Pages Are No Longer the Unit of Measurement. Answers Are.

Last week, I got an email from my client Bev with a screenshot of her recent victory: 

Because that’s not a traditional #1 ranking, you might not think of that as worth celebrating. You might even think of it as a “failure” because Google didn’t send traffic to Bev’s post about Kit vs. Substack. Instead, it extracted her answer, synthesized her explanation, and cited her business as the authoritative source.

This example of zero-click search is working in Bev’s favor because now her answer is essentially being endorsed by Google as “the best possible way to explain this.”

She didn’t win the click… but she became the answer.

The other part of the victory? A few months ago, that same post wasn’t being cited at all. 

Without changing the topic or writing anything new, we made some key adjustments to how she structured the answer — and that shift completely altered whether AI systems could find it, use it, and attribute it back to her.

If you’re still thinking the only metric to celebrate is “my page should rank #1,” you’re optimizing for a measurement system that no longer exists.

This is the new victory in the AI era: become the citable source that Google implicitly endorses.

The Shift From Optimizing for Pages to Answers

For two decades, the unit of measurement in search was the page.

You optimized pages so each page could rank for a separate set of keywords, then measured success by whether your page showed up in position 1, 2, or 3 (or 67, or 92 …).

The question search engines used to ask was: “Which page is most relevant to this query?”

But now, AI-mediated search systems — Google’s AI Overviews, Perplexity, ChatGPT search, Gemini — are asking: “Who has already explained this clearly, consistently, and credibly?”

You’re no longer competing page-to-page and seeing if you can shift up 1-2 positions by making tweaks to your content headline or meta description. 

Now, you’re competing entity-to-entity against other experts in your field, and the AI Overviews are being optimized on an answer-to-answer basis.

What matters now is whether your explanation is clear enough, structured enough, and reinforced enough that an AI system will choose you when it needs to answer that question. The page is just the container where the information lives, not the search end game.

What Bev’s Post Looked Like Before

Bev shared her post with me as part of the Signal content review process, and I was impressed at how much she packed into the post. 

For context, Bev’s an email marketing strategist who’s been working with clients for years, helping them choose between platforms like Kit and Substack based on their business models and audience needs.

So her blog post was chock full of rich insights … but the structure buried her expertise.

The post opened with an explanation of why she cares about email platforms before answering the question of “kit vs. substack.” The comparison people were searching for was halfway down the page.

And when it finally appeared, it wasn’t in a text paragraph, which is a more difficult format for AI systems to easily extract than structured content like bullets and tables. 

This is the expert’s paradox in action: Bev has so much knowledge about email service providers, email automation, and newsletter platforms that she wanted to explain why the answer mattered before giving the answer.

That’s good teaching… but it’s terrible for retrieval-based AI search systems. 

Because AI systems don’t read your post like a student in a classroom. They scan for clear, explicit answers to specific questions.

If your answer is wrapped in a narrative arc, hedged with qualifiers, or structured like an essay instead of a diagnostic? They move on to somebody who says clearly, “choose this tool if X, or this tool if Y.”

How to Format Your Content for Answer-Level Retrieval

If you want AI systems to retrieve, cite, and reuse your content, here’s what to do:

1. Answer-First Structure

Start with the answer, then explain why it matters, then show how to do it.

This is called answer-first formatting: Move the solution to the top of the post instead of making the AI system (or the reader) hunt for it.

Don’t bury the answer in narrative context or build suspense. The system doesn’t care about your journey. It cares about your answer.

That’s why Bev moved the answer of which tool is best for each business model to the introduction. She explicitly contrasted the tools in the first 200 words, then used the rest of the post to support that answer with more nuanced details.

Why this worked: AI systems retrieve chunks of content that directly answer questions. If the answer is buried under 500 words of context, they often won’t find it — or they’ll find a competitor’s answer that’s easier to extract.

Answer-first structure makes your content immediately useful to both humans and AI.

2. Clear Headings That Match Real Questions

Not: “Our Philosophy”
Yes: “How does X work?”
Yes: “What’s the difference between X and Y?”
Yes: “When should you choose X over Y?”

Headings should mirror the actual questions people are asking. This makes it easier for AI systems to identify which section answers which query.

So Bev changed her title from “:Kit vs. Substack: Which Email Platform is Better for Your Newsletter?” to “Kit vs. Substack: Which Platform is Right for Your Business??”

Why this worked: AI systems scan headings to identify which section answers which query. When your headings match the question structure, you make it easy for the system to route to the relevant part of your post.

3. Make your positioning explicit.

Don’t just describe the tools. Tell people when to choose one over the other.

Vague statements like “both tools have pros and cons” don’t give AI anything they can’t find from the standard software feature comparison tables. 

Bev gave clear, declarative guidance along these lines: “Choose Substack for pure writing/thought leadership; choose Kit for robust business email marketing.”

Why this worked: Bev’s diagnostic criteria of which tool worked best based on the reader’s business model and goals was specific, defensible, and useful. AI can extract it, paraphrase it, and cite Bev as the source.

4. Comparison Tables and Structured Breakdowns

If you’re comparing two things, use a table. If you’re listing steps, use numbered lists. If you’re contrasting approaches, make the contrast explicit. Don’t bury the important information in a wall of text.

Structure makes it easier for AI to extract information, which leads to more citations.

Bev created a table to show the types of features that might be beneficial for the different audiences that she was serving (business owners vs. thought leaders): 

Why this worked: Giving a concept structure makes it easier for tools to extract. AI systems can parse tables, pull out key differences, and cite you as the source of that comparison.

5. FAQ Sections That Actually Answer Questions

Answer real questions people ask with clear, standalone answers. If you’re not sure what to include, go Google it and look at the “People Also Ask” and “Related Searches” like this: 

Not all of those questions would be in Bev’s wheelhouse (she doesn’t need to dip her toe into the political issues around the platform) but she could clearly answer the first one based one what else is in the post.

FAQ sections work because they:

• Match the question-answer format AI systems expect
• Provide reusable chunks of information
• Signal that you’ve anticipated common follow-ups

FAQ schema can be added at the end of your post as confirmation of the questions and answers you’re already provided within the post — but now they’re structured in a way that are more likely to get surfaced in places like Google’s People Also Ask or Featured Snippets.

Why this works: Schema makes content easier for systems to parse and reuse the most common clarifying questions. .

6. Reinforce your entity across channels.

If you’ve only written about a topic once, on one page, on one platform? That’s not enough to show up in that AI Overview. 

Bev isn’t just the author of this one blog post. She’s also:

• A person entity with expertise in email marketing automation strategy
• She owns a recognizable business, Your Personal Tech Fairy
• Her home base is a website with consistent expertise signals around email marketing
• She’s regularly a guest expert about email marketing in multiple places: podcasts, LinkedIn, etc.

This is where omnichannel becomes critical. AI systems don’t just look at one page in isolation. They look for pattern consistency across domains.

So if you’ve:

• Written about it on your own site
• Discussed it on a podcast (with transcripts published)
• Been quoted in someone else’s article
• Published a newsletter on Substack or Kit’s Creator Network
• Mentioned it in a LinkedIn post or video

The system sees the same person, same expertise, and same positioning across multiple surfaces.

When I said earlier that the AI answers are entity-to-entity competition, this is what I meant. If somebody else wrote an even better post than Bev about kit vs. substack but had never written another word about email marketing, AI would still trust Bev more because she’s proven her topical authority on her website and around the internet. 

Why this worked: That cross-channel reinforcement tells AI systems: This isn’t a one-off opinion. This is her semantic territory.

And when Google’s AI Overview cited her, it didn’t just pull a sentence, it also named her business and linked to her post. It attributed the explanation to her as an entity.

(If you want to understand the mechanics of how AI systems actually decide what to cite — why some content gets chosen and other content gets ignored — I wrote a separate post that breaks down the retrieval process in detail.)

How the Definition of SEO Success Has Shifted

Traditional keyword-first SEO measured:

• Rankings
• Traffic
• Click-through rate

AI-mediated search requires new metrics:

• Are you being cited? Not just ranked, but actually named as the source.
• Are you being recognized? Do people see your business name when they search?
• Are you being remembered? When the same question comes up again, does the AI system return to you as the answer?

This is not about gaming a new algorithm. This is about becoming the kind of source that AI systems want to cite — because your explanations are clear, credible, and reusable.

The Real Question to Ask Before Publishing Anything

“If an AI needed to explain this to someone tomorrow — would it choose me?”

If the answer is no, the issue isn’t your keywords. It’s your clarity, your confidence, and your credibility.

Ask yourself:

• Did I answer the question in the first 200 words?
• Are my headings clear enough that an AI system could identify which section answers which query?
• Have I made my positioning explicit, or am I hedging?
• Is this explanation reusable, or does it require narrative context to make sense?
• Have I reinforced this explanation anywhere else — podcast, newsletter, interview, another post?

If you can’t say yes to most of those, you’re not optimizing for how information actually surfaces anymore. 

What Happened to Bev

Even though Bev still doesn’t rank #1 in the traditional sense, she became the answer.

And here’s what that means for her business:

• People searching that query now see her as a credible source
• Her entity recognition just got stronger
• The next time someone asks an AI system a related question, she’s more likely to surface again
• She’s not just accumulating traffic, she’s earning attribution as a credible expert

Pages are no longer the unit of measurement. Answers are.

And if you’re still optimizing for pages, it’s time to rethink how you’re organizing your content.