Product roadmap
Where Backburner is headed. These extend the owned-only recommendation core into
adjacent backlog pains. Items are listed in suggested build order — priorities and
effort are estimates, not commitments.
Already shipped
Explicit game status (backlog / playing / completed / dropped)
Semantic library search
Resume context — LLM "get back in" for in-progress games
Planned features
High-value ideas from competitive research and product analysis
04 — Achievement-aware completion coaching
Medium effort
Priority 1
Surface Steam achievements and recommend games you are close to finishing — "3 achievements from done" — as a sharper alternative to HLTB-only finish-started picks.
Why it fits: Complements finish-started and Resume with a harder completion signal than playtime alone.
05 — Persistent play queue + tiered snooze
Medium effort
Priority 3
A user-ordered shortlist of what to play next, plus snooze windows of 1 day / 1 week / 1 month instead of the current fixed skip.
Why it fits: Natural extension of pin-for-tonight and the feedback loop; high daily-use value.
06 — Community-sentiment-informed pitch
Medium effort
Priority 5
Weave trustworthy review highlights into vibe pitches and resume blurbs — "players say combat clicks after chapter 2" — instead of opaque review percentages.
Why it fits: Makes the LLM copy more credible without touching the recommendation core.
07 — "You already own this" sale guard
Low–Medium effort
Priority 2
Paste a game you are tempted to buy and use embeddings to surface similar games you already own — the ethical inverse of sale-tracking apps.
Why it fits: On-brand ("play what you own") and reuses existing embeddings for a novel positioning.
08 — Recurring time profiles
Low effort
Priority 4
Saved profiles like "Weeknight ~1h" or "Weekend ~3h" that pre-fill the Coach vibe form instead of re-entering minutes and platform each session.
Why it fits: Pure UX polish on the headline flow; no backend required for v1.
09 — Streaming AI + spotlight in library
Medium effort
Priority 7
Stream the LLM pitch token-by-token and highlight the recommended game in the library grid for a more premium Coach experience.
Why it fits: Perceived quality upgrade without changing selection logic.
10 — Cross-platform libraries (Xbox / GOG / PSN)
High effort
Priority 8
Import and unify libraries beyond Steam so recommendations can span every platform you own. The largest strategic lever — and the largest design effort.
Why it fits: Most recurring cross-competitor request, but conflicts with the current Steam-first architecture.
New ·
11 — Decky plugin integration (Steam Deck)
Medium–High effort
Priority 6
A Backburner plugin for Decky Loader so you can get your next-game recommendation straight from the Quick Access Menu in Game Mode and launch the pick without leaving the couch. Reuses our ProtonDB tier and Steam Deck compatibility data to filter the 3-game playlist to Deck-playable titles, then launches via steam://run/{appid}.
Why it fits: The Deck is the canonical backlog device, and our existing Deck/Proton enrichment makes a QAM panel a thin client over the Coach API — distribution where the decision actually happens.
Explicitly out of scope
Deliberate non-goals — they pull against "play what you own" or feel like obligation.
Social / multiplayer (friend compare, co-op discovery)
Heavy goal/streak gamification
Sale-tracking that nudges buying
Sign in with Steam