Privacy Policy — Citation Gap

What Citation Gap collects, why, how long it is kept, and how to have it deleted.

Last updated: 11 August 2026

What Citation Gap does. Citation Gap is a Model Context Protocol connector — used in Claude, ChatGPT and other MCP-compatible assistants — that analyses the on-page Generative Engine Optimization (GEO) signals of a web page you specify, optionally compares it against competitor pages you supply, and returns a report.

What we collect.

• From your sign-in. Citation Gap uses Google Sign-In. Google sends us your email address, display name and account identifier so we can authenticate you. We do not store your email or name in our records. Instead we keep a pseudonymous, one-way identifier derived from your Google account id — it cannot be reversed to your name or email — together with when you first signed in, when you last signed in, and how many times you have signed in. Your email and name are used only in the moment (to verify your account and label the connection) and are held solely as part of your active connection by the sign-in layer; they are removed when you disconnect.
• From your use of the tool. The page URL and query you submit, and the competitor URLs you supply.
• From the pages analysed. We fetch the public content of the page you submit and of any competitor pages you supply (and their robots.txt, llms.txt and entitymap.json, if present) in order to analyse them.

We request only the basic Google sign-in permissions — your name, email address and account identifier. We do not request access to Gmail, Google Drive, your contacts or any other Google service, and we could not read them if we tried.

What we do not collect. We do not ask for or store your conversation with the assistant, your documents, your contacts, or any personal data beyond what is described above. We do not store your email address or your name in our records. We do not track you across other websites, build behavioural or advertising profiles, or use your data for advertising.

Why we keep a sign-in identifier. To recognise your account across sessions, to power your report history, and — when paid plans arrive — to know which plan you are on. Because we do not store an email address, we cannot send you email; if paid plans later introduce email contact, that will be separate and explicitly opt-in, and any such email will carry an unsubscribe link.

How it is used and stored.

• Your sign-in record — the pseudonymous identifier, sign-in dates and count — is stored in a database hosted by Cloudflare. It contains no name or email.
• Generated reports. When you produce a report, the rendered report is stored so its link keeps working and so you can find it again in your report history. Reports are kept until you delete them (see "Deleting your data"). Each link is an unguessable random address, and anyone who has the link can view that report.
• Report inputs, briefly, for quality. To verify that reports are accurate, we keep the analysed content of the pages involved — the extracted data and the page content as fetched, for your page and any competitor pages — for about 7 days, after which it is deleted automatically. It is used only to check report quality and is never shown to other users.
• Aggregate audit records. We keep the audited URL and its scores, with no link to who ran the audit, to show a page's score over time and to build anonymous, aggregate benchmarks.
• We use the URL, query and competitor list you provide only to run the analysis.

Link verification. When Citation Gap generates an EntityMap, it checks the Wikidata and Wikipedia links by requesting those public pages to confirm they resolve. No personal data is sent.

Third parties. The service runs on Cloudflare (hosting, database, key-value storage and headless page rendering) and uses Google for sign-in (OAuth 2.0). We do not sell your data, and we do not share it with any other third party.

Data retention. Your reports and your pseudonymous sign-in record are kept until you delete them. The short-lived quality copies of analysed page content are kept for about 7 days, then deleted automatically. Anonymous aggregate audit records (URL and scores, not linked to you) are retained to measure trends.

Deleting your data. You can delete everything we hold for you at any time: while signed in, ask Citation Gap to "delete my data." This immediately and permanently removes your pseudonymous record and every report you have generated (including the stored report pages and the short-lived quality copies). Anonymous aggregate records that are not linked to you are unaffected, because they cannot be traced back to you. Disconnecting the connector (in your assistant's settings — e.g. Claude or ChatGPT → Settings → Connectors) revokes its access at any time.

Your rights. You can delete your data yourself as described above, or email the address below with any question, correction or request. Because we do not store your email or name, acting on a request sent by email will require you to be signed in so your (pseudonymous) records can be identified.

Security. Access is authenticated with OAuth 2.0 via Google. All traffic is served over HTTPS. The administrative view shows only aggregate, pseudonymous usage counts (no names or emails), is protected by a separate secret, and is never publicly reachable.

Contact. Questions, data requests or security reports: [email protected].