Does ChatGPT citing you mean Perplexity cites you too?

How much do the assistants' sources overlap?

Very little in Orbit Media's study.

Orbit Media collected 13,184 citations across 1,765 answers from ChatGPT, Claude, Gemini and Perplexity, using 72 prompts across three B2B brands run weekly, with data through August 23, 2026. Orbit Media found that on one brand's question set ChatGPT and Perplexity shared only 7.6% of their cited domains, and that all four models cited the same domain for the same question in only 1.7% of the question and domain pairs it compared.

Because Orbit Media worked with three B2B brands and 72 prompts, the study shows how the assistants behave in that setting and does not give a rate for every market.

Can one assistant name you while another names a competitor?

Yes, and the same company can look very different from one assistant to the next.

Agenzy audited a B2B compliance software company, name masked, with 75 buyer prompts that did not contain a brand name, across ChatGPT, Perplexity and Google AI Overviews in four markets on one day in July 2026. Agenzy found that the company was named in 23.59% of 2,153 AI answers and ranked sixth of 18 brands, but in only 11.03% of answers in ChatGPT in the US, and in 1 of 376 answers to buyer questions framed around a problem instead of a product category.

One company, one set of questions, and the result depends on the assistant, the market and the way the question is put.

Can an assistant read your page and still recommend someone else?

It happens, and it is a separate problem from being unseen.

Digital Commerce tested 25 direct-to-consumer and e-commerce brands in 14 categories on ChatGPT, Claude and Perplexity, in a non-personalized setting, across 450 answers. Digital Commerce found that 20 of the 25 brands (80%) were invisible when the buyer described a need and had not yet named a brand, and that 3 of the 25 (12%) had what it calls ghost rankings: their own content was cited as a source while a competitor was recommended.

Ghost rankings matter because a company can see its pages cited and conclude that all is well, while the recommendation goes to someone else.

Does it matter if the assistant sends no clicks?

Yes, because the answer itself can be the whole buying moment.

Researchers posting to arXiv followed one month of browsing data from a representative panel of 900 US adults. The arXiv study found that only 1% of visits to pages with an AI Overview produced a click on a cited source, and that users clicked results on 8% of pages with an AI Overview against 15% of pages without one. If few readers click through, being named in the answer is worth more than being listed as a link beneath it.

How do you check which assistants name you?

Run the same set of buyer questions on every assistant and keep a record.

1. Step 1Write ten questions a new customer would ask. Use the words a buyer uses, for example the best provider for a type of need or a company to solve a named problem. Leave your brand name out.
2. Step 2Ask each assistant in a fresh conversation. Use ChatGPT, Claude, Gemini, Grok and Perplexity, and paste the same text each time.
3. Step 3Record who is named. Note which companies each answer recommends, which it only mentions and whether yours appears.
4. Step 4Record the sources. Write down which pages and sites each assistant cites when it names a company.
5. Step 5Compare the assistants. Mark the companies and sources that appear in several answers and those that appear in only one.
6. Step 6Compare with your own pages. Check what each assistant says about you against what your site says today.
7. Step 7Fix, then ask again. Update your own pages and the public profiles the answers rely on, then repeat the same questions with the same wording and compare.

What does EntityRise do with this?

EntityRise is an AI visibility audit for agencies and companies that shows how ChatGPT, Claude, Gemini, Grok and Perplexity describe and recommend a company and what to improve first.

Each audit asks up to 40 buyer questions: you mark 15 as high priority and they go to all five assistants, and up to 25 more are each asked on one assistant, so one audit keeps up to 100 answers with their sources. The report is a PDF of around 40 pages with the analysis assistant by assistant, your main competitors compared area by area and an ordered list of what to improve. You can start the audit, and agencies can run it for clients under their own name.