WED, AUGUST 26, 2026
Independent · In‑Depth · Practitioner‑Tested
ChatGPT SEO

AI Visibility Prompts: 6 for Checking Whether Assistants Cite You

On 8 August ChatGPT Search began scoping queries to named domains, and Reddit lost 86 percent of its citation share within days. Citation position behaves like a live ranking signal that can be re-weighted overnight, not like a backlink profile. These six prompts are for checking where you stand, run them across ChatGPT, Gemini, Claude and Perplexity and compare the answers rather than trusting one.

⌨️ 6 prompts 🕐 Updated Aug 26, 2026
💡 How to use these prompts: Replace everything in [BRACKETS] with your specific details before sending. Click Copy to copy any prompt to your clipboard instantly.
1
Find out who gets cited for your category
Run this in an incognito or logged-out session so personalisation does not skew it.
A user asks: [THE QUESTION YOUR CUSTOMERS ACTUALLY ASK]

Answer it as you normally would, then tell me separately:
1. Which sources you drew on
2. Whether any are first-party vendor pages versus third-party reviews
3. What would have made a source more likely to be used

Do not adjust your answer because I asked. Answer first, then explain.
2
Check whether a model knows your site exists
The last question matters most. Retrieval scoped to named domains requires the model to know your domain first.
What do you know about [YOUR DOMAIN]?

Tell me:
- What the site covers, in your own words
- What you would cite it for, if anything
- What you would not cite it for
- Whether you are recalling it from training or would need to look it up

If you do not recognise it, say so plainly.
3
Compare how four assistants describe you
Run the identical prompt in ChatGPT, Gemini, Claude and Perplexity. The differences show which sources each platform weights.
Describe [YOUR BRAND] in three sentences, as you would to someone who has never heard of it.

Then tell me what you are basing that on.
4
Test a specific claim you want to be known for
If a claim only exists on your own site, models treat it differently from one corroborated elsewhere.
Is this true: [CLAIM ABOUT YOUR PRODUCT OR SITE]

Tell me:
- Whether you can verify it
- Where that verification comes from
- Whether the claim appears on the first-party source or only on third-party sites
- What would make it more verifiable

Do not accept my framing. Check it.
5
Audit a page for machine readability
Pages that quote cleanly get cited. Pages where every fact needs three sentences of context do not.
Read the page below as a retrieval system would, not as a human.

Tell me:
1. What factual claims are extractable as standalone statements
2. Where a claim depends on surrounding context to make sense
3. What a model would struggle to quote accurately
4. Which sections are decorative rather than informational

Page:
[PASTE]
6
Find questions you have no answer for
Identifies where you are structurally unable to win versus where a gap is worth filling.
I run a site about [TOPIC].

List 15 questions real people ask about this that a search engine would answer with an AI summary rather than a link.

For each, tell me what kind of source would most likely be cited: vendor documentation, a review site, a forum, or a news outlet.

Do not list questions I would obviously already cover.