Is ChatGPT recommending my business?
The quick answer: ask ChatGPT the questions your buyers ask, not your brand name, several times each, and count how often you are named. If you only ask once, or you ask "what is <my brand>", you will learn very little. Here is how to do it properly, and how to get a quick read for free.
The mistake most people make
Asking an assistant about your own brand by name tells you what it knows about you, not whether it recommends you. Buyers do not type your name; they type the problem. The question that matters is "which tool should I use for…", and whether you appear in the answer.
The second mistake is trusting one answer. Assistants generate a fresh answer each time. You might be named once and missing the next three times, or the reverse.
A do-it-yourself method
You can get a rough picture in an hour:
- Write five to ten questions a buyer in your category would ask, from broad to specific.
- Ask each one three to five times, in a fresh chat each time so earlier answers do not influence later ones.
- Record which brands each answer names, in order.
- Count the share of answers that name you, and note which competitors appear most.
- Repeat the same exercise on Perplexity and Google, because they often disagree with ChatGPT.
Reading the result honestly
With small samples, differences are noisy. Being named in 2 of 10 answers versus 4 of 10 is not a clear difference. That is why Shortlist reports a 95% interval next to every rate. If your interval overlaps a competitor's, treat it as roughly level.
Also look at position. If you are named but always last of six, buyers skimming the answer may not get that far.
Faster: the free check and the leaderboards
The free check runs three buyer questions from your category on Google AI Overview, Google AI Mode and Perplexity and shows each answer's brands, with yours highlighted. If your category has a public leaderboard, you can see your mention rate across all six engines, based on hundreds of answers, and read every one of them.
If the result shows you are missing, the next question is why. That is what the $390 audit answers: which sources the assistants rely on for your category and what you can change.
Choosing the right questions
The questions decide the result, so choose them with care. Include the broad question a newcomer would ask, the specific questions a buyer asks once they know what they need, and at least one “alternative to” question naming the market leader. Use the words your customers use, not your internal product terms. If you sell in several markets, write questions for each, because answers can differ by country and language.
Write the questions down before you ask them, and keep them. If you change them every time, you cannot compare one month with the next.
See it for a real category
- AI recommendations for crm for small business
- AI recommendations for whatsapp business api provider
- AI recommendations for email warm-up tool
- How Shortlist measures
Questions
Should I test in my own ChatGPT account?
You can, but past conversations and saved memories can colour the answers. Use a fresh chat, or a logged-out session where possible, so the answer is closer to what a new buyer would see.
Why not just ask ChatGPT about my brand by name?
Because buyers ask about the problem, not your name. The useful test is whether you appear when they ask for options.
How many times should I ask each question?
At least three to five times, in fresh chats. Shortlist leaderboards use five repeats per engine.
Is the free check really free?
Yes. One check per email and per website per day, with no card needed.
What if I am named on one engine but not another?
That is common. Engines rely on different sources. The per-engine view shows where to focus.