Why does AI recommend my competitors instead of me?

Q: Why does AI recommend my competitors instead of me?

We diagnose five, and they need different fixes: the crawler cannot reach your pages, your content is readable but not quotable, your brand name does not resolve to a single entity, no independent source corroborates your claims, or a competitor is simply present in every source the model reads and you are not. Diagnosing which one applies is most of the work — the remedies do not overlap.

Order	Cause we check	Time to rule out
01	Crawlers blockedA bot rule returns 403, or the page needs JavaScript to show its text.	5 min
02	Nothing quotableReadable, but no paragraph survives being pulled out of the page.	15 min
03	Entity unresolvedYour name exists in several forms. None of them wins.	1 hour
04	Nobody corroboratesOnly your own site states your specification.	half a day
05	Competitor owns corpusYou are absent from the reading list, not from the market.	days

Cheapest first, because a diagnosis that starts at the bottom spends the budget before it reaches the answer. These are the windows we scope against, not measured averages — we have run no client audits, so there is no average to publish.

1. AI crawlers cannot reach your pages

GPTBot, ClaudeBot or PerplexityBot gets a 403 from your bot-management rules, or your product copy renders client-side and the HTML they receive is an empty shell. This is the cheapest cause to rule out — three curl requests — and no other cause is worth investigating until it is clear.

2. Your pages are readable but not quotable

The page is readable but no paragraph survives extraction. This is what translation can leave behind — grammatically clean, rhetorically smooth, and containing no sentence that makes a specific claim a model could quote. If you cannot find a two-to-four-sentence passage on your product page that would answer a buyer’s question on its own, a model cannot either.

3. Your brand name resolves to more than one entity

The brand appears as three or four different strings across your own properties: a romanised name, a legal entity in the footer, a marketplace seller name. A model that cannot resolve which entity it is describing takes the safe route and names a brand it can.

4. No independent source corroborates your claims

You state SPF 50. Only your own site says so. A model asserting a specification sourced solely to the party who benefits from it is taking a risk, and it may decline. Once sources you do not own state the same figure, the claim no longer rests on your word alone.

5. A competitor owns the corpus your category is answered from

The uncomfortable one. A competitor appears in every round-up, comparison page and forum thread the model reads for your category, so it is named by default. This is an absence from the corpus rather than a verdict on the product, it is the slowest cause to fix, and it is the constraint that takes months rather than weeks.

Seven surfaces one answer is assembled fromillustrative

Your siteyours to write
Amazon listingyours to correct
Retailer pagesyours to correct
Review round-upssomeone else's
Reddit threadssomeone else's
YouTube reviewssomeone else's
Korean reviewswritten, wrong language

How many you write

1/7is a page you controlThe other six decide the shortlist.

You write one of these seven. Two more you can get corrected. Four are written by other people, and an English answer is assembled almost entirely from English sources — which is why the last row matters: a decade of reviews already exists in Korean, and the answer your US buyer reads was written without any of it. Which rows apply to you is what an audit establishes. The last four rows are the work.

How to tell which applies

Check them in order, because they get progressively more expensive to diagnose. The times below are what we scope against, not measured averages.

Cause	How to check	Time
1. Access	`curl -A "GPTBot"` against three key pages	5 minutes
2. Liftability	Read one product page and try to find a quotable paragraph	15 minutes
3. Entity ambiguity	Search your brand name across your own properties and listings	1 hour
4. Corroboration	Find every external page stating your specification	Half a day
5. Corpus ownership	Read the sources engines cite for your category	Days

The last two require looking at what the engines actually cited, which is why our free report lists the sources alongside the brands.

Buyer question	Engine	Named
best Korean sunscreen for oily skin	ChatGPT	not named
K-beauty sunscreen that does not pill	Claude	named
Korean skincare brands sold at Ulta	AI Overviews	not named
…17 more questions, with the verbatim answer and every source cited