.
Aio needs a much deeper diveinto why someone is asking a
question, what's theirunderlying goal, and then you
build whole content ecosystemsaround those intents.
So for A-B testing you don'tjust have one page.
You'd have say what is A-Btesting?
Then maybe how to run an A-Btest in Klaviyo, a-b testing
(07:55):
ideas for e-commerce, maybe evenwhat to do when your A-B test
fails.
Speaker 1 (08:02):
Each piece tackles a specific angle of that core
intent.
That makes sense.
You're providing comprehensivecoverage without making one page
impossibly long, and you'resignaling to the AI hey, we
really know this topic insideand out.
Speaker 2 (08:12):
Exactly Builds authority.
Second tactic target long tailscenario-based queries.
Generic is out, context isdefinitely in.
So instead of just emailmarketing tips, you'd aim for
something much more specific,like how do I improve open rates
in Gmail versus Outlook,specifically for B2B emails?
Speaker 1 (08:31):
Wow, that's ultra specific.
Speaker 2 (08:33):
It is and it sounds exactly like something a real
person would actually type orask an AI.
Right, You're meeting a veryprecise need.
Speaker 1 (08:39):
Yeah, you're anticipating how someone would
genuinely ask for help in a realsituation.
Speaker 2 (08:43):
Precisely.
Third, use headings asquestions and answer them
clearly.
Think modular LLM skim.
They look for structure whenyour H2 heading directly matches
a likely user query like how doI increase my email open rate
and outlook?
Speaker 1 (08:57):
Instead of something vague like improve your email
strategy.
Speaker 2 (09:00):
Right and then the paragraph right below it gives a
clear, concise answer.
You make it incredibly easy forthe AI to just grab that
section and cite you.
Speaker 1 (09:07):
Get straight to the point for the AI, but that also
makes it clear for the humanreader too.
Win-win.
Speaker 2 (09:12):
Absolutely.
Fourth point structure contentfor AI tools.
This is where formatting reallymatters.
Things like bullet points,numbered lists, step-by-step
instructions, good internallinking LLMs apparently eat this
stuff up.
Also include clear definitions,good examples, maybe subtle
clarifiers.
These things make your contentvery quotable for an AI.
Speaker 1 (09:35):
So it's not just the writing, it's how you present
the information, making it superdigestible, AI friendly.
Speaker 2 (09:40):
That's a huge part of it.
And fifth, maybe the mostimportant account for the human
behind the query, when youreally optimize for the user's
actual mindset, you know, arethey short on time, are they
budget conscious, are theyskeptical?
When you address that oursources suggest, you don't just
rank, you resonate, you buildactual trust.
Speaker 1 (09:58):
Okay, resonate, I like that.
Now you mentioned AI tools.
Think differently.
This is where it gets reallyinteresting, right, how search
intent itself is shifting.
Speaker 2 (10:05):
Yeah, this is a big one.
Ai doesn't just answer faster,it approaches the type of answer
differently.
We used to talk abouttraditional search like a buffet
of links.
Right, you pick what looks good.
Ai search is more like gettingone plate, one answer curated
for you.
One plate, okay.
So how does that change intent?
(10:26):
Well, the classic intent typesinformational, transactional,
navigational.
They're still relevant, yeah,but our sources are really
emphasizing two additional typesthat seem to carry a lot more
weight for LLMs instructionaland scenario-based.
Speaker 1 (10:39):
Instructional and scenario-based.
Why are those so important now?
Speaker 2 (10:42):
Because they signal really high intent.
Speaker 1 (10:44):
Yeah.
Speaker 2 (10:44):
And they provide rich context.
Someone asking an instructionalhow do I?
Or scenario-based what's thebest?
X for Y situation query they'reusually ready to do something
or make a decision.
That's where the money ispotentially.
Speaker 1 (10:56):
And there's data backing this up, isn't there
Something from SEMrush?
Speaker 2 (11:00):
Yeah, exactly Our sources mentioned.
Semrush found that whenGoogle's AI overviews, search,
gpt or active standardinformational queries, plummeted
from like 81% down to about 39%and get this, 70% of queries
suddenly fell into this unknownbucket because they didn't fit
the old models.
Neatly, 70%.
But that unknown bucket, that'snot a problem, it's the
(11:22):
opportunity.
It's full of those neurons,context-rich, help-me-decide
type questions that AI isperfectly suited to answer.
Speaker 1 (11:29):
Okay, so this changes how we need to approach
optimization, which leads us tothe AIO process, itself
described as speed dating withrobots.
Speaker 2 (11:38):
Huh yeah, that's a memorable way to put it, because
you really might only get onequick shot to make an impression
and get cited.
So the process needs to besharp.
Step one do smart queryresearch.
Forget just looking at keywordvolume.
You need to understand how realpeople actually ask questions
in AI chat interfaces.
Speaker 1 (11:53):
How do you do that effectively?
Speaker 2 (11:54):
Well, one great pro tip from the sources is actually
ask to chat, GPT or similartool.
Ask it.
How would a beginner ask aboutyour topic?
You get instant insight intonatural language queries.
Speaker 1 (12:04):
Oh, that's clever.
Use the AI to understand howpeople use the AI.
Speaker 2 (12:08):
Exactly Step two.
Map intent like a human Reallyfigure out.
Is this person trying to learnsomething new?
Compare options fix a problem.
Are they ready to buy?
The examples given were great.
What is schema markup?
That's clearly informational.
Schema markup versus meta tagsthat's consideration.
How to add schema in Shopifydefinitely instructional.
(12:28):
Best plugin for schema markup.
They're likely purchase ready,knowing that intent shapes
everything.
Speaker 1 (12:34):
Okay, mapping the real goal.
Speaker 2 (12:35):
Right Step three craft clear, modular answers.
This is critical.
Write in a way that an AI canbasically copy paste your answer
without needing to rewrite it.
That means one question, oneclear answer, short paragraphs.
Use those bullets and numberedsteps, and absolutely no
throat-clearing intros or cleverdetours.
Get straight to the point.
You're not telling a story,you're solving the user's
(12:56):
immediate need.
Speaker 1 (12:57):
Direct concise structured yes.
Speaker 2 (13:00):
Step four Monitor and improve.
Aio isn't something you set uponce and walk away from.
You have to track it.
Are you getting cited in LLMresponses?
What's the tone like when youare?
Where are your competitorsshowing up?
Speaker 1 (13:13):
Active monitoring is key.
Speaker 2 (13:14):
Absolutely.
Step five optimize the tech.
We mentioned technicalexcellence earlier Page speed,
mobile crawlability, schema theyall still matter, but the focus
shifts slightly.
It's about how easily machinescan interpret your page
specifically in the context of aquery.
And step six structure like amachine learner would want.
Speaker 1 (13:34):
Structure like a machine learner?
Okay, yeah.
Speaker 2 (13:35):
LLMs don't browse visually like humans.
They parse data to make itsuper easy for them.
Use headings that are fullquestions.
Include definitions andexamples.
Clearly Use specific schematypes.
How-to QI page article thatexplicitly tell the AI the
purpose of your content.
Break answers into logicalchunks.
It's all about making itmachine readable.
The sources stress this wholeprocess requires discipline.
Speaker 1 (13:55):
Discipline.
Okay, that makes sense.
Now, doing all this, itprobably requires different
tools than we might have usedfor traditional SEO.
Right, this isn't a game fortools built back in 2015.
What are some of theinteresting players emerging in
this AIO space?
Speaker 2 (14:10):
Yeah, the tool landscape is definitely evolving
.
Our sources mentioned a fewinteresting ones.
There's Peak.
Speaker 1 (14:14):
Peak.
Speaker 2 (14:15):
Yeah, P-E-C.
Apparently, it helps you seeexactly where and how your brand
is getting mentioned in AIanswers across ChatGPT, Gemini,
Perplexity, those kinds ofplatforms.
It basically turns that AIblind spot into a measurable
scoreboard.
Speaker 1 (14:29):
Okay, visibility tracking for AI answers.
That sounds useful.
Speaker 2 (14:36):
For sure.
Then there's Athena.
This one seems focused oncompetitive intelligence within
AI search, so seeing who'sreally owning which types of
queries in the AI results.
Speaker 1 (14:42):
Competitor analysis for AI.
Oh God.
Speaker 2 (14:44):
Goodie.
This tool is designed to helpmarketers actually optimize
their content for AI results.
Competitor analysis for AI ohgood Goodie.
This tool is designed to helpmarketers actually optimize
their content for AI engines.
It looks at things like tone,structure, scheming, use and
gives recommendations based onhow AI seem to be behaving.
Speaker 1 (14:55):
So more on the content optimization side.
Speaker 2 (14:58):
Exactly.
And then, of course, you haveestablished players evolving.
The Semrush AI toolkit wasmentioned.
It's extending theirtraditional SEO tools into this
new territory, things liketracking your AI visibility,
analyzing the sentiment aroundyour brand in AI mentions.
Speaker 1 (15:14):
Sentiment Like is the AI describing your brand as
expensive or easy to use?
Speaker 2 (15:20):
Precisely that kind of thing, Plus suggesting topics
tailored specifically for thoseAI-style long-tail queries we
talked about.
Speaker 1 (15:26):
Okay, so there are tools emerging.
It sounds like the key takeawayhere is, like one source said,
the best tool is the one thatcloses your current gap.
Whatever you need most rightnow.
That seems right, but the worstmove is just ignoring this
whole AI layer completely.
Speaker 2 (15:40):
Yeah, ignoring it is definitely the only wrong move.
This AI discovery layer it'salready shaping how people
perceive brands and makepurchase decisions.
Action is really needed now.
Speaker 1 (15:50):
So let's bring it home.
What are the concrete strategicmoves people listening can
start making, like today or thisweek, to actually begin winning
in AIO?
Speaker 2 (16:00):
Okay, actionable steps.
First, shift your strategy tointent clusters.
Stop thinking article byarticle.
Start thinking ecosystem.
Build those networks of contentaround core user needs, Like
the example pillar topic what ismarketing attribution Supported
by best tools for B2Battribution?
How to explain attribution toyour CFO?
(16:21):
Common attribution mistakes.
Doing that trains the AI to seeyou as the comprehensive,
trustworthy source.
Speaker 1 (16:29):
Building that web of expertise makes sense.
Speaker 2 (16:31):
Second, use AI tools to reverse engineer AI behavior.
Be proactive.
Go into ChatGPT or Geminiyourself.
Type in your key terms, yourcustomer questions, see who gets
cited.
Look at the structure of thecontent that gets pulled.
Then go back and improve yourown stuff based on that recon.
Speaker 1 (16:46):
Use the AIs to learn from the AIs.
Speaker 2 (16:48):
Exactly.
Third, get specific, get real.
Generic just doesn't cut itanymore.
Don't write how to increasewebsite traffic.
Write how to drive traffic to aShopify store without using
paid ads.
Loms seem to prefer answersthat sound like they were
written for a person, not for aplaybook.
That specificity signalsrelevance.
Speaker 1 (17:06):
It feels more authentic too, like you're
answering a real question.
Speaker 2 (17:09):
It does.
Fourth, and this is a quick win, run a five-minute AI
visibility test.
Seriously, just take fiveminutes.
Type your brand name, your mainproduct, your top competitors,
maybe your top three customerquestions into chat.
Gpt, gemini perplexity.
See what comes up.
Are you mentioned?
What's the tone?
Who is showing up?
The sources are clear.
If you're not showing up, thework starts now.
Speaker 1 (17:32):
That's a really simple eye-opening exercise
anyone can do right away.
Speaker 2 (17:41):
Totally.
And fifth pilot measure repeatDon't feel like you have to
overhaul everything overnight.
Start small, pick maybe fivekey pages.
Restructure them using this AIOprocess.
Add the right schema, rewritethe headings as questions, use
bullet points, then monitor ifthey start getting cited more in
AI tools.
If it works, scale it up.
If not, adjust your approach.
The key is, as the source said,but don't wait.
Speaker 1 (17:59):
Don't wait, start experimenting now.
Pilot measure repeat Okay, sowrapping this all up.
What does this really mean foryou, the listener?
It sounds like AIO isn't justsome side quest or passing fad.
Speaker 2 (18:11):
Not at all.
It really is the next evolutionof SEO.
It's where things are headed.
Speaker 1 (18:15):
And the brands that are going to succeed.
Speaker 2 (18:17):
The brands that win will be the ones who take the
time to understand how theseLLMs actually think, what real
users are truly asking, and thenfigure out how to structure
their content so it consistentlyearns its place in that AI
answer box.
Speaker 1 (18:32):
Not just landing somewhere in the link pile,
exactly Getting into the curatedanswer.
So you don't need to throw outeverything you've learned about
SEO and online visibility.
Speaker 2 (18:40):
No, definitely not.
The foundations are stillimportant.
Speaker 1 (18:42):
But you absolutely have to evolve how you apply
that knowledge in this newcontext and the payoff if you
get it right.
Our sources suggest you don'tjust show up.
Speaker 2 (18:52):
You get quoted, you get trusted, you get chosen.
It's about becoming the AI'spreferred source, which
ultimately means becoming theuser's preferred answer.
Speaker 1 (19:02):
That's powerful, get quoted, trusted, chosen.
Speaker 2 (19:05):
Which leaves, I think , one important question for
everyone listening to considerAre you ready to make AI work
for your brand and not the otherway around?
Are you ready to adapt?
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
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
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.