It can hand off to any ESP that has a hosted form or API endpoint. The system prompt links the right URL and the bot does not store the email itself; it routes the subscriber to your ESP's hosted form, where the actual subscription happens. Tags can be preselected via URL parameters so a topic-specific interest carries through to your welcome sequence and segmentation.

Only if you choose to expose them. Most writers index the paid issues' titles and excerpts and gate the full text behind the subscription, which lets the chatbot work as a tasting menu without giving away the paid product. A potential paid subscriber asking 'do you cover pricing experiments' sees the relevant paid issue titles and excerpts, then the subscribe link.

Issue date and number are part of the post meta, so answers stay chronologically accurate. A reader asking 'what was your last issue about' gets the actual most recent issue, not a frozen snapshot. For newsletters with seasonal content, the bot can also explain when an issue was published and whether it has been updated since.

Yes. Each answer can include a 'you might also like' set drawn from semantically similar posts. This works the same way as the main archive search, just scoped to issues semantically near the question rather than directly answering it. For long-running newsletters, this is often where the magic shows up, because readers find issues they would never have surfaced through chronological browsing.

Importing your Substack into WordPress as a custom post type is enough for SleekAI to index it. Tools like the Substack-to-WP importer handle the conversion in a single pass. Once the issues live as WordPress posts, the chatbot reads them just like native content. You can keep publishing on Substack and mirror to WP, or migrate fully and use WP for both publishing and the chatbot.

All logs and indexes are stored on your site. SleekAI does not relay data to third parties beyond the OpenAI calls your key authorizes. For writers with strict subscriber-privacy commitments, this matters more than the chatbot itself, because the alternative is a hosted SaaS that holds reader conversations on someone else's servers.

If your ESP exposes subscriber tier in user meta or via a logged-in cookie, the chatbot can scope answers accordingly. A logged-in paid subscriber asking about a paid-only issue gets the full answer; a free reader gets the title and a paywall prompt. The same prompt can also route inquiries differently, for example sending paid subscribers to a community link and free subscribers to an upgrade page.

If your newsletter has distinct sections, like a tech essay, a links roundup, and a personal note, you can index them as separate post types or categories so the bot can answer 'show me your links roundups about pricing'. Multibot also lets you run a section-specific chatbot, scoped to one part of the newsletter, which works well for paid editions with a Q&A column or office-hours format.