Guide: Why does ChatGPT recommend our competitor?
The five reasons a competitor gets named in your place, and how to tell which one it is.
The short answer
In two paragraphs.
Because, for that question, the assistant found more to say about them. ChatGPT writes each answer from what it reads at that moment: the websites it can open, the reviews, lists and forum threads it searches, and the way the question was asked. A competitor who is named and you are not is ahead on one of five things, and the audit's first job is to say which.
The five: their website can be read and yours cannot; the sources the assistant cites for your market name them and leave you out; their pages answer the question your pages leave open; the budget or place in the question favours them; or you asked one assistant and your customers use another. Each has a different fix, and only the first is quick.
The five reasons
Which one is it? Each leaves a different trace in the answers.
| Reason | What you see in the answers | What changes it |
|---|---|---|
| Your website cannot be read | You are never named, for any question, and your own pages are never cited | Fix the basics: robots, access, words the script builds, speed. See the guide on readability |
| The sources name them | The citations under the answer are reviews, lists and forums where they appear and you do not | Be present in those sources: reviews, listings, trade pages, the places your market reads |
| Their page answers the question | The answer repeats a sentence from their page: a price, a size guide, a comparison | Write the answer on your own page, in plain words, with the figures |
| The question's framing favours them | You are named for some budgets or places and absent for others | Decide which questions to win, and say so on your pages |
| Another assistant names you | ChatGPT names them, Gemini or Perplexity names you | Measure all three and work on what they share |
The framing
A brand can own one question and be eleventh on the next.
In our perfume study in the United Arab Emirates, Lattafa was named on 67.2% of the 12 low-budget questions, first of all brands. On the 54 questions with no price limit it ranked eleventh, with 6.3%. On the 10 premium and niche questions Amouage led with 52.3%. The same three assistants, the same day, the same market. A competitor who seems to be recommended everywhere is usually recommended for a kind of question, and the kind can be named.
The assistants differ too. Across the 76 brand-free questions in that study, ChatGPT named Amouage most, Gemini named Lattafa most, and Perplexity named Ajmal most. Checking one assistant tells you about one assistant.
The shop and the shelf
Being named as a place to buy is a different thing from being named for a product.
In our running-shoe study in Lebanon, 342 answers came back from three AI systems across 38 buying decisions. Fewer than half of them, 149 of 342, named any shop at all. Mike Sport, a retailer, was named as a place to buy more often than any other shop, 52 times against 48 for Decathlon. In 12 of the 342 answers the system also recommended a product Mike Sport sells and named Mike Sport as a place to buy it. The retailer was known; the connection between its shelf and the shoes was rarely made.
So when a competitor is recommended, look at what for. Named as a shop, named for a product, or named for both: each is a different gap, and the audit reads all three.
ChatGPT says something wrong about us. Can we correct it?
Indirectly. The assistant repeats what it reads, so the correction goes on the sources it reads: your own pages first, then the listings and reviews that describe you. The audit's information consistency check lists where your products, prices and claims disagree with each other.
Our competitor is bigger. Can we still be named?
For some questions, yes. The perfume study ranked 209 brands, stores and shopping places across 76 questions, and no brand held a fifth of the shelf. Size wins the broad questions; the specific ones, a budget, a place, a problem, are won by whoever answers them.
How do we find out which reason it is for us?
Ask the questions your customers ask, of all three assistants, several times each, and read the citations. Senpai.ai's AI search audit does that for one market and ranks what to change first. The question map on this site is the first step, in a minute.
The question map
What are your customers asking AI?
They ask ChatGPT what to buy and who to trust before they ever reach you. Type your website and see the questions they ask, and who AI names instead of you.
People ask AI things like
- Where can I try on running shoes in Dubai today?
- Which running shoes stay cool in very hot weather?
- How do I know my running shoe size?
Find what's worth changing next.
Bring the market, the product or the question. We'll start there.
Or start with the question map: see what your customers ask AI.