Google Web Guide: How It’s Reshaping The SERP And What It Means For Your SEO Strategy via @sejournal, @cyberandy

For decades, the digital world has been defined by hyperlinks, a simple, powerful way to connect documents across a vast, unstructured library. Yet, the foundational vision for the web was always more ambitious.

It was a vision of a Semantic Web, a web where the relationships between concepts are as important as the links between pages, allowing machines to understand the context and meaning of information, not just index its text.

With its latest Search Labs experiment, Web Guide (that got me so excited), Google is taking an important step in this direction.

Google’s Web Guide is designed to make it easier to find the information, not just webpages. It is optimized as an alternative to AI Mode and AI Overview for tackling complex, multi-part questions or to explore a topic from multiple angles.

Built using a customized version of the Gemini AI model, Web Guide organizes search results into helpful, easy-to-browse groups.

This is a pivotal moment. It signals that the core infrastructure of search is now evolving to natively support the principle of semantic understanding.

Web Guide represents a shift away from a web of pages and average rankings and toward a web of understanding and hyper-personalization.

This article will deconstruct the technology behind Web Guide, analyzing its dual impact on publishers and refining a possibly new playbook for the era of SEO or Generative Engine Optimization (GEO) if you like.

I personally don’t see Web Guide as just another feature; I see it as a glimpse into the future of how knowledge shall be discovered and consumed.

How Google’s Web Guide Works: The Technology Behind The Hyper-Personalized SERP

At its surface, Google Web Guide is a visual redesign of the search results page. It replaces the traditional, linear list of “10 blue links” with a structured mosaic of thematic content.

For an exploratory search like [how to solo travel in Japan], a user might see distinct, expandable clusters for “comprehensive guides,” “personal experiences,” and “safety recommendations.”

This allows users to immediately drill down into the facet of their query that is most relevant to them.

But, the real revolution is happening behind the scenes. This curation is powered by a custom version of Google’s Gemini model, but the key to its effectiveness is a technique known as “query fan-out.”

When a user enters a query, the AI doesn’t just search for that exact phrase. Instead, it deconstructs the user’s likely intent into a series of implicit, more specific sub-queries, “fanning out” to search for them in parallel.

For the “solo travel in Japan” query, the fan-out might generate internal searches for “Japan travel safety for solo women,” “best blogs for Japan travel,” and “using the Japan Rail Pass.”

By casting this wider net, the AI gathers a richer, more diverse set of results. It then analyzes and organizes these results into the thematic clusters presented to the user. This is the engine of hyper-personalization.

The SERP is no longer a one-size-fits-all list; it’s a dynamically generated, personalized guide built to match the multiple, often unstated, intents of a specific user’s query. (Here is the early analysis I did by analyzing the network traffic – HAR file – behind a request.)

To visualize how this works in semantic terms, let’s consider the query “things to know about running on the beach,” which the AI breaks down into the following facets:

The WebGuide UI is composed of several elements designed to provide a comprehensive and personalized experience:

• Main Topic: The central theme or query that the user has entered.
• Branches: The main categories of information generated in response to the user’s query. These branches are derived from various online sources to provide a well-rounded overview.
• Sites: The specific websites from which the information is sourced. Each piece of information within the branches is attributed to its original source, including the entity name and a direct URL.

Let’s review Web Guide in the context of Google’s other AI initiatives.

Web Guide’s unique role is that of an AI-powered curator or librarian.

It adds a layer of AI organization while preserving the fundamental link-clicking experience, making it a strategically distinct and potentially less contentious implementation of AI in search.

The Publisher’s Conundrum: Threat Or Opportunity?

The central concern surrounding any AI-driven search feature is the potential for a severe loss of organic traffic, the economic lifeblood of most content creators. This anxiety is not speculative.

Cloudflare’s CEO has publicly criticized these moves as another step in “breaking publishers’ business models,” a sentiment that reflects deep apprehension across the digital content landscape.

This fear is contextualized by the well-documented impact of Web Guide’s sibling feature, AI Overviews.

A critical study by the Pew Research Center revealed that the presence of an AI summary at the top of a SERP dramatically reduces the likelihood that a user will click on an organic link, a nearly 50% relative drop in click-through rate in its analysis.

Google has mounted a vigorous defense, claiming it has “not observed significant drops in aggregate web traffic” and that the clicks that do come from pages with AI Overviews are of “higher quality.”

Amid this, Web Guide presents a more nuanced picture. There is a credible argument that, by preserving the link-clicking paradigm, it could be a more publisher-friendly application of AI.

Its “query fan-out” technique could benefit high-quality, specialized content that has struggled to rank for broad keywords.

In this optimistic view, Web Guide acts as a helpful librarian, guiding users to the right shelf in the library rather than just reading them a summary at the front desk.

However, even this more “link-friendly” approach cedes immense editorial control to an opaque algorithm, making the ultimate impact on net traffic uncertain to say the least.

The New Playbook: Building For The “Query Fan-Out”

The traditional goal of securing the No. 1 ranking for a specific keyword is rapidly becoming an outdated and insufficient goal.

In this new landscape, visibility is defined by contextual relevance and presence within AI-generated clusters. This requires a new strategic discipline: Generative Engine Optimization (GEO).

GEO expands the focus from optimizing for crawlers to optimizing for discoverability within AI-driven ecosystems.

The key to success in this new paradigm lies in understanding and aligning with the “query fan-out” mechanism.

Pillar 1: Build For The “Query Fan-Out” With Topical Authority

The most effective strategy is to pre-emptively build content that maps directly to the AI’s likely “fan-out” queries.

This means deconstructing your areas of expertise into core topics and constituent subtopics, and then building comprehensive content clusters that cover every facet of a subject.

This involves creating a central “pillar” page for a broad topic, which then links out to a “constellation” of highly detailed, dedicated articles that cover every conceivable sub-topic.

For “things to know about running on the beach,” (the example above) a publisher should create a central guide that links to individual, in-depth articles such as “The Benefits and Risks of Running on Wet vs. Dry Sand,” “What Shoes (If Any) Are Best for Beach Running?,” “Hydration and Sun Protection Tips for Beach Runners,” and “How to Improve Your Technique for Softer Surfaces.”

By creating and intelligently interlinking this content constellation, a publisher signals to the AI that their domain possesses comprehensive authority on the entire topic.

This dramatically increases the probability that when the AI “fans out” its queries, it will find multiple high-quality results from that single domain, making it a prime candidate to be featured across several of Web Guide’s curated clusters.

This strategy must be built upon Google’s established E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) principles, which are amplified in an AI-driven environment.

Pillar 2: Master Technical & Semantic SEO For An AI Audience

While Google states there are no new technical requirements for AI features, the shift to AI curation elevates the importance of existing best practices.

• Structured Data (Schema Markup): This is now more critical than ever. Structured data acts as a direct line of communication to AI models, explicitly defining the entities, properties, and relationships within your content. It makes content “AI-readable,” helping the system understand context with greater precision. This could mean the difference between being correctly identified as a “how-to guide” versus a “personal experience blog,” and thus being placed in the appropriate cluster.
• Foundational Site Health: The AI model needs to see a page the same way a user does. A well-organized site architecture, with clean URL structures that group similar topics into directories, provides strong signals to the AI about your site’s topical structure. Crawlability, a good page experience, and mobile usability are essential prerequisites for competing effectively.
• Write with semiotics in mind: As Gianluca Fiorelli would say, focus on the signals behind the message. AI systems now rely on hybrid chunking; they break content into meaning-rich segments that combine text, structure, visuals, and metadata. The clearer your semiotic signals (headings, entities, structured data, images, and relationships), the easier it is for AI to interpret the purpose and context of your content. In this AI-gated search environment, meaning and context have become your new keywords.

The Unseen Risks: Bias In The Black Box

A significant criticism of AI-driven systems like Web Guide lies in their inherent opacity. These “black boxes” pose a formidable challenge to accountability and fairness.

The criteria by which the Gemini model decides which categories to generate and which pages to include are not public, raising profound questions about the equity of the curation process.

There is a significant risk that the AI will not only reflect but also amplify existing societal and brand biases. A compelling example is to review complex issues to test the fairness of the Web Guide.

Medical diagnostic queries are complex and can easily reveal biases.

Once again, UGC is used and might not always bring the right nuance between doom narratives and overly optimistic positions.

Since the feature is built upon these same core systems of traditional Search, it is highly probable that it will perpetuate existing biases.

Conclusion: The Age Of The Semantic AI-Curated Web

Google’s Web Guide is not a temporary UI update; it is a manifestation of a deeper, irreversible transformation in information discovery.

It represents Google’s attempt to navigate the passage between the old world of the open, link-based web and the new world of generative, answer-based AI.

The “query fan-out” mechanism is the key to understanding its impact and the new strategic direction. For all stakeholders, adaptation is not optional.

The strategies that guaranteed success in the past are no longer sufficient. The core imperatives are clear: Embrace topical authority as a direct response to the AI’s mechanics, master the principles of Semantic SEO, and prioritize the diversification of traffic sources. The era of the 10 blue links is over.

The era of the AI-curated “chunks” has begun, and success will belong to those who build a deep, semantic repository of expertise that AI can reliably understand, trust, and surface.

More Resources:

---

Featured Image: NicoElNino/Shutterstock

Why Semantic HTML Matters For SEO And AI

I’ve had this post in drafts for a while, mostly as a container for me to drop bits into for when I get time to expand it into a proper newsletter.

Then, my good friend Jono Alderson published his excellent piece on semantic HTML, and for a few weeks, I lost the will to complete mine.

But, I thought I should finish my version anyway, as my focus is slightly different and perhaps a bit more practical than Jono’s.

You should still definitely read Jono’s blog; it says all I want to say and more.

Semantic HTML

Let’s start with a quick overview of what semantic HTML is.

As the language upon which the web is built, HTML is a markup that surrounds text to provide it with structure.

The

tag around a block of content indicates that it is a paragraph of text.

The

tag around a sentence shows that it is the page’s main heading.

The

tag indicates the start of an ordered (usually numbered) list.

The tag indicates you’ll be loading an image onto the webpage. And so forth.

Semantic HTML was used to code every webpage.

Content was surrounded by specific tags that indicated what each bit of content was meant for, and then CSS was applied to make it look good. It wasn’t perfect by any means, but it worked.

It also meant that you could look at the raw HTML source of a webpage and see what the page was trying to deliver, and how. The HTML signposted the structure and meaning of each bit of content on the page. You could see the purpose of the page just by looking at its code.

Then WYSIWYG editors and later JavaScript frameworks arrived on the scene, and HTML took a backseat. Instead of

and

, we got endless nestings of
and tags.

The end result is webpage HTML that lacks structure and has no meaning, until it is completely rendered in the browser and visually painted onto a screen. Only then will the user (and a machine system trying to emulate a user) understand what the page’s purpose is.

It’s why Google goes through the effort of rendering pages as part of its indexing process (even though it really doesn’t want to).

We know Google doesn’t usually have the time to render a news article before it needs to rank it in Top Stories and elsewhere. The raw HTML is therefore immensely important for news publishers.

Good HTML allows Google to effortlessly extract your article content and rank your story where it deserves in Google’s ecosystem.

Semantic HTML is a key factor here. This is the reason why SEOs like me insist that an article’s headline is wrapped in the

heading tag, and that this is the only instance of

on an article page.

The H1 headline indicates a webpage’s primary headline. It signposts where the article begins, so that Google can find the article content easily.

Which HTML Tags Are Semantic?

Beyond the

heading tag, there are many other semantic HTML elements you can implement that allow Google to more easily extract and index your article content.

In no particular order, the elements you should be using are:

• Paragraphs: Don’t use
  and tags to format the article into paragraphs. There’s been a tag for that for as long as HTML has existed, and it’s the

  tag. Use it.

• Subheadings: Use
  

  /

  /

  subheading tags to give your page structure. Use subheadings in an article to preface specific sections of content in your article. Use subheadings for the headers above concrete structural elements, such as recommended articles.

Images: Always use the tag if you want to show an image that you’d like Google to see as well. Google explicitly recommends this.

Clickable Links: When linking to another page, either internal or external, use the tag with an “href” value containing the target URL. It’s the only kind of link that Google will definitely follow.

Relational Links: The tag allows you to create a relationship between the current URL and another URL. This can be a canonical page, a stylesheet, an alternative language version of the current page, etc.

Lists: Bullet lists should use the

tag, and numbered lists should use
tag. You can make them look however you want with CSS, but do use the list tags as the foundation.

Emphasis: When you want to highlight a specific word or phrase, there are semantic HTML tags you should use for that: for italics, and for bold.

All the above tags, with the exception of , are intended for the content of the webpage, providing structure and meaning to the text.

There are additional semantic HTML tags that are intended to provide structure and meaning to the code of the page.

These tags allow Google to identify different elements on the page, such as the navigation vs. a sidebar, and process them accordingly.

Semantic HTML image from W3Schools.com (Image Credit: Barry Adams)

• The  and  tags exist to separate the page’s metadata (in the ) from the actual content (in the ). Every HTML page starts with those two.
• 
  

   can be used to wrap around the head section of the page, where the logo, navigation, and other stylistic elements sit.

• 
  

   should be used for your site’s main navigation. Mega menus, hamburger menus, top navigation links, whatever form your navigation takes, you should wrap it in the

• You can use 
  

  tags to divide your page into multiple sections. One section could be the article; another could be the comments below the article.

• 
  

   is the tag that shows where the page’s actual main article text begins (including the headline). This is a very valuable tag for news publishers.

• With 
  

   you can indicate blocks of content like a sidebar of trending stories, recommended articles, or the latest news.

• 
  

   is used for, you guessed it, the footer of the webpage.

These structural semantic tags help search engines understand the purpose and value of each section of HTML.

It enables Google to rapidly index your content and process the different elements of your pages appropriately.

There are many more semantic HTML tags at your disposal, for various different purposes. Chances are, there’s an HTML element for every imaginable use case.

Rather than cram your code full of

tags to make something happen, first see if there’s a proper HTML element that does the trick.

How Does It Help AI?

We know that LLMs like ChatGPT and Perplexity crawl the open web for training data, as well as for specific user queries that require content from the web.

What some of you may not know is that LLMs do not render JavaScript when they process webpages.

Google is the exception to the rule, as it has devoted a great deal of resources to rendering webpages as part of indexing.

Because Google’s Gemini is the only LLM built on Google’s index, Gemini is the only LLM that uses content from fully rendered webpages.

So, if you want to have any chance of showing up as a cited source in ChatGPT or Perplexity, you’d do well to ensure your complete page content is available in your raw, unrendered HTML.

Using semantic HTML to structure your code and provide meaning also helps these LLMs easily identify your core content.

It’s much simpler for ChatGPT to parse a few dozen semantic HTML tags rather than several hundred (or even thousand) nested

tags to find a webpage’s main content.

If and when the “agentic web” comes to life (I’m skeptical), semantic HTML is likely a crucial aspect of success.

With meaningless

and tags, it’s much easier for an AI agent to misunderstand what actions it should perform.

When you use semantic HTML for things like buttons, links, and forms, the chances of an AI agent failing its task are much lower.

The meaning inherent in proper HTML tags will tell the AI agent where to go and what to do.

What About Structured Data?

You may think that structured data has made semantic HTML obsolete.

After all, with structured data, you can provide machine systems with the necessary information about a page’s content and purpose in a simple machine-readable format.

This is true to an extent. However, structured data was never intended to replace semantic HTML. It serves an entirely different purpose.

Structured data has limitations that semantic HTML doesn’t have.

Structured data won’t tell a machine which button adds a product to a cart, what subheading precedes a critical paragraph of text, and which links the reader should click on for more information.

By all means, use structured data to enrich your pages and help machines understand your content. But you should also use semantic HTML for the same reasons.

Used together, semantic HTML and structured data are an unbeatable combination.

Build Websites, Not Web Apps

I could go off on a 2,500-word rant about why we should be building websites instead of web apps and how the appification of the web is anathema to the principles on which the World Wide Web was founded, but I’ll spare you that particular polemic.

Suffice it to say that web apps for content-delivery websites (like news sites) are almost always inferior to plain old-fashioned websites.

And websites are built, or should be, on HTML. Make use of all that HTML has to offer, and you’re avoiding 90% of the technical SEO pitfalls that web apps tend to faceplant themselves into.

That’s it for another edition. Thanks for reading and subscribing, and I’ll see you at the next one!

---

This post was originally published on SEO For Google News.

---

Featured Image: N Universe/Shutterstock

Cohorts, Clusters, And The Coming AI Ad System via @sejournal, @DuaneForrester

The funnel didn’t disappear. It went invisible.

Marketers spent decades perfecting the funnel: awareness, consideration, conversion. We built personas. We mapped content to stages. We watched users click, scroll, bounce, convert. Everything was visible.

But GenAI doesn’t show its hand.

The funnel still exists, it’s just hidden inside the model. Every time someone prompts ChatGPT or Perplexity, they reveal their place in a decision journey.

Not by filling out a form or triggering a pixel, but through the prompt fingerprint embedded in their question.

That’s the new funnel. You’re still being evaluated. Still being chosen. But the targeting is now invisible, inferred, and dynamic.

And most marketers have no idea it’s happening. In fairness, I think only the cohort portion of this is actively happening today.

The ad system I explore here is purely theoretical (though Google appears to be working in a similar direction currently, and its rollout could be realistic, soon – links below).

TL;DR: This article doesn’t just explain how I think GenAI is reshaping audience targeting; it introduces three new concepts I think you’ll need to understand the next evolution of paid media: Prompt Fingerprints, Embedding Fingerprints, and Intent Vector Bidding. 

The funnel isn’t gone. It’s embedded. And it’s about to start building and placing ads on its own.

About the terminology: 

Prompt Fingerprint and Intent Vector Bidding, I believe, are net-new terms for our industry, coined here to describe how future LLM-based systems could group users and auction ad space.

Conceptually, Intent Vector Bidding aligns with work already being done behind the scenes at Google (and I’m sure elsewhere), though I don’t believe they use this phrase. 

Embedding Fingerprint draws from AI research but is reframed here as a brand-side construct to power targeting and retrieval inside GenAI systems.

This article was written over the last three weeks of July, and I was happy to find an article on August 4 talking about the concepts I’m exploring for a future paid ads bidding system.

Coincidental, but validating. The link to that article is below.

What Cohort Targeting Used To Be

In the pre-AI era, cohort targeting was built around observable behaviors.

• Retargeting audiences built from cookies and pixels.
• Segments shaped by demographics, location, and device.
• Lookalikes trained on customer traits and CRM lists.

We mapped campaigns to persona types and funnel stages. A 42-year-old dad in Ohio was mid-funnel if he clicked a product video. An 18-year-old in Mumbai was top-funnel if he downloaded an ebook.

These were guesses, good ones, often, but still blunt instruments. And they were built on identifiers that don’t necessarily survive the GenAI shift.

Prompts Are The New Personas

Large language models don’t need to know who you are. They don’t really need to track you. They don’t care where you came from. They only care what you ask, and how you ask it.

Every prompt is vectorized. That means it’s turned into a mathematical representation of meaning, called an embedding. These vectors capture everything the model can glean from your input:

• Topical domain.
• Familiarity and depth.
• Sentiment and urgency.
• Stage of intent.

LLMs use this signal to group prompts with similar meaning, even if they come from completely different types of people.

And that’s how new cohorts can form. Not from identity. From intent.

Right now, most marketers are still optimizing for keywords, and missing the bigger picture. Keywords describe what someone is searching for. Prompt fingerprints describe why and how.

Someone asking “quietest portable generator for camping” isn’t just looking for a product, they’re signaling lifestyle priorities (minimal noise, portability, outdoor use) and stage (comparison shopping).

That single prompt tells the model far more than any demographic profile ever could.

And crucially, that person is joining a cohort of other prompters asking similar questions in similar ways. If your content isn’t semantically aligned with that group, it’s not just less visible. It’s excluded.

New Concept: Prompt Fingerprint

A unique embedding signature derived from a user’s language, structure, and inferred intent within a prompt. This fingerprint is your new persona.

It’s what the model actually sees and what it uses to determine which answers (and potentially which ads) you receive. (More on those ads later!)

When Context Creates The Cohort

Let’s say the Toronto Maple Leafs just won the Stanley Cup (hey, a guy can dream, right?!). Across the city, thousands of people start prompting:

• “Where to celebrate in Toronto tonight?”
• “Best bars near Scotiabank Arena open late?”
• “Leaf’s victory parade time and location?”

None of these users knows each other. Some are teenagers, others are retirees. Some are local, others are visiting. Some are hardcore fans, some just like to party. But to the model, they’re now a momentary cohort; a group connected by real-time context, not long-term traits.

This is a fundamental break from everything digital marketers are used to. We’ve always grouped people by identity: age, interests, behavior, psychographics. But LLMs group people by situational similarity.

That creates new marketing opportunities and new blind spots.

Imagine you sell travel gear. A major snowstorm is forecast to slam into the Northeast U.S.

Within hours, prompts spike around early departures, snowproof duffel bags, and waterproof boots. A travel-stress cohort forms: people trying to escape before the storm hits. They’re not a segment you planned for. They’re a moment the system saw before you did.

If your content or product is aligned with that moment, you need a system that detects, matches, and delivers immediately. That’s what makes system-embedded ad tech essential.

You’re not buying audiences anymore. You’re buying alignment with the now, with a moment in time.

And this part is real today.

While the inner workings of commercial GenAI systems remain opaque, cluster-like behavior is often visible within a single platform session.

When you ask a string of similar questions in one ChatGPT or Gemini session, you may encounter repeated phrasing, brand mentions, or answer structure. That consistency suggests the model is grouping prompts by embedded meaning, not demographics or declared traits.

I cannot find studies or examples of this behavior being recorded, so please drop a comment if you have a source for such data. I keep hearing about it, but cannot find dedicated data.

Looking Forward

Entire classes of micro-cohorts may form and disappear within hours. To reach them, you’ll need AI-powered, system-embedded ad systems that can:

• Detect the cohort’s emergence through real-time prompt patterns.
• Generate ads aligned with the cohort’s immediate need.
• Place and optimize those ads before the window closes.

Humans can’t move at that speed. AI can. And it has to because the opportunity vanishes with the context.

Sidebar: What I Think Is Real Vs. What I Think Is Coming

• Prompt Fingerprints – Live Today: Every GenAI system turns your prompt into a vector embedding. It’s already the foundation of how models interpret meaning.
• Cohort Clustering by Prompt Similarity – Active Now: You can observe this in tools like ChatGPT and Gemini. Similar prompts return similar answers, meaning the system is clustering users based on shared intent.
• Embedding Fingerprints – Possible Today: If brands structure their content for vectorization, they can create an embedding signature that aligns with relevant prompts. Most don’t yet.
• Intent Vector Bidding – Emerging Theory: Almost in the market today. Given current ad platform trends, this kind of bidding system is likely being explored widely across platforms.

Why Old-School Personas Will Work Less Effectively

Age. Income. ZIP code. None of that maps cleanly in vector space.

In the GenAI era, two people with radically different demographics might prompt in nearly identical ways and be served the same answers as a result.

It’s not about who you are. It’s about how your question fits into the model’s understanding of the world.

The classic marketing persona is much less reliable as a targeting unit. I’m suggesting the new unit is the Prompt Fingerprint, and marketers who ignore that shift may find themselves omitted from the conversation entirely.

The Funnel Is Still There — You Just Can’t See It

Here’s the thing: LLMs do understand funnel stages.

They just don’t label them the way marketers do. They infer them from phrasing, specificity, and structure.

• TOFU: “Best folding kayaks for beginners”
• MOFU: “Oru Inlet vs. Tucktec comparison”
• BOFU: “Oru kayak discount codes July 2025”

These are prompt-level indicators of funnel stage. And if your content doesn’t align with how those prompts are formed, it likely won’t get retrieved.

Want to stay visible? Start mapping your content to the language patterns of funnel-stage prompts, not just to topics or keywords.

Embedding Fingerprints: The New Targeting Payload

It’s not just prompts that get vectorized. Your content does, too.

Every product page, blog post, or ad you write forms its own Embedding Fingerprint, a vector signature that reflects what your message actually means in the model’s understanding.

Repurposed Concept: Embedding Fingerprint

Originally used in machine learning to describe the vector signature of a piece of data, this concept is reframed here for content strategy.

An embedding fingerprint becomes the reusable vector signature tied to a brand, product, or message – a semantic identity that determines cohort alignment in GenAI systems.

If your content’s fingerprint aligns closely with a user’s prompt fingerprint, it’s more likely to be retrieved. If not, it’s effectively invisible, no matter how “optimized” it may be in traditional terms.

Intent Vector Bidding: A Possible New Advertising Paradigm

So, what happens when GenAI systems all start monetizing this behavior?

You could get a new kind of auction. One where the bid isn’t for a keyword or a user profile, per se, but for alignment.

New Concept: Intent Vector Bidding

A real-time ad bidding mechanism where placement is determined by alignment between a user’s prompt intent vector and an advertiser’s content vector.

To be clear: this is not live today in any public, commercial ad platform that I am aware of. But I think it’s well within reach. Models already understand alignment. Prompt clustering is already happening.

What’s missing is the infrastructure to let advertisers fully plug in. And you can bet the major players (OpenAI, Google, Meta, Microsoft, Amazon, etc.) are already thinking this way. Google is already looking at this openly.

We’ve Been Heading Here All Along

The shift toward LLM-native ad platforms might sound radical, but in reality, we’ve been headed this way for over a decade.

Step by step, platform by platform, advertisers have been ceding control to automation, often without realizing they were walking toward full autonomy.

Before we trace the path, please keep in mind that while I do have some background in the paid ad world, it’s much less than many of you.

I’m attempting to keep my date ranges and tech evolutions accurate, and I believe they are, but others may have a different view.

My point here isn’t historical accuracy, it’s to demonstrate a continual, directional progression, not nail down on which day of which year did Google do X.

And, I’ll add, maybe I’m entirely off base with my thinking here, but it’s still been interesting to map all this out, especially since Google has already been digging in on a similar concept.

1. From Manual Control To Rule-Based Efficiency

• Early 2000s – 2015

In the early days of search and display, marketers controlled everything: keyword targeting, match types, ad copy, placements, and bidding.

Power users lived inside tools like AdWords Editor, manually optimizing bids by time of day, device type, and conversion rate.

Automation started small, with rule-based scripts for bid adjustments, budget caps, and geo-targeting refinements. You were still the pilot, just with some helpful instruments.

2. From Rule-Based Logic To AI-Guided Bidding

• 2015 – 2018

Then came Smart Bidding.

Google introduced Target CPA, Target ROAS, and Enhanced CPC: bid strategies powered by machine learning models that ingested real-time auction data (device, time, location, conversion likelihood) and made granular decisions on your behalf.

Marketers set the goal, but the system chose the path. Control shifted from how to what result you want. This was a foundational step toward AI-defined outcomes.

3. From AI-Guided Bidding To Creative Automation

• 2018 – 2023

Next came the automation of the message itself.

Responsive Search Ads let advertisers upload multiple headlines and descriptions and Google handled the permutations and combinations.

Meta and TikTok adopted similar dynamic creative formats.

Then Google launched Performance Max (2021), a turning point that eliminated keywords entirely.

• You provide assets and conversion goals.
• The system decides where and when to show your ads, whether across Search, YouTube, Display, Gmail, Maps, and more.
• Targeting becomes opaque. Placement is more invisible. Strategy becomes trust.

You’re no longer steering the vehicle. You’re defining the destination and expecting the algorithm gets you there efficiently.

4. From Creative Automation To Generative Execution

• 2023–2025
  

The model doesn’t just optimize messages anymore; it writes them.

• Meta’s AI Sandbox generates headlines and CTAs from a prompt.
• TikTok’s Creative Assistant produces hook-driven video scripts on demand.
• Third-party tools and GPT-based agents build full ad campaigns, including copy and targeting.
• Google’s Veo 3 and Veo 3 Fast now live on Vertex AI, generate polished ads and social clips from text or image-to-video inputs, optimized for rapid iteration and programmatic use.

This isn’t sci-fi. It’s what’s coming to market today.

5. What Comes Next – And Why It’s Inevitable

The final leap is where you don’t submit an ad, you instead submit your business.

A fully LLM-native ad platform would:

• Accept your brand’s value propositions, certifications, product specs, creative assets, brand guidelines, company vision statements, and guardrails.
• Monitor emergent cohorts in real time based on prompt clusters and conversation spikes.
• Inject your brand into those moments if, and only if, your business’s vector aligns with the cohort’s intent.
• Charge you automatically for participation in that alignment.

You wouldn’t target. You wouldn’t build campaigns. You’d just feed the system and monitor how well it performs as a semantic extension of your business.

The ad platform becomes a meaning-based proxy for your company, an intent-aware agent acting on your behalf.

That’s not speculative science fiction. It’s a natural endpoint of the road we’re already on, I believe. Performance Max removed the steering wheel. Generative AI threw out the copywriter. Prompt-aligned retrieval will take care of the rest.

Building The LLM-Native Ad Platform

This is a theoretical suggestion of what could be our future for paid ads within AI-generated answer systems.

To make Intent Vector Bidding real at scale, the underlying ad platform will have to evolve dramatically. I don’t see this as a plug-in bolted onto legacy PPC infrastructure.

It will be a fully native layer inside LLM-based systems, one that replaces both creative generation and ad placement management.

Here’s how it could work:

1. Advertiser Input Shifts From Campaigns To Data Feeds

Instead of building ads manually, businesses upload:

• Targeted keywords, concepts, and product entities.
• Multimedia assets: images, videos, audio clips.
• Credentials: certifications, affiliations, licenses.
• Brand guidelines: tone, voice, claims to avoid.
• Business limitations: geography, availability, compliance.
• Structured value props and pricing tiers.

2. The System Becomes The Creative + Placement Engine

The LLM:

• Detects emerging prompt cohorts.
• Matches intent vectors to advertiser fingerprints.
• Constructs and injects ads on the fly, using aligned assets and messaging.
• Adjusts tone and detail based on prompt stage (TOFU vs BOFU).

3. Billing Becomes Automated And Embedded

• Accounts are pre-funded or credit-card linked.
• Ad spend is triggered by real-time participation in retrieval or output injection.
• No ad reps. No auctions you manage. Just vector-aligned outcomes billed per engagement, view, or inclusion.
• Ad creation and placement become a single-price-point item as the system manages all, in real time.

If you want some more thoughts on this concept, or one that’s closely related, Cindy Krum was recently on Shelley Walsh’s IMHO show, where she talked about whether she thinks Google will put ads inside Gemini’s answers, and it was an interesting discussion.

You should give it a listen. And this report on Google suggests this is not only here now, but expanding.

The Human Role Doesn’t Disappear – It Evolves

Marketers and ad teams won’t be eliminated. Instead, they’ll become the data stewards and strategic interpreters of the system.

• Expectation setting: Clients will need help understanding why their content shows up (or doesn’t) in GenAI outputs.
• Data maintenance: The system is only as good as the assets you feed it, and relevance and freshness matter.
• Governance and constraints: Humans will define ethical limits, messaging boundaries, and exclusions.
• Training and iteration: AI ad visibility will rely on live outputs and observed responses, not static dashboards. You’ll tune prompts, inputs, and outputs based on what the system retrieves and how often it surfaces your content.

In this model, the ad strategist becomes part translator, part data curator, part retrieval mechanic.

And the ad platform? It becomes autonomous, context-driven, and functionally invisible, until you realize your product’s already been included in the buyer’s decision … and you’ve been billed accordingly.

A Closer Look: Intent Vector Bidding In Action

Imagine you’re an outdoor gear brand and there’s a sudden heatwave hitting the Pacific Northwest. Across Oregon and Washington, people begin prompting:

• “Best ultralight tents for summer hiking”
• “Camping gear for extreme heat”
• “Stay cool while backpacking in July”

The model recognizes a spike in semantically similar prompts and data from news sources, etc. A heatwave cohort forms.

At the same time, your brand has a product page and ad copy about breathable mesh tents and high-vent airflow systems.

If your content has been vectorized (or if your system embeds an ad payload with a strong Embedding Fingerprint), it’s eligible to enter the auction.

But this isn’t a bid based on demographic data or historical retargeting. It’s based on how closely your product vector aligns with the live cohort’s prompt vectors.

The LLM chooses the most semantically aligned match. The better your alignment, the more likely your product is included in the AI’s answer, or inserted into the contextual ad slot within the response.

No campaign setup. No segmented audience targeting. Just semantic match at machine speed. This is where creative, product, and performance converge, and that convergence rewrites what it means to “win” in modern advertising.

What Marketers Can Do Right Now

There’s no dashboard that will tell you which Prompt Fingerprints you’re aligned with. That’s the hard part.

But you can start by thinking like a model until tools start to develop features that allow you to model your Prompt Fingerprint.

Start with:

• Simulated prompt testing: Use GPT-4 (or Gemini or any other) to generate sample queries by funnel stage and see what brands get retrieved.
• Create content for multi-cohort resonance: for example, a camping blog that aligns with both eco-conscious minimalists and adventure-seeking parents.
• Build your own prompt libraries: Classify by intent stage, specificity, and phrasing. Use these to guide creative briefs, content chunking, and SEO.
• Track AI summaries: In platforms like Perplexity, Gemini, and ChatGPT, your brand might influence answers even when you’re not explicitly mentioned. Your goal is to become the attributed source, not just a silent contributor.

In this new, genAI version of search, you’re no longer optimizing for page views. You’re optimizing for retrievability by semantic proximity.

The Rise Of The Prompt-Native Brand

Some brands will begin designing entire messaging strategies around prompt behavior. These prompt-native brands won’t wait for traffic to arrive. They’ll engineer their content to surf the wave of prompt clusters as they form.

• Product copy structured to match MOFU queries.
• Comparison pages written in prompt-first language.
• AI ad copy tuned by cohort spike detection.

And eventually, new brands will emerge that never even needed a traditional website. Their entire presence will exist in AI conversations.

Built, tuned, and served directly into LLMs via vector-aligned content and Intent Vector Bids.

Wrapping Up

This is the next funnel, and it’s not a page. It’s a probability field. The funnel didn’t disappear. It just went invisible.

In traditional marketing, we mapped clear stages (awareness, interest, decision) and built content to match. That funnel still exists. But now it lives inside the model. It’s inferred, not declared. It’s shaped by prompts, not click paths.

And if your content doesn’t align with what the model sees in that moment, you’re missing in the retrieval.

More Resources:

---

This post was originally published on Duane Forrester Decodes.

---

Featured Image: NicoElNino/Shutterstock

AI Crawler Optimization Tips

Generative AI platforms such as ChatGPT, Perplexity, and Claude now execute live web searches with all prompts. Ensuring a site is crawlable by AI bots is therefore essential for mentions and citations on those platforms.

Here’s how to optimize a website for AI crawlers.

Disable JavaScript

Make sure your pages are readable with JavaScript disabled.

Unlike Google’s crawler, AI bots are immature. Many tests from industry practitioners confirm AI crawlers cannot always render JavaScript.

Most publishers and businesses no longer worry about JavaScript crawlability since Google has rendered those pages for years. Hence there’s a huge number of JavaScript-heavy sites.

The Chrome browser can render a site without JavaScript. To activate:

• Go to your site using Chrome.

• Open Web Developer tools at View > Developer > Developer Tools.

• Click Settings (behind the gear icon) on the right side of the panel.

• Scroll down and check the option “Disable JavaScript” under “Debugger.”

Now browse your site, making sure:

• All essential content is visible, especially behind tabs and drop-down menus.

• The navigation menu and other links are clickable.

• For video embeds, there’s an option to click to the original video, access a transcript, or both.

You can use Aiso, an AI optimization platform, to ensure AI bots can access and crawl your site. With a free trial, the platform allows a few free checks. Go to the “Website crawlability” section and enter your URL.

The tool will conduct a thorough review with suggestions on improving access for AI crawlers and even show the appearance of pages with JavaScript disabled.

Ensure AI Access

Make sure your site allows access for AI bots. Some content management platforms and plugins disallow AI access by default — site owners are often unaware.

To check, review your robots.txt file at [yoursite.com]/robots.txt.

The AI platforms themselves can interpret the file to ensure it allows access. Paste your robots.text URL into a ChatGPT prompt, for example, and request an analysis.

Structured Data

Structured data markup, such as from Schema.org, can also help ensure visibility.

Schema markup makes it easier for AI bots to extract essential information from a page (or bypass a block) without crawling it in full.

For example, many website FAQ sections have collapsible elements that prevent access to AI bots. Schema’s FAQPage Type replicates all questions and answers, enabling bot visibility.

Similarly, Schema’s Article Type can communicate context and authorship of content.