The system prompt is fully editable, so you can tell the bot to avoid major plot points unless explicitly asked. Most authors instruct the bot to confirm which book a reader is on before answering arc questions, which catches accidental spoilers from earlier-book readers asking about later-book events. The default settings err on the side of avoiding spoilers entirely.

Yes, via OpenAI Files. Authors often upload the first chapter so the bot can answer 'what's the opening like' with an actual quoted passage rather than a marketing blurb. Some authors also upload character bios, world-building notes, and series glossaries so the bot can answer fan questions in depth without revealing material that has not been published yet.

If you sell signed editions through WooCommerce, the bot can read product data and stock just like for any store. A reader asking 'do you have any signed paperbacks of book two left' gets a real answer from inventory, not a generic 'check the shop'. Out-of-stock items can route to the next-best alternative or the email-when-available form.

Yes. Multibot supports a chatbot scoped to a specific series page or universe, with its own spoiler rules, character context, and tone. A romance author writing both contemporary and historical can run two distinct chatbots, one per genre, with different voices and different recommendations. Both bots can route to the same newsletter or to genre-specific lists.

If translated editions are listed in WordPress, the bot can mention publishers, formats, and release dates per language. A French reader asking when the German edition arrives gets the German release date if it is in your system. WPML and Polylang are honored, so the chatbot replies in the reader's language with citations to the language-matched book pages.

Logs are stored in your WordPress database, fully under your control. Authors who treat reader questions as editorial input often export logs monthly and use them to plan blog posts, newsletter content, and bonus material. For authors with strict reader-privacy commitments, the fact that conversations never touch a third-party SaaS is often the deciding factor.

If pre-order pages and dates are in WordPress, the bot can answer 'when can I pre-order the next book' and link to the right retailer. For multi-retailer pre-orders, the system prompt can route to a books2read-style page or directly to a reader's preferred store. The bot will not invent pre-order dates that have not been announced.

Yes, and arguably better, because nonfiction readers ask sharper, more answerable questions. A reader of a productivity book might ask 'which chapter covers weekly review systems' and the bot can answer from the table of contents and uploaded chapter excerpts. Course and workbook authors also use it to point readers at supplementary material that ships with the book.