Strategic AI Search Approach: The Strategic Questions Behind AI Search and How to Answer Them

Traditional analytics were built for a world of keywords and rankings. You typed a query, a list of results appeared, and your job was to move up the list. Rankings, clicks, and conversions followed in a line you could measure.

Now a growing share of people ask ChatGPT, Perplexity, or Google's AI Mode instead. They get one written answer, sometimes with a few cited sources, and often never visit a results page. So brands face a plain problem: how do you measure what you cannot see?

This guide is organized around the strategic questions that problem raises. Is SEO still worth the investment? 

How does AI visibility connect to the rest of marketing? How do you measure it, and how do you know the measurement is trustworthy? What is AI traffic actually worth? How do you size demand when there is no keyword volume? 

The answers come from the thinking of Baruch Toledano, Similarweb's VP of Solutions, in a February 2026 article laying out how the company approaches AI search. This guide works through those answers and what they mean for a marketing team.

Strategic AI Search Approach

The source's position comes down to a handful of ideas. AI search is a structural shift in how people look for things, not a feature to bolt on. 

SEO is evolving, not dying, and deserves more investment, not less. AI visibility is fragile if it is not supported by real brand demand, so it has to be read alongside branded search, traffic, and conversion. 

Measurement is only worth trusting if it is built on the questions real people ask and the responses chatbots actually give. 

AI traffic is small in volume but unusually high in intent. And demand in AI should be sized by topic, not by keyword, because chatbots are used conversationally. The rest of this guide takes each in turn.

Is SEO Dying, or Changing?

The source's answer is direct: it is doubling down on SEO, not slowing down. SEO is not disappearing; it is evolving. The fundamentals still matter: fresh data, site audits, rank tracking, and keyword research. What changes is that these tools need to show the effects of AI.

Two examples make this concrete. In rank tracking, you can now see AI Overviews directly in the search results landscape, so you know which results pages carry an AI-generated answer on top. 

In keyword research, you can see whether a keyword triggers an AI Overview, how much zero-click behavior surrounds it (people getting their answer without clicking), and how AI is affecting traditional search demand. The point is that your SEO strategy adapts to AI disruption instead of being blindsided by it.

There is a second pressure worth noticing. The source points to research by Aleyda Solis suggesting that SEO is losing clicks not only to AI but also to paid results, as search pages are monetized more heavily. 

In a market where organic clicks are shrinking from two directions, it argues that paid search needs closer protection too: controlling spend and being alerted when competitors bid on your own brand terms.

For a team deciding where to put effort, the practical reading is that AI search does not replace the SEO foundation. It sits on top of it, and the foundation needs to be watched more closely, not abandoned.

How Does AI Visibility Connect to The Rest of Marketing?

The source makes two compact points. "AI visibility without brand demand is fragile. Traffic without conversion insights is incomplete."

The first half means that being mentioned by an AI engine is not a standalone achievement. If people are not searching for or choosing your brand elsewhere, AI mentions rest on thin ground. 

The second half means that counting visits tells you little until you know what those visitors do. So AI search does not replace fundamentals like branded search, loyalty signals, conversion analysis, and your channel mix. It joins them.

The source also draws a distinction between treating AI as a feature and treating it as a structural shift in search behavior. A feature is something you add to a dashboard. 

A structural shift means you review how every channel behaves, because in many cases the channels complement each other along the same customer journey. 

Someone may meet your brand in an AI answer, search for it by name later, read a review, and convert from a paid ad. Looking at only one of those steps gives a distorted picture.

The source describes this as the difference between reacting to AI and building for it.

What Should You Measure?

The source treats AI visibility measurement as the area where the market is demanding clarity. Its list of what a brand needs to understand is useful whichever tool you use: where you appear in AI-generated answers, who is cited instead of you, how citation patterns change over time, and how your visibility shifts across the whole landscape of topics you care about.

Two terms are worth defining plainly. A mention is your brand appearing in an answer. A citation is an answer pointing to a source, such as your page or someone else's page about you. They are not the same thing, and tracking both tells you more than either alone. The source also puts weight on sentiment: whether the way you are described is positive, neutral, or negative.

It describes its measurement approach with three pillars.

Clarity means clear measurement of presence in AI environments, so you can say simply whether you are in the answer or not. 

Real user data means measurement grounded in actual usage instead of assumptions about what people ask. Actionability means deep citation insight, so a team knows what to fix and where to compete, not just a score to report.

That third pillar is the one teams often miss. A visibility number that cannot tell you which source is beating you, or on which question, does not tell you what to do next.

How Do You Know the Measurement is Trustworthy?

There are already dozens of AI brand visibility tools. The source says its priority was not to launch another one but to build something reliable, grounded in strong data, and useful. It reports hundreds of customers tracking visibility across multiple AI engines.

Its clearest argument is about sentiment. Some providers, it says, simply ask a chatbot, "What do you think about this brand?" and report the reply. 

That can be interesting, but brands can do it themselves, and it does not reflect how real buyers meet your brand. A single self-directed question tells you how a model answers one prompt you chose. It says little about how it talks about you when thousands of real people ask real questions.

The approach it describes instead has four parts. Track the topics and prompts that matter to a brand, based on real user data. 

Record every response, every day. Analyze those responses properly, judging sentiment in the context of both the question and the answer. And keep refining how sentiment is read, because the same sentence can be positive for one question and neutral for another.

You can apply this standard to any tool you evaluate. Ask where the prompts come from, how often responses are collected, and how sentiment is judged. The discipline behind the data, not the dashboard in front of it, is what makes a number worth acting on.

The source also notes that this data can show how AI platforms themselves are changing. As an example, it points to a surge in Gemini's traffic that the market was not necessarily predicting. Knowing which platforms are rising and falling tells you where to look for your audience.

What is AI Traffic Actually Worth?

The source is candid that first-party measurement is imperfect, but useful for seeing how much traffic reaches your own site from different AI sources. 

It adds an outside view: traffic from AI chatbots to any site, including direct competitors, with trends over time so you can see who is gaining or losing share, side-by-side comparisons, and a breakdown of which landing pages attract AI visitors. 

For an ecommerce brand, that could show whether a competitor's AI traffic lands on blog content, category pages, product pages, or the homepage. It also shows the likely prompts that led people there, so you see the journey, not just the visit.

At the time of writing in February 2026, the source says ChatGPT drives most of that traffic, though that could shift as Google's AI Mode expands.

The important point is the framing. The primary reason to be visible in AI search is not traffic, because there is simply not much traffic to be won. But the traffic that does exist tends to be very high intent. 

The source says Similarweb's own studies show that in markets like US retail, traffic from generative AI sources converts at a higher rate than any other channel, because these visitors arrive with a clear purpose. Treat that as Similarweb's finding, and check it against your own conversion data.

This changes what you aim for. If you judge AI search by volume, it looks like a rounding error. If you judge it by intent and influence, it looks like a place where buyers form opinions before they ever click. 

The sensible approach is to track it, find the landing pages that win, learn from competitors' successes, and optimize for the demand that is available.

How Do You Size Demand When There Are No Keywords?

This is the open problem the source says it is working on. Many tools have tried to copy traditional keyword search volume for AI. 

The source argues that AI does not work like Google. People talk to chatbots conversationally and with context, in long, specific requests, so the neat one-to-one link between a keyword and a monthly search count does not hold.

Its alternative is to use real user data to assess demand across topics instead of keywords. That is a more nuanced way to see where attention sits in AI-driven discovery. The source frames this as work still in progress, saying more is to come, so treat it as a direction instead of a finished answer.

For your own planning, the lesson holds even without any particular tool. Stop looking for a "search volume" number for each prompt. Group questions into topics, watch which topics gain attention, and judge your coverage at the topic level.

How Do You Do This With a Tight Team and Budget?

The source's answer is that it treats answer engine optimization (AEO) as a natural part of the broader marketing and SEO ecosystem, which is why most of its customers use its AI tools through marketing packages, with a self-serve option for those who want to test.

The broader point is not about any one price list. It is that AEO does not need to be a separate program with its own team. It sits alongside SEO, content, and analytics work, uses much of the same data, and is better planned together. 

The source also describes two ways of reducing the manual effort. One is an SEO agent that analyzes whole groups of keywords at once, shows the competitive landscape and your own performance, identifies actions, and can even draft content briefs. 

The other is making data available through MCP (a standard way for AI assistants to connect to data sources), which some SEO teams use for deep competitor research or for dashboards that combine traditional SEO metrics with AI visibility.

The second goal the source names is simplification. AI has added complexity to an already fragmented landscape, and the aim is to consolidate AI visibility, SEO performance, traffic, and wider marketing signals into one coherent story.

Putting The Questions Together

The table below is a synthesis of the source's positions, arranged as a working checklist. It is a summary of the reasoning above, not a framework from the source itself.

The New Search Reality

The source's closing argument is that the shift to AI-powered search is accelerating. Brands that treat it as a distant problem will find themselves invisible in the channels where their customers increasingly turn for answers.

But visibility alone will not win. Success depends on the full picture: where your brand appears in AI-generated responses, how those citations turn into traffic, what that traffic does on your site, and how all of it compares with your competitors. 

The strategic approach is to build for that picture on purpose, with real data and clear priorities, instead of reacting to each change as it arrives.

Frequently Asked Questions

Is Traditional SEO Still Worth Investing in Now that People Use AI Search?

The source says yes, and goes further: it is increasing its investment in SEO. The fundamentals of site health, rank tracking, and keyword research still matter, and they need to be adapted to show AI's effects, such as whether a keyword triggers an AI Overview. SEO is changing, not going away.

Why Not Just ask a Chatbot What it Thinks of My Brand?

You can, and it can be interesting. But it tells you how a model answers one question you chose, once. The source argues for tracking the topics and prompts that matter to your brand, based on real user behavior, recording the responses every day, and analyzing sentiment in the context of each question and answer.

If AI Traffic is so Small, Why Does it Matter?

Because the traffic that arrives is high intent, and because AI answers influence buyers before they click. The source cites Similarweb studies showing that in US retail, traffic from generative AI converts at a higher rate than any other channel. The primary reason to be visible, it says, is not volume.

Is There a Keyword Search Volume Equivalent for AI?

Not in the same sense. The source argues that people use chatbots conversationally and with context, so a one-to-one keyword volume model does not fit. It is working on assessing demand by topic, using real user data, and says more is coming on that.

Does AI Visibility Replace Brand Building and Conversion Work?

No. The source is firm that AI visibility without brand demand is fragile and that traffic without conversion insight is incomplete. It should be read together with branded search, loyalty signals, conversion analysis, and your channel mix.

Do I Need a Separate Budget for AEO?

The source treats AEO as a natural part of the broader marketing and SEO ecosystem, which is why most of its customers use its AI tools through marketing packages. In practice, that suggests planning AEO alongside your existing SEO and content work, though your own budgeting depends on your situation.

Related Searches:

• Black Hat GEO: Tactics, Risks, and Why It Never Wins
• How to Pitch Journalists: A Practical Guide
• Negative GEO: How Competitors Can Harm Your Reputation on AI