Among the 55 brands scanned on LensCore in September 2026, 39 were named by none of the five AI engines. For 21 of those (53.8%) a relevant competitor was named instead. For the other 18, no brand was named at all — the engines answered the buying question without recommending anyone. Those two situations look identical on a scorecard and call for opposite responses.
This reads the same September 2026 scans as our visibility study, through a different lens, and comes out of the same script — scripts/september-visibility-study.mjs. We name competitors below because that is what the engines said. We do not name any brand that was scanned.
Being absent and being beaten are not the same problem
| Among the 39 brands no engine named | Brands | Share |
|---|---|---|
| A relevant competitor was named instead | 21 | 53.8% |
| No brand was named at all | 18 | 46.2% |
If a rival was named and you were not, the engine had an answer and chose somebody else. The category has an established answer, and you are competing against a specific name for a specific question.
If nobody was named, the engine answered the question generically — advice, criteria, a category explanation, no brands. That is a different and much better position to be in. Nobody owns the answer yet.
Where a rival existed, it usually won
Across every brand with a detectable relevant competitor — 32 of the 55 — the top rival was named more often than the brand itself in 27 cases, or 84.4%. That figure covers all brands, not only the invisible ones: being named sometimes is not the same as being named more than your competition.
126 distinct rivals were named in total across the 55 scans. Most appeared once. A handful recurred across several unrelated brands’ question sets:
| Rival named | Appeared in this many brands' scans |
|---|---|
| boAt | 11 |
| Lenskart | 7 |
| H&M | 3 |
| Mamaearth | 3 |
| Zara | 2 |
| Semrush | 2 |
A name recurring across several different brands’ questions is doing something our data can see but cannot explain: it has become a default answer in its category. We did not test why, so we are not going to tell you why.
What this does not show
The same sample caveats apply. These are the brands that chose to run an AI-visibility scan, not a random sample of anything, and each engine was queried once per scan. 53 of the 55 brands were measured with 8 questions and 2 with 25.
Only 32 of 55 brands had a detectable relevant competitor. The 84.4% figure is of those 32, not of all 55. Our rival extractor also returns adjacent names it is not confident about, and those are excluded from every number on this page.
An empty category today is not an empty category tomorrow. The 18 brands whose questions named nobody were measured in one month. That window closing is exactly the thing worth watching, and it is not something a single scan can tell you.
Nothing here is causal. We did not run an intervention. We cannot tell you that publishing a comparison page would have moved any brand from the 39 into the named group.