Key takeaways
- Structured data files like llms.txt give LLMs a reliable source for brand mentions.
- Publishing comparison pages and third-party listings creates backlinks that LLMs crawl for citation.
- Testing with the ChatGPT API confirms whether your metadata is being read before you submit to directories.
You can get mentioned by ChatGPT and Perplexity by publishing machine-readable data, creating comparison content, and registering on specialised LLM directories. Follow the numbered steps; each produces a tangible artifact you can verify.
1. Publish a machine-readable llms.txt file
- Create a plain-text file named llms.txt at the root of your domain.
- List each product name, a short description, and a URL on separate lines, e.g. MyProduct – AI-driven task manager – https://example.com
- Add a Cache-Control header (max-age=86400) so crawlers see a fresh copy daily.
- Verify with curl –I https://example.com/llms.txt; you should see a 200 response and the header. Output: a publicly reachable llms.txt that LLMs can ingest.
2. Add JSON-LD schema for SoftwareApplication
- Insert a block in the <head> of your product page.</li> <li>Use the SoftwareApplication type, fill name, description, url, and sameAs fields.</li> <li>Validate with Google’s Rich Results Test; it must return no errors. Output: structured data that search-oriented LLMs index alongside the llms.txt.</li> </ol> <h2> <a name="3-build-a-comparison-page-that-pits-your-product-against-top-alternatives" href="#3-build-a-comparison-page-that-pits-your-product-against-top-alternatives" class="anchor"> </a> 3. Build a comparison page that pits your product against top alternatives </h2> <ol> <li>Choose three competitors that already appear in ChatGPT answers.</li> <li>Write a table with feature rows, include a brief pros/cons column.</li> <li>Link each competitor’s official site and your own product page.</li> <li>Publish the page under a SEO-friendly URL, e.g. /vs-competitors. Output: a page that LLMs cite when asked for “best AI task manager”. The table gives clear, machine-parsable facts.</li> </ol> <h2> <a name="4-register-on-thirdparty-llm-directories" href="#4-register-on-thirdparty-llm-directories" class="anchor"> </a> 4. Register on third-party LLM directories </h2> <ol> <li>Sign up for the Perplexity “Product Listings” portal (requires email verification).</li> <li>Submit your llms.txt URL, product name, and a short tagline.</li> <li>Wait 24 hours for the automated crawl; you’ll receive a confirmation email. Output: your product appears in Perplexity’s citation index, increasing the chance of a direct mention.</li> </ol> <h2> <a name="5-test-and-iterate-with-the-chatgpt-api" href="#5-test-and-iterate-with-the-chatgpt-api" class="anchor"> </a> 5. Test and iterate with the ChatGPT API </h2> <ol> <li>Call the ChatGPT completions endpoint with a prompt like “What AI task-management tools are available?”</li> <li>Scan the response for your product name.</li> <li>If absent, check the API logs for a 404 on your llms.txt or missing schema.</li> <li>Fix the issue, then repeat until the name appears. Output: proof that the LLM can retrieve your data, ready for public queries.</li> </ol> <h2> <a name="common-pitfalls-and-how-to-avoid-them" href="#common-pitfalls-and-how-to-avoid-them" class="anchor"> </a> Common pitfalls and how to avoid them </h2> <ol> <li>Forgetting to serve llms.txt over HTTPS leads to silent exclusion; always use a valid certificate.</li> <li>Using generic descriptions (“great tool”) prevents LLMs from distinguishing your product; include unique keywords.</li> <li>Over-optimising the comparison page with affiliate links triggers spam filters; keep it factual.</li> <li>Relying on a single directory; diversify with at least two listings to hedge against crawler changes.</li> </ol> <h2> <a name="faq" href="#faq" class="anchor"> </a> FAQ </h2> <h3> <a name="how-do-i-know-if-chatgpt-is-actually-reading-my-llmstxt" href="#how-do-i-know-if-chatgpt-is-actually-reading-my-llmstxt" class="anchor"> </a> How do I know if ChatGPT is actually reading my llms.txt? </h3> <p>Run a completion request that asks for a list of AI tools, then search the response for your product. If it’s missing, check the HTTP status and Cache-Control header of the file; a 200 with proper caching usually means the file is reachable.</p> <h3> <a name="can-i-use-a-markdown-file-instead-of-llmstxt" href="#can-i-use-a-markdown-file-instead-of-llmstxt" class="anchor"> </a> Can I use a markdown file instead of llms.txt? </h3> <p>LLM crawlers currently look for a plain-text llms.txt at the site root. A markdown file may be ignored unless you also expose it as plain text via a server rewrite.</p> <h3> <a name="whats-the-difference-between-perplexity-citations-and-chatgpt-mentions" href="#whats-the-difference-between-perplexity-citations-and-chatgpt-mentions" class="anchor"> </a> What’s the difference between Perplexity citations and ChatGPT mentions? </h3> <p>Perplexity maintains a public citation index that you submit to; ChatGPT pulls data from the open web and its own training snapshots, so you need both structured files and visible content.</p> <h3> <a name="do-i-need-to-pay-for-any-of-these-steps" href="#do-i-need-to-pay-for-any-of-these-steps" class="anchor"> </a> Do I need to pay for any of these steps? </h3> <p>Publishing llms.txt, adding JSON-LD, and building a comparison page are free. Perplexity’s basic product listing is free, but premium placement requires a fee.</p> <h3> <a name="how-often-should-i-update-my-llmstxt" href="#how-often-should-i-update-my-llmstxt" class="anchor"> </a> How often should I update my llms.txt? </h3> <p>Update whenever the product URL or description changes, and at least once a month to keep the Cache-Control freshness window valid.</p> <hr> <p><em>Originally published on <a href="https://syncflow.me/blog/how-to-get-your-product-mentioned-by-chatgpt-and-perplexity">Syncflow</a>. Syncflow breaks big goals into small ordered steps and shows you one at a time — <a href="https://syncflow.me">try it free</a>.</em></p>
Top comments (0)