How do I get ChatGPT and Perplexity to cite my business?
Give ChatGPT and Perplexity accessible pages that directly support a useful answer, with clear business identity and verifiable evidence. Check each provider’s crawler guidance and test citations in its actual search experience; a brand mention without a source link is not a citation.
What makes a useful citation target?
Choose a stable URL that contains the fact or explanation the answer needs. A precise service page, documented method, or sample can be more useful than a homepage full of broad promises. Include enough context for the claim to stand alone: who or what it concerns, the conditions, and any limitation. Do not hide the supporting evidence behind an email form while expecting the gated content to function as a public source.
Which access checks differ by provider?
OpenAI publishes guidance for OAI-SearchBot, while Perplexity publishes guidance for PerplexityBot. Check both the robots file and security layer. A successful request from your own browser is not proof that a provider’s crawler receives the same response. When investigating a bot request, verify the provider’s documented identity information rather than trusting a user-agent string alone.
How should you run a fair citation test?
Use the same buyer question in each product with search enabled. Save the answer and open the source list. Record the cited URL and the claim it supports. Test branded and unbranded questions separately. If your company is mentioned but another site supplies the citation, record an external-source mention rather than a direct citation to your website. Keep the unsuccessful tests in the denominator.
What should you change when you are absent?
Inspect the cited sources for missing information rather than copying their wording. In an illustrative consultancy example, competitors may document a discovery deliverable while your page never explains what the customer receives. Publish an accurate sample and its limits, then repeat the tests later. Source selection remains outside your control, so pair citation tracking with observed referral visits and confirmed subscriptions.
What should you put on your working checklist?
| Recorded result | Classification |
|---|---|
| Your page is linked as a source | Direct website citation |
| Your business is named without a link | Unlinked mention |
| Another site describes your business | Third-party mention or citation |
| Your link receives a real visit | Referral visit, measured separately |
Which related questions should you explore?
Who is behind this guidance?
Which sources can you check?
What else should you know?
Is a mention enough to count as a citation?
No. Record the source link separately.
Will submitting a sitemap force a citation?
No. Discovery is only one step before source selection.
Can I test using an API instead?
You can record API observations, but label them separately from the consumer product experience.
Exploring AI too?
Have an interesting experience or a question? Feel free to get in touch. I am always happy to exchange perspectives on practical uses of AI.
hello@mikloskovacs.io
