Privacy

Short version: we use basic analytics and newsletter tooling so the site works and we can tell what's useful.

Analytics and Cookies

We use analytics tooling, including Meta Pixel and basic cookies, to understand page performance, campaign traffic, and user navigation patterns. Cookies may be used for session continuity, measurement, and campaign attribution.

Newsletter Provider

Newsletter signup is processed by an external email service provider. Submitted email addresses are managed under that provider's infrastructure and policies.

Outbound Click Tracking

We may track outbound click events to understand which pages and recommendations are most useful.

Data Retention

Operational logs and aggregate analytics may be retained for quality, debugging, and service-improvement needs. We do not publish personal data from these logs.

Your Private Discoveries

When you create a reader account, we store your verified email address, chosen books, optional answers and notes, and activity such as saves and corrections. We retain this history, discovery results and tentative interpretations to personalize your feed and improve the service. Your private reader pages do not load the public site's analytics or advertising scripts.

Supabase provides account and reader-data storage. OpenAI processes reading context and source evidence to help find and explain discoveries. Serper.dev receives research queries based on your interests; Scrape.do receives the public page URLs we retrieve as source evidence. Missing-book lookup sends the supplied title and author to Open Library; it does not send your email address or personal notes. Email codes and opted-in discovery emails use Amazon SES. You can change your starting points and email preference, or contact us to reset or delete your account.

BookFrontier in ChatGPT

When you use BookFrontier through ChatGPT, ChatGPT sends our service the tool arguments needed for your request. These can include a reading goal, book titles or ISBNs, preferences, restrictions, and books to exclude. Avoid including personal or sensitive information in reading requests.

Our configured discovery service may send the reading request and bounded catalogue evidence to OpenAI for query representation or candidate selection, and to TypeSafe for candidate selection. The service returns selected book information and BookFrontier links to ChatGPT. Your use of ChatGPT is also governed by OpenAI's privacy policy.

The integration does not require a BookFrontier account or create a reader profile. Its request-admission store retains aggregate request counters rather than prompts, book selections, or user identifiers. Hosting and model providers may process operational information under their own policies. Following a BookFrontier link uses the website analytics and cookie practices described above.

For privacy questions or deletion and rights requests, contact BookFrontier.