In Recro simulations, brands appeared in 33% of answers when buyers asked about them by name; in decision questions that didn’t name them, the rate fell to 5.5%. The gap comes from whether the buyer already knows the brand. A buyer who knows you types your name to confirm. A buyer who doesn’t yet know you asks about the category, and if you’re not in the answer, never meets you. Growth is won in the second person’s question.
The test marketing teams run most often is typing the brand’s name into AI: “What is [Brand] and what does it do?” If the answer is accurate and detailed, everyone relaxes. That test confirms one thing: the model knows the brand. It doesn’t answer the commercially important question: does a buyer who doesn’t know the brand yet come across it when asking about the category?
How big was the gap between the two question types in our simulations?
Recro’s simulations for 6 brands between April and September 2026 included two kinds of questions. The first were questions buyers asked along their decision journey without naming the brand, such as “Which reliable coffee suppliers serve the food sector?” The second were questions in which the buyer researched a specific brand by name.
| Question type | Runs | Brand mention rate |
|---|---|---|
| Decision-journey questions without the brand name | 348 | 5.5% |
| In-depth comparison questions | 60 | 18.3% |
| General category research | 22 | 27.3% |
| Research questions naming the brand | 30 | 33.3% |
The closer the brand’s name gets to the question, the higher visibility climbs. But the questions where buyers build the shortlist are the ones in the top row: no brand name yet, only a need and a few conditions.
The same pattern shows up even in a highly visible brand
The most visible brand in our simulations was a corporate services brand. It was named explicitly in 75% to 88% of answers. But one line in the report stood out: the model never identified the brand by description, that is, by its attributes without its name. This “semantic recognition” rate was 0%.
In practice, the model can describe the brand to a user who already knows its name. But it doesn’t match a description of the brand’s specialty and the type of client it serves to that brand on its own. The brand’s place within its category stays hidden behind its name.
The key person: the buyer who doesn’t know you yet
A keyword approach can’t see this distinction. Ranking first for branded keywords shows that people who already know you can find you. Growth comes from people who don’t: a procurement director entering a new sector, an operations manager wanting to switch suppliers, a CFO who hasn’t heard of you from a peer.
That’s why Recro builds measurement around the key person. The persona is a buyer entering the decision process without knowing the brand. Questions grow out of that person’s need and conditions. The brand’s name enters those questions only if the model recommends it on its own.
Why do you disappear in category questions?
- Your category definition doesn’t match the buyer’s language. The brand describes itself in its own words; the buyer asks about their problem in theirs. The model can’t build a bridge between the two.
- The brand’s attributes aren’t tied to its name. When sector, scale, specialty and difference statement aren’t written together consistently, the model can’t match the description to the brand.
- You’re missing from category lists. In discovery questions, the model relies on industry lists and comparison articles. In our simulations, brands that were never mentioned had one thing in common: they were absent from the lists their competitors appeared in.
- Your proof lives only on your own site. When the model finds the brand by name, it reads the brand’s own site. In a category question, it looks for independent sources. In Ahrefs’ analysis of 75,000 brands, the number of web mentions showed the strongest relationship with AI visibility.
This pattern is the clearest signature of an Authority GAP. We explained how the shortlist forms in these category questions in the invisible shortlist.
What does it take to be recognized by description?
- Write the brand’s description sentence. “[Brand] is a [category] firm offering [specialty] to [scale] companies in [sector].” This sentence should stay the same on the site, in profiles and in press materials.
- Build category pages. Pages that bring the brand’s name and attributes together with the category names buyers use.
- Get into category lists. Map the lists, directories and comparison articles cited in discovery questions.
- Change your test. Instead of asking about the brand by name, ask the buyer’s category question and see whether the brand comes up on its own.
How does Recro make this gap visible?
The Recro Insight Model measures a brand’s recognition by name and its recommendability within the category separately. The gap between the two numbers shows the brand’s room to grow.
One simulation · 4 key pillars · 90+ steps
What sets Recro apart is four pillars built from scratch for each brand. Together they form a single simulation of 90+ steps, with each pillar producing the input for the next.
- Persona builder: Decision-maker profiles that represent the brand’s real buyers, specific to its sector and sales structure.
- Question builder: The brand-specific decision questions these people ask AI along their buying journey.
- Report builder: A report that reads the answers against the brand’s goals and competitors, together with the sources behind them.
- Action recommendation builder: A prioritized list of actions that closes the sector-specific signal gaps.
For this gap, each pillar produces the following:
- The persona builder sets up the key person who doesn’t know the brand yet: a buyer at the start of a decision with only a need and conditions.
- The question builder produces decision questions in the buyer’s language that don’t contain the brand’s name; questions that name the brand are kept separate.
- The report builder reports explicit mentions and recognition by description separately and shows the order in which the brand enters the answer.
- The action recommendation builder prioritizes the content and source steps that tie the brand’s attributes to its name.
From a Recro simulation · anonymized
- Brand: Corporate services brand · Turkish
- Named mentions: Named explicitly in 75–88% of answers.
- Recognition by description: 0%; the model never matched the brand by its attributes without its name.
- Diagnosis: Proof sits mostly on the brand’s own site; third-party sources are weak.
- Class: Authority GAP + Communication GAP
- Action: A consistent description sentence; category pages; attribute-based mentions in industry lists and publications.
Executive summary
- In Recro simulations, brands appeared in 33% of answers when asked by name and 5.5% in decision questions that didn’t name them.
- Even the most visible brand was never recognized by description; its visibility worked only for users who already knew its name.
- Growth is won in the category question of the key person who doesn’t know the brand yet; measurement and content should be built around that person.
Frequently asked questions
We get an accurate answer when we ask about our brand by name. Isn’t that enough?
It shows the model knows the brand. It doesn’t show whether the brand is recommended in the category question of a buyer who doesn’t know it. That second question decides growth.
What is semantic recognition?
The model matching the brand on its own in a question that describes its attributes without naming it. It shows the model has learned the brand’s place within the category.
How long does it take to close this gap?
A description sentence and category pages can be published within weeks. Attribute-based mentions in third-party sources take months; the effect shows earlier in search-grounded answers.
If branded search traffic is rising, are we fine?
It shows that people who know you are looking for you. Whether new buyers meet you in category questions needs to be measured separately.
What data are these rates based on?
Recro simulations for 6 brands with 460 question-answer runs. The data is anonymized and should be read as an observed pattern, not a sector average.
When a buyer who doesn’t know you yet asks about your category, is your brand in the answer? Recro builds a sample simulation with brand-specific personas and shares the result in a demo report.
Request a demo insight report →
Sources
- Recro simulation data · April–September 2026 · 6 brands, 460 question-answer runs, 73 reports (internal, anonymized; run on a single AI model)
- Ahrefs · An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied)
This article was prepared with AI assistance and published under the review of Mehmet Semih İpek.
