Does AI Name Your Sub-Brands, or Just Your Company?

1. Legion, or Just Lenovo?

Ask an AI assistant "what is the best gaming laptop" and it might say "Lenovo also makes gaming laptops." Technically that is a mention. Commercially it is a near-miss, because a competitor got recommended by product name — "get the ASUS ROG Zephyrus" — while your brand got recommended by company name. The buyer wanted to hear "Legion", the exact product they can go and search for, and the answer never handed them that name.

The same gap exists in every portfolio: does AI name the specific product line, or only the parent brand? Answer-level visibility metrics cannot see this distinction — to them, any mention of your company counts. The Sub-Brand Portfolio is the drilldown built to separate "AI named my company" from "AI named my specific product." For most brands, the gap between those two is large, quiet, and losable.

2. What the Sub-Brand Portfolio Shows

The panel is a matrix. Down the side are the brand keywords you configure — your product names, sub-brands, and feature brands (for a laptop maker, that might be ThinkPad, Legion, and Yoga). Across the top are prompt categories, so you can see whether AI names your products in "comparison" questions but not in "how-to" questions.

The data comes straight from the answers: every stored AI response records which exact keyword matched in the text. Each cell then tells you how often that specific keyword was named in that category, and in how many distinct prompts. True to the receipt principle that runs through Citation Analytics, click any cell and you drop into the prompts behind it — the actual answers where AI did (or did not) say your product's name. (Like every drilldown, it reads the latest run of each prompt per model inside your selected time range.)

Two accuracy notes up front:

• You define the keywords. The matrix only appears once brand keywords are configured. If the panel is missing, that is the first setup step, not a bug.
• The category columns are an Enterprise capability. Splitting the matrix by prompt category (via prompt Tags) is available on Enterprise. On other plans the matrix shows a single "All prompts" column, which still surfaces the most important row of all — the one we turn to next.

3. The "Generic (No Keyword Named)" Row Is the Whole Story

At the bottom of the matrix sits a row labelled "Generic (no keyword named)." It counts every answer that mentions the parent brand without naming any specific product line. Mechanically, when AI says your brand, Ayzeo checks the wording against your keyword list; anything that does not match a tracked product falls into this generic bucket.

For many brands with a rich portfolio, the row looks something like this (illustrative numbers):

AI talks about the company constantly and names a specific product essentially never. For a company with a rich product portfolio, that is a precise diagnosis: the brand is known, the products are not. Every one of those generic mentions is a buyer who heard the company name but was never handed the specific product to search for, ask about, or buy.

A high Generic count is not a failure — it is a map. It says: the brand equity is already there; you just have not taught AI the product names yet.

4. AI Recommends Entities, Not Companies

Here is why the generic gap costs deals rather than just looking untidy. AI models recommend entities — specific, named things they have learned exist in the world. "Legion" is an entity; "a gaming laptop from Lenovo" is a description. When your product line is not an entity in the model's world, you cannot win the specific-recommendation slot: the model has nothing precise to name, so it either falls back to your company name or reaches for a competitor's product that is a known entity.

That is the real, losable difference no rate metric captures. "Get the ASUS ROG Zephyrus" and "Lenovo also makes gaming laptops" score identically as one mention each, but they send the buyer to two very different places: one to a product page and a price, the other to a homepage and a decision they still have to make. Specificity also reads as confidence — the precisely-named product looks like the surer choice.

5. Reading the Matrix by Prompt Category

On Enterprise, the category columns turn the matrix into a targeting tool. The pattern you are looking for is where the generic mentions concentrate:

• Generic high in "comparison" prompts — AI is comparing you against rivals but describing you at the company level while it names competitors' specific products. This is a direct competitive disadvantage.
• Generic high in "how-to" or "use-case" prompts — AI knows your brand exists but does not associate a named product with the job the buyer is trying to do. That is a content gap: the product-to-use-case link is missing from the sources AI reads.
• A specific sub-brand strong in one category, absent in others — you have earned recognition for that product in one context; the opportunity is to extend it into adjacent questions.

Because every cell drills into its prompts, you never have to guess. Click the generic cell in the category that matters most to your pipeline and read what AI actually said instead of your product name.

6. How to Close the Generic Gap

A high Generic count is one of the more tractable problems in GEO, because the brand awareness is already won. The work is entity-building — teaching AI, and the sources AI reads, the product names.

1. Configure your keywords first. List every sub-brand, product line, and feature brand you want AI to name. Without them, the matrix cannot separate generic from specific.
2. Name the product line consistently, everywhere. In titles, comparisons, and spec pages, stop writing "our platform" and write the product name. AI mirrors the language of the pages it reads; if your own site is generic, the answers will be too.
3. Fix the source layer, not just your homepage. Pair this with the Sources & Destinations drilldown to see which third-party pages AI consults for the generic prompts, then get your specific product named on and around those pages — reviews, comparisons, community threads, docs.
4. Structure the product-to-use-case link. Clear headings, comparison tables, and structured data that explicitly tie a named product to the job it does give AI an unambiguous entity to reuse. See our citation strategies guide for the on-page tactics.
5. Re-check the row monthly. As the specific rows climb and Generic falls, you are watching brand equity convert into product equity in AI answers — a trend worth putting in the board deck.

7. Frequently Asked Questions

Q: What does the "Generic (no keyword named)" row mean?

A: It counts every answer that mentions the parent brand without naming any of the specific sub-brands or products you track. A high Generic count means AI knows your brand but not your individual products — a clear, fixable content and PR opportunity.

Q: How does Ayzeo decide whether a mention is generic or a specific sub-brand?

A: Every stored AI answer records which exact keyword matched in its text. If the wording matches a tracked product name, it counts toward that product's row; if it only names the company, it falls into the Generic bucket.

Q: The Sub-Brand Portfolio is not showing. Why?

A: The matrix only appears once brand keywords are configured. Add your sub-brands and product names, and it populates on the next analysis run.

Q: Can I see which sub-brands AI names in comparison prompts versus how-to prompts?

A: The category (prompt-tag) columns are an Enterprise capability. On other plans the matrix shows a single "All prompts" column, which still exposes the Generic row and every per-keyword count.

Q: Why does naming the product line matter more than naming the company?

A: AI models recommend entities — specific named things. If your product line is not an entity in the model's world, you cannot win the specific-recommendation slot, and the model falls back to your company name or a competitor's named product. A named product is also a searchable next step; a company name is a homepage and a decision still to make.

Q: How do I move mentions from Generic into a specific product row?

A: Name the product line (not just the company) consistently in your own content and in the third-party pages AI consults for those prompts, and structure the product-to-use-case link with clear headings and data. Then track the row over time as Generic falls and the product rows climb.

Key Takeaways

• Company mention ≠ product mention: the Sub-Brand Portfolio separates "AI named my company" from "AI named my specific product."
• The Generic row is the headline: a high count means strong brand awareness but unrecognised products — a fixable, high-value gap.
• AI recommends entities: if your product line is not a named entity, you lose the specific-recommendation slot to a competitor that is.
• Category columns target the fix (Enterprise): find where generic mentions concentrate and aim your content there. Every cell has receipts — click through to the real answers.
• Closing the gap converts equity: teaching AI your product names turns brand awareness into searchable, higher-margin product demand.