Use two or more engines. Domains are compared as sets, duplicates and protocol/paths are ignored, just the host matters.
Overlap scores appear here once two engines have domains.
What Jaccard overlap tells you
Engines do not share a retrieval index. For the same prompt, their cited domain sets typically overlap only 16 to 20 percent. This tool computes the Jaccard coefficient, the size of the intersection over the size of the union, for every pair of engines you paste.
Low overlap is normal and it is the point: it means you cannot run one AI-search strategy. Each engine consults a different set of third-party domains, so your off-page targets differ by engine. Overlap above 0.35 usually means your prompt set is too narrow to generalise.
How to use it
- Paste the domains each engine cited for one prompt.
- Read the pairwise Jaccard overlap scores.
- Target the distinct domains each engine favours.
Frequently asked questions
What is a normal cross-engine source overlap?
Low. For the same prompt, engines typically share only 16 to 20 percent of cited domains, because each runs a different index and re-ranker. Overlap above 0.35 usually means your prompt set is too narrow to generalise.
What do I do with a low overlap score?
Treat each engine as a separate off-page target. The domains ChatGPT cites are not the ones Perplexity cites, so a single link-building or PR list will not cover them. The matrix shows which domains to pursue for which engine.
Is the Cross-Engine Source Overlap Calculator free to use?
Yes. The Cross-Engine Source Overlap Calculator is completely free, with no sign-up, no usage limits and no watermark on the output.
Is my data private?
Yes. The Cross-Engine Source Overlap Calculator runs entirely in your browser. Nothing you paste or enter is uploaded, stored or sent to any server.
Related tools and reading
Keep going with the playbooks and tools behind this one.