Your story, told by someone else
A brand used to be what you said about yourself, repeated by people who had met you. Now there is a third storyteller. An AI assistant answers a buyer's question with a short account of who you are, what you offer and how you compare, assembled from sources you mostly do not control.
That account is often accurate. Sometimes it is out of date, incomplete or set up so a rival looks like the better choice. The buyer cannot tell the difference, and you do not see the conversation.
What our study shows about how firms are represented
In our study of 240 AI conversations about money, four patterns showed how a firm's story gets told for it.
Your content can work for someone else. One firm's pages were cited in 22 Gemini conversations without the firm being named once. Claude did the same with several banks' explainer pages.
A missing fact gets filled in. When buyers asked what advice would cost, ChatGPT and Claude went looking for price pages and named the firms behind them. Gemini usually gave a price range with no page behind it. A firm without a readable price is either skipped or described by the average.
The category can be ruled out before any brand. Gemini told buyers that paid help was not worth it in 15 of 80 answers, and all three engines steered investors towards the same textbook choice.
Others tell your story. Claude leaned on intermediaries and comparison sites. Gemini named firms and prices from memory without flagging them. Some of the intermediaries were also put forward as the place to go.
Four ways the story goes wrong
When an engine names you, it can get you wrong in three ways, and there is one warning sign that it might.
Who can help us with an expat mortgage, and what does it cost?
AI answerYour firm mainly advises Dutch homeowners on refinancing.
Wrong. You specialise in mortgages for expats, and the line comes from an old directory listing.
Illustrative
- Misrepresented. What it says is wrong: an old price, a service you stopped offering, or another company's facts under your name.
- Undersold. A fact the buyer needs is missing: a price in a PDF, a certification on a page the engine cannot read.
- Out-positioned. Everything is accurate, but the comparison is set up so a rival looks stronger, usually because the rival has proof the engine can cite and you do not.
- The warning sign: unsourced. You are named, but from memory or other people's pages, not your own. This is where most wrong descriptions start.
Being described wrongly is often worse than being left out. A buyer who does not find you may keep looking. A buyer who finds a wrong description has no reason to, and if they do contact you, they arrive with that picture.
The lens applied: what you can do yourself
Each row is a finding from the study, read through brand and narrative, with something you can do yourself. The last column names the gate and state in our framework: Found (does AI name you?), Understood (does it get you right?) and Actionable (where does it send a buyer who is ready?). The overview explains each state.
| Finding | What it means for your story | What you can do yourself | Gate · state |
|---|---|---|---|
| Published prices get firms named | Without your price, the buyer hears the category average | Publish your price with what it includes, so it says what you want it to say | Understood · Undersold |
| Two languages, two markets | You have one reputation per language | State your positioning the same way in every language you sell in | Found · Overlooked |
| Figures move between runs | Other people's numbers become the starting point for your sales conversation | Become the source of the numbers in your field: publish your method and worked examples | Actionable · arrives misinformed |
| Some engines wait, some volunteer | On some engines you are the explainer; on others, the name | Make sure your best explainer pages carry your name and a next step | Found, and for buyers who never ask, not considered |
| Few brands appear on every engine | Your story differs by engine | Keep a one-page list of confirmed facts and check each engine against it | Across gates: how stable your position is |
| Engines handle sources differently | What an engine "remembers" about you may be years old | Keep directories, registers, review sites and your LinkedIn page in line with your website | Found · Unsourced |
| The textbook answer wins | You can be ruled out before anyone compares you | Publish an honest page on when a buyer needs a solution like yours and when they do not | Found · Overlooked, framed out |
| Specialists win specific questions | A big name does not protect you on niche questions | Publish proof on the criteria buyers compare on: cases, reviews, certifications, results | Understood · Out-positioned |
| Each engine shows different proof | Intermediaries tell your story, and some compete for the buyer | Check how you appear on the comparison and review sites in your field, correct them, and give buyers a reason to come direct | Actionable · via aggregator |
| Pages used without the firm's name | Your expertise builds someone else's answer | Attach your name to your claims, data and examples, close to where engines quote them | Found · Ghost-cited |
Try it yourself: a 20-minute brand check
- Write a one-page list of what is true about you: what you do, for whom, where, at what price, with which certifications, and why buyers choose you.
- Ask ChatGPT, Gemini and Claude: "What does [your company] do, who is it for, and what does it cost?"
- Mark each claim in the answers: matches, has drifted, is wrong, or is missing.
- For anything wrong or missing, look at the sources the engine linked. Fix the source, whether it is your own page, a directory or a profile.
- Ask a buyer's question without your name and see how you are compared with the firms that come up.
- Run the same check again in a month and see what moved.
Why one lens is not enough
Brand tells you what story is being told. It does not tell you which buyers hear it, or why the engine tells it that way. Read the other two lenses: how buyers decide inside an AI conversation and how AI engines decide which companies to name. For all three applied to one industry, see One study, three lenses.
Frequently asked questions
Why is ChatGPT giving wrong information about my company? Usually because its sources are wrong or missing: an outdated directory listing, an old profile, or a key fact that only exists in a PDF. Look at the sources it links, fix them where you can, and make sure your own pages state the facts plainly.
Can I correct what AI says about my business? Not directly in the answer. You can change what the engines read: your own pages, your listings and profiles, and the third-party pages that describe you.
Is being described wrongly worse than not being mentioned? Often, yes. A buyer who does not find you may keep looking. A buyer who finds a wrong description has no reason to, and if they do contact you, they arrive with the wrong picture.
What is a ghost citation? When an engine uses your page as a source but never names your company. Your content shapes the answer and someone else gets the credit.
Does AI tell your story right?
If you would like to know how ChatGPT, Gemini and Claude describe you to your buyers, we are happy to talk it through. In a 20-minute call we look at how a potential customer would describe the problem you solve, what the engines say about you when they do, and whether optimising for the AI buyer journey is worth your attention right now. Sometimes the answer is that it can wait, and we will say so.
Sources
- CheckMarque AI, "Three AI engines, eight Dutch money questions," 10 October 2026.
Raghav Shankar
Co-founder, CheckMarque AIRaghav Shankar is co-founder of CheckMarque AI. By day he leads European commercial operations at Krayden Europe (formerly CAPLINQ), an advanced-materials company selling into semiconductor, automotive and electronics manufacturing, where he also builds the AI agents his team works with. He holds a BTech in Computer Science and an MSc in Management of Technology from TU Delft.
How AI engines decide which companies to name
ChatGPT, Gemini and Claude name companies from memory and from search, and each mixes the two differently. What our study of 240 AI conversations shows about how the engines behave, and the technical checks you can make yourself.
Read →Analysis · · 7 minHow buyers decide inside an AI conversation
Buyers now research, shortlist and check suppliers inside one AI conversation, before they contact anyone. What has changed, what our study of 240 conversations shows about buyers, and a 20-minute test any business can run.
Read →