How I Used Query Fan-Out to Build an AI Search Content Strategy for Aesthetic Brands

Last week, in the first edition of *Thursday Morning by VicAILab*, I wrote that AI Search starts before the click.

More links inside AI Search matter. But the deeper question is not whether a brand can be clicked. The deeper question is whether a brand can be understood, selected, summarized, and trusted before the click happens.

This week, I want to show what I built when that idea stopped being abstract.

Not a universal framework. Not a guarantee.

A working hypothesis — tested, refined, and still evolving.

For years, content strategy started with keywords.

Find the keyword. Build the page. Optimize the page. Wait for the ranking. Hope for the click.

But AI Search changes the architecture.

Keywords still matter. But in AI Search, they are the starting point, not the architecture.

One visible question can trigger many hidden sub-questions. One user intent can expand into definitions, comparisons, risks, safety concerns, alternatives, evidence, expectations, objections, and trust signals.

That is why query fan-out matters.

The question I kept returning to was simple:

”If AI expands one query into multiple sub-questions before building an answer, what does that mean for the content an aesthetic brand actually needs to create?”

Query fan-out gave me a name for something I had been circling for a long time.

And once I had the name, the content strategy became clearer.

By the end of this edition, I want you to see one thing clearly:

• In AI Search, a content strategy is no longer only a publishing calendar.

• It is a structured map of the questions an AI system may ask before it decides what to show, what to cite, and what to ignore.

One Signal

In May 2025, Google gave this shift a clear technical name: query fan-out.

In AI Mode, Google says it uses query fan-out by breaking down a question into subtopics and issuing multiple queries at the same time on the user’s behalf.

Google Search Central later made the point even clearer: both AI Overviews and AI Mode may use query fan-out, issuing multiple related searches across subtopics and data sources to develop a response.

That matters because query fan-out is not just “more search.”

It means that one visible query can be decomposed into several hidden sub-questions before the answer is built.

Some of those sub-questions may be semantically close to the original query.

Others may look semantically distant.

But they may still be very close to the real intent behind the search.

A person may ask: “What is the best skin treatment before summer?”

But the system may need to explore several other questions before building a useful answer.

• What does “best” mean?

• Best for pigmentation?

• Best for acne?

• Best for redness?

• Best for skin tightening?

• Best for minimal downtime?

• Best for safety before sun exposure?

• Best for a patient afraid of side effects?

• Best for someone comparing clinics, devices, products, prices, evidence, expectations, and trust?

That is the shift.

The visible query is only the surface.

The real strategy lives in the question space behind it.

This is why content strategy in AI Search cannot stop at keywords.

The old path was: keyword → page → ranking → click

The new path is different: keyword → question vectors → entity relationships → knowledge graph → answer eligibility

This does not mean keywords disappear.

Google itself says that the best practices for SEO remain relevant for AI features in Search.

So keywords still matter.

But in AI Search, keywords are no longer the architecture.

They are the entry point into a larger system of questions, entities, associations, evidence, and trust signals.

For aesthetic brands, this changes the work completely.

A content strategy is no longer only a list of keywords to target or topics to publish.

It becomes a structured map of the questions AI may need to resolve before it decides which brand, product, clinic, device, treatment, or expert deserves to appear.

One Interpretation

This is where query fan-out becomes more than a technical phrase.

For me, it changed the way I look at content strategy.

Because if AI Search can break one question into several sub-questions before building an answer, then the old content planning model becomes incomplete.

The old model was built around visibility.

• Find keywords.

• Create pages.

• Optimize titles.

• Publish content.

• Track rankings.

• Measure clicks.

That model still matters.

But it is no longer enough.

In AI Search, the system does not only look for a page.

It may look for:

• The best explanation.
• The safest clarification.
• The most credible source.
• The clearest comparison.
• The missing context.
• The limitation.
• The risk.
• The alternative.
• The evidence.
• The objection.
• The practical next step.

That means the content problem changes.

”You are no longer writing only for a user who types a query and chooses a link.”

• You are writing for a retrieval layer that may inspect your content before the user ever reaches your website.
• You are writing for an answer engine that may extract one paragraph, one claim, one definition, one comparison, or one caution.
• You are writing for a citation layer that may decide whether your page deserves to support the answer.
• And, with the right caution, you are also writing for future machine readability.

Not because anyone can guarantee that a specific article will enter a future training corpus.

That would be an illusion.

But because clear, stable, well-structured, well-sourced content has a better chance of being understood, reused, connected, and preserved by future AI systems.

This is why I no longer see content strategy as a publishing calendar.

A calendar tells you when to publish.

AI Search forces you to ask a harder question:

• What must exist so that the brand can be understood correctly when AI decomposes the user’s intent?

That is the real work.

When I build an AI Search content strategy for an aesthetic brand, I do not start only with keywords.

I start with the visible question:

• Then I ask what the hidden questions could be.
• Then I separate the questions by intent.
• Then I filter those questions through 30+ years in dermatology and 20+ years in anti-aging, because not every question carries the same clinical, reputational, or decision-making weight.

Some questions are informational.

• What is the treatment?
• How does it work?
• Who is it for?
• What results can be expected?

Some questions are comparative.

• How is this different from another treatment?
• How does this product compare with another product?
• What is better for pigmentation, redness, acne, skin laxity, texture, or prevention?

Some questions are safety-driven.

• What are the risks?
• Who should avoid it?
• What happens before summer?
• What happens after sun exposure?
• What are the contraindications?
• What should the patient ask before deciding?

Some questions are trust-driven.

• Who explains this?
• What evidence supports it?
• Is the source medically credible?
• Is the brand precise?
• Is the claim exaggerated?
• Is the information consistent across the website, social media, articles, product pages, FAQs, and expert content?

Some questions are decision-driven.

• What should I choose?
• What should I avoid?
• What should I compare?
• What should I ask the doctor?
• What is the next reasonable step?

This is where the keyword becomes only the beginning.

A keyword can tell you where the user starts.

But a question vector tells you where AI may need to go.

And a knowledge graph shows how the pieces connect.

The treatment connects to the indication.

The indication connects to the patient concern.

The concern connects to safety.

Safety connects to evidence.

Evidence connects to authority.

Authority connects to trust.

Trust connects to selection.

This is the real interpretation.

In AI Search, proximity is not only semantic. It is decisional.

A sub-question may look far away from the original query, but still be close to the real decision.

A user may search for “best skin treatment before summer.”

But AI may need to understand downtime, pigmentation risk, photosensitivity, melasma relapse, post-procedure SPF, skin phototype, contraindications, clinic expertise, product alternatives, and realistic expectations.

Those are not random branches.

They are the hidden brief behind the visible question.

The user writes a question.

AI builds a brief.

Your content has to be ready for the brief.

For aesthetic brands, this matters even more because they are high-interpretation brands.

• They do not only compete for attention.
• They compete for correct interpretation.
• A device can be misunderstood.
• A treatment can be oversimplified.
• A product can be reduced to one ingredient.
• A clinic can become invisible.
• An expert can be omitted.
• A strong brand can be flattened into generic category language.

And once AI builds that first interpretation, the user may never see the full story.

This is why the strategy must move from isolated content pieces to answer infrastructure.

• Not one standalone article.
• Not one isolated landing page.
• Not one disconnected campaign.

But a connected system of questions, answers, entities, evidence, comparisons, definitions, limitations, and trust signals.

That is what I actually build.

• Not content for content’s sake.
• Not SEO text with better vocabulary.
• Not a calendar filled with posts.

I build a map that helps AI understand what the brand is, what it does, what it should be associated with, what it should not be confused with, and why it deserves to be part of the answer.

This is the shift.

Content strategy for AI Search is not only about being found.

It is about being correctly understood before selection happens.

One Consequence

The consequence is simple.

“Aesthetic brands do not need isolated content pieces. They need topic campaigns built from multiple points of view.” — Victor G. Clatici, MD

In old content planning, a brand could say:

• “We need an article about pigmentation.”

Or:

• “We need a post about skin tightening.”

Or:

• “We need a landing page for this treatment.”

That is not enough in AI Search.

Because query fan-out may not evaluate only one page, one keyword, or one article.

• It may move across fragments.
• It may inspect a definition.
• Then a comparison.
• Then a safety clarification.
• Then an expert explanation.
• Then a FAQ.
• Then a product page.
• Then a clinic page.
• Then a scientific source.
• Then a patient-oriented explanation.
• Then a limitation or contraindication.

So the brand is not judged only by one beautiful page.

It may be judged by the coherence of the entire signal.

This changes the way a topic should be built.

A topic is no longer one article.

A topic becomes a campaign of perspectives.

• One piece may explain the problem.
• One piece may define the treatment.
• One piece may compare alternatives.
• One piece may explain safety.
• One piece may answer objections.
• One piece may clarify who is not a candidate.
• One piece may connect the topic to seasonality.
• One piece may explain evidence.
• One piece may translate the expert view for the patient.
• One piece may show the brand’s specific point of view.

Together, these pieces do not repeat the same message.

They build the same semantic territory from different angles.

That is the difference.

A normal campaign repeats.

An AI Search campaign reinforces.

• It reinforces the entities.
• It reinforces the associations.
• It reinforces the safety signals.
• It reinforces the evidence.
• It reinforces the expert perspective.
• It reinforces the brand’s position inside the question space.

“In AI Search, every unanswered question becomes a competitor’s entry point.” — Victor G. Clatici, MD

• If your brand does not explain downtime, another source may define it.
• If your brand does not explain safety, another source may frame the risk.
• If your brand does not compare options, another source may control the comparison.
• If your brand does not clarify realistic expectations, another source may create confusion.
• If your brand does not state what it should not be confused with, AI may make the connection for you.

And when AI has to guess, the brand loses control.

This is especially important for aesthetic brands.

Because aesthetic decisions are rarely simple. They involve beauty, fear, comparison, trust, timing, budget, safety, expectations, social proof, and medical credibility.

A person may search for one thing.

But the decision behind that search is layered.

That is why the content must also be layered.

Not longer for the sake of being long.

Not more content for the sake of volume.

But modular, extractable, answer-ready content.

If AI builds answers from fragments, your content must be built in answer-ready fragments.

• A clear definition.
• A precise comparison.
• A short safety explanation.
• A realistic expectation.
• A documented limitation.
• A credible expert statement.
• A simple next step.
• A strong FAQ answer.
• A consistent brand claim.

Each fragment should be able to stand alone.

But each fragment should also connect to the larger map.

This is where the question space becomes the strategic asset.

The most valuable asset is no longer only the content library.

It is the question map behind the content library.

For an aesthetic brand, the question map shows what AI may need to resolve before it can decide whether the brand deserves to appear.

• What is the treatment?
• Who is it for?
• Who should avoid it?
• What problem does it solve?
• What problem does it not solve?
• How does it compare with alternatives?
• What evidence supports it?
• What risks must be mentioned?
• What expectations should be corrected?
• What seasonality matters?
• What should the patient ask before deciding?
• What makes this brand different?
• What makes this source credible?

That map should guide the campaign.

Not the other way around.

This also means that beautiful content is not enough.

In AI Search, content must be structured enough to be understood, verified, extracted, compared, reused, and cited.

Aesthetic brands often invest in polished visuals, elegant copy, and attractive claims.

That still matters.

But AI Search needs more than decoration.

• It needs structure.
• It needs consistency.
• It needs evidence.
• It needs clear entities.
• It needs stable definitions.
• It needs visible expertise.
• It needs answers that are easy to parse and hard to misunderstand.

This is where many brands will be exposed.

• Not because they are weak.
• But because their signals are fragmented.
• Marketing says one thing.
• Medical says another.
• Legal removes the useful detail.
• SEO optimizes for a keyword.
• Social media simplifies the message.
• The product page says less than it should.
• The FAQ does not answer the real fear.
• The expert voice is missing.

AI Search rewards coherence.

Fragmented teams create fragmented signals.

So the consequence is not “publish more.”

The consequence is: Build a question-led, evidence-backed, modular topic campaign.

• One topic.
• Many perspectives.
• One semantic territory.
• Many answer-ready fragments.
• One brand position.
• Many reinforcing signals.

That is how an aesthetic brand can start building content that survives query fan-out.

Not by producing random articles.

Not by filling a calendar.

Not by chasing keywords one by one.

But by building a structured campaign around the questions AI may need to answer before the user ever clicks.

One Move

Do not start with a content calendar.

Start with one strategic topic.

Choose one topic that matters for your aesthetic brand.

• Not ten topics.

One.

• Pigmentation.
• Skin tightening.
• Acne.
• Facial redness.
• Healthy aging.
• Laser safety.
• Injectables.
• Photoprotection.
• Barrier repair.

Then do not ask only:

• “What keyword should we target?”

Ask:

• “What hidden brief could AI build around this topic?”
• What definitions would it need?
• What comparisons would it need?
• What risks would it need to clarify?
• What evidence would it need to trust?
• What objections would it need to resolve?
• What patient concerns would it need to address?
• What entities would it need to connect?
• What expert signals would it need to recognize?
• What brand position should it understand?

Then build the campaign.

• One topic.
• Multiple perspectives.
• Answer-ready fragments.
• Clear entities.
• Visible evidence.
• Consistent claims.
• Expert interpretation.
• That is the move.

Do not publish isolated content and hope AI connects the dots.

Build the dots so clearly that AI has less room to misunderstand them.

Because in AI Search, strategy is no longer only about what you publish.

It is about what your content helps AI understand before the click.

In AI Search, the brands that win will not be the brands with the most content. They will be the brands with the clearest answer infrastructure.

Victor G. Clatici, MD LLM Nutritionist | MedAIMark | VicAILab AI-ready authority.

Zero mistakes. By design.

Bucharest, Romania May 21, 2026

The article was first published on LinkedIn.