Key takeaways
- Being named and being cited are different outcomes. A name appears in the answer text; a citation is a page the answer shows as one of its sources. A fintech can have either without the other.
- Our study counts how often each vendor is named, written as named in N of M answers, one assistant at a time. It does not count citations, and the figures below carry their intervals.
- A fintech is cited when a page of its own states a specific claim plainly, an assistant can reach it and read it, and the claim is easy to lift. Nobody outside an assistant's provider can promise a citation.
A B2B fintech gets cited in AI-generated answers when an assistant that searches the web finds a page of the fintech's own, or an independent page about it, that states a specific claim plainly enough to use as a source for something a buyer asked. The page has to be reachable by the assistant's search, readable as text, and clear about who is making the claim. Whether an assistant then shows it as a citation is the assistant's decision, and nobody outside its provider can promise it.
This page separates two things buyers and fintech marketers often run together, being named and being cited. It then shows what our UK B2B technology study counts for the two categories that serve fintech, and sets out what a fintech can do. Nothing here judges any vendor's product.
Named, cited and recommended are different outcomes
An AI answer can treat a company in three different ways, and each answers a different question.
- Named. The company's name appears in the answer text, perhaps in a list or a passing comparison. This tells you whether the answer connects the company with the question.
- Cited. A page is shown as one of the answer's sources. This tells you whether the answer used someone's evidence, and whose.
- Recommended. The company is put forward as the fit for the buyer's stated need, ideally with the reason attached.
A fintech can be named without any of its pages being cited, because the answer may have taken its name from a review site or a comparison. It can also have a page cited for a general point without the answer naming it. Mention, citation and qualified recommendation goes through the six common patterns and the first piece of work each one points at. The difference matters for this page's question, because getting cited is about your pages, and being named is about whether the answer connects your name to the question.
What our study counts
Our UK B2B technology study counts how often each vendor is named. It does not count citations. We put the same buyer questions to three assistants through their APIs, each with its own web search: Gemini with Google Search, GPT through Azure OpenAI with Bing search, and Perplexity through its Sonar Pro model. Each category has a fixed list of vendors, chosen before any answer was collected.
A vendor counts as named when the answer names its product, one of its listed aliases, or an edition or plan of it. A company named only as an owner or parent does not count.
Each figure is written as named in N of M answers: M is the number of that assistant's answers for the category that were stored and classified, and N is how many of them named the vendor. Every figure belongs to one assistant, because the assistants search differently and answer differently. Every figure also carries an interval, a range that says how high or low the true rate could plausibly be. Reading an AI visibility figure honestly explains why.
The fintech figures
Two categories in the study serve fintech. The first is payments infrastructure: card acceptance, direct debit, open banking payments, embedded payments and business payouts for UK B2B companies. The second is finance and ERP for the mid-market: the accounting, finance and ERP systems that mid-sized UK companies run.
| Vendor | Gemini (API, Google Search) | GPT (Azure OpenAI API, Bing search) | Perplexity (API, Sonar Pro) |
|---|---|---|---|
| Stripe | named in 16 of 18 answersInterval 0.64 to 1.00 | named in 18 of 20 answersInterval 0.70 to 1.00 | named in 18 of 20 answersInterval 0.70 to 1.00 |
| Adyen | named in 13 of 18 answersInterval 0.44 to 1.00 | named in 15 of 20 answersInterval 0.50 to 1.00 | named in 14 of 20 answersInterval 0.40 to 1.00 |
| GoCardless | named in 14 of 18 answersInterval 0.50 to 1.00 | named in 16 of 20 answersInterval 0.50 to 1.00 | named in 12 of 20 answersInterval 0.30 to 0.90 |
| Worldpay | named in 5 of 18 answersInterval 0.10 to 0.50 | named in 10 of 20 answersInterval 0.20 to 0.75 | named in 8 of 20 answersInterval 0.10 to 0.70 |
| Checkout.com | named in 5 of 18 answersInterval 0.05 to 0.56 | named in 10 of 20 answersInterval 0.20 to 0.80 | named in 5 of 20 answersInterval 0.00 to 0.50 |
| Mollie | named in 5 of 18 answersInterval 0.05 to 0.56 | named in 8 of 20 answersInterval 0.15 to 0.70 | named in 4 of 20 answersInterval 0.00 to 0.50 |
| PayPal | named in 5 of 18 answersInterval 0.00 to 0.59 | named in 5 of 20 answersInterval 0.00 to 0.50 | named in 6 of 20 answersInterval 0.05 to 0.55 |
| Wise | named in 6 of 18 answersInterval 0.05 to 0.64 | named in 4 of 20 answersInterval 0.00 to 0.50 | named in 4 of 20 answersInterval 0.00 to 0.45 |
| TrueLayer | named in 3 of 18 answersInterval 0.00 to 0.45 | named in 2 of 20 answersInterval 0.00 to 0.30 | named in 2 of 20 answersInterval 0.00 to 0.30 |
| Modulr | named in 4 of 18 answersInterval 0.00 to 0.53 | named in 1 of 20 answersInterval 0.00 to 0.15 | named in 0 of 20 answersInterval 0.00 to 0.28 |
Answers behind these figures
- Gemini (API, Google Search): 18 of 20 answers stored, 0 not classified
- GPT (Azure OpenAI API, Bing search): 20 of 20 answers stored, 0 not classified
- Perplexity (API, Sonar Pro): 20 of 20 answers stored, 0 not classified
Not yet compared with the app
| Vendor | Gemini (API, Google Search) | GPT (Azure OpenAI API, Bing search) | Perplexity (API, Sonar Pro) |
|---|---|---|---|
| Sage | named in 19 of 20 answersInterval 0.85 to 1.00 | named in 20 of 20 answersInterval 0.72 to 1.00 | named in 20 of 20 answersInterval 0.72 to 1.00 |
| Microsoft Dynamics 365 (Business Central and Finance) | named in 20 of 20 answersInterval 0.72 to 1.00 | named in 20 of 20 answersInterval 0.72 to 1.00 | named in 18 of 20 answersInterval 0.70 to 1.00 |
| Oracle NetSuite | named in 20 of 20 answersInterval 0.72 to 1.00 | named in 20 of 20 answersInterval 0.72 to 1.00 | named in 18 of 20 answersInterval 0.70 to 1.00 |
| SAP | named in 10 of 20 answersInterval 0.25 to 0.75 | named in 12 of 20 answersInterval 0.30 to 0.85 | named in 9 of 20 answersInterval 0.20 to 0.70 |
| Xero | named in 14 of 20 answersInterval 0.40 to 0.90 | named in 8 of 20 answersInterval 0.15 to 0.65 | named in 6 of 20 answersInterval 0.10 to 0.60 |
| QuickBooks | named in 11 of 20 answersInterval 0.25 to 0.85 | named in 4 of 20 answersInterval 0.05 to 0.40 | named in 8 of 20 answersInterval 0.10 to 0.70 |
| Acumatica | named in 7 of 20 answersInterval 0.15 to 0.55 | named in 8 of 20 answersInterval 0.10 to 0.70 | named in 6 of 20 answersInterval 0.00 to 0.60 |
| Unit4 | named in 1 of 20 answersInterval 0.00 to 0.15 | named in 3 of 20 answersInterval 0.05 to 0.30 | named in 3 of 20 answersInterval 0.00 to 0.35 |
Answers behind these figures
- Gemini (API, Google Search): 20 of 20 answers stored, 0 not classified
- GPT (Azure OpenAI API, Bing search): 20 of 20 answers stored, 0 not classified
- Perplexity (API, Sonar Pro): 20 of 20 answers stored, 0 not classified
Not yet compared with the app
What these figures show
Read the two tables as a small sample and as a direction.
- A small group of vendors tends to be named in nearly every answer, and the rest far less often. The study reports that pattern across most categories, and it is visible in each table above.
- The assistants do not agree. A vendor can be named in most of one assistant's answers and in few of another's, so read the three columns separately.
- The intervals are wide. Each figure rests on the answers to one category's questions, and two vendors whose intervals overlap may not differ at all.
What these figures do not show
- They do not show citations. A vendor named in many answers may have had few of its own pages cited, and the reverse. The study does not say.
- They are not about recommendation. An answer can name a vendor to recommend it, to compare it or to warn against it. The study counts only that it was named.
- They are not about the consumer apps. These are API answers, and the apps can answer differently. We have not yet compared these figures with answers from the apps.
- They are a snapshot, and they cover the tracked list only. Assistants change their models and search often, and a vendor outside a category's list is not counted.
- They say nothing about a product. Being named often is not the same as being the right choice.
What a B2B fintech can do to be cited
Citations come from pages. The work is to make the pages an answer would want to use exist, be reachable and be easy to lift from.
1. Start from the questions fintech buyers ask, in their words. Buyers in payments and finance tend to name constraints: the kind of payment, the markets covered, the systems it must connect to, how it is regulated, what it costs to run. An answer can only connect you to a constraint through something public. Building a buyer-question set shows how to collect the wordings and freeze them.
2. See which sources the assistants actually cite for those questions. Citations differ between assistants. Perplexity shows numbered sources in its answers, and ChatGPT shows them when it searches. How to see which sources ChatGPT and Perplexity cite sets out what a source view needs to show to be useful: the exact page, the answer it appeared in, the run and the assistant.
3. Make your own claims easy to cite. Put each specific claim on a page of its own, stated as plain text a crawler receives, with the company's name in the title and the opening, so a good page is not used without you being named. State what is true and date the pages that change. How a B2B tech company gets recommended more often by ChatGPT and Claude covers access for each assistant's search crawler, which is separate from its training crawler, and the pages that answer buyers' questions.
4. Trace what the cited sources look like for the vendors that are named. Pick one answer, read its citations, sort each source by kind and owner, and compare what the named vendor has with what you have. What evidence AI assistants rely on when they recommend B2B vendors sets out the trace.
5. Turn the gaps into briefs, then measure again. Turning AI visibility data into a content plan shows how to rank gaps and write one brief for each. Then ask the same questions again, on each assistant separately, and count citations and names on their own bases.
Read the study, or request the report
The findings page sets out every figure in every category, the method and its limits. The full report is free with your work email.
If you would rather have the measurement run for you, our services page lays out the steps, from a first read of your own buyer questions to monthly measurement.
