FAQ
Frequently asked questions
Clear answers about wallet credit, usage, subscriptions, and how Tycoon charges for work.
What scoring model does the AI Product Manager use?
RICE by default — Reach × Impact × Confidence ÷ Effort — because it's consistent and forces the AI to show its work on each input. If your company has a reason to prefer ICE (smaller teams, faster iteration) or a custom model (strategic bets weighted differently), you tell the AI PM once in chat and it sticks. The key property isn't the specific model; it's that the same model is applied to every backlog item every week, so priorities move based on evidence rather than mood. Paperclip and Polsia don't enforce this consistency — they let each agent score differently, which produces the roadmap drift the workflow is designed to prevent.
How is this different from Productboard or Aha!?
Productboard and Aha! are dashboards — they surface data you still have to interpret. This workflow is a ritual with built-in synthesis: the AI PM reads the raw feedback, clusters themes, produces the verbatim quotes, rescores the backlog, drafts the retrospective, and proposes next week. You read one page and make 1-2 decisions. Dashboards make you go look; this workflow brings interpretation to you every Friday. Pairing works too — you can run this workflow against your existing Productboard or Aha! instance without migrating.
What if I don't ship every week?
The workflow still runs — just with less to retrospect on. The synthesis, rescoring, and next-week planning steps remain valuable even in weeks you don't ship. For early-stage products or larger features that take multiple weeks, the retrospective reviews progress instead of shipped features: are we on track, what risks emerged, what changed in the feedback signals. The ritual matters more than the ship cadence; the consistency is what keeps priorities honest over months.
Can I customize which signals feed into the synthesis?
Yes — tell the AI PM in chat. 'Add Twitter DMs as a feedback source' or 'drop sales call notes from roadmap input — those are sales's domain, not product's' and the next review reflects it. You can also per-role weight signals: if customer interviews matter more than survey responses for your stage, the AI PM adjusts. Unlike Paperclip where every customization is a config change, this is a conversation — the AI remembers preferences across weeks.
How do I hand off from this workflow to actual engineering execution?
Approved priorities go into Linear automatically, tagged with the week's theme, owner, and success metric. The AI CTO picks them up on Monday morning in the daily standup workflow, breaks them into sub-issues, and ships through the normal engineering cadence. At Friday's next review, the AI PM pulls back the shipped items from Linear and loops them into the retrospective. The planning-to-execution loop is closed end-to-end, which is how good product teams work at scale — Tycoon just compresses the coordination layer into one chat.