Roadmap

What’s coming next

Honest statuses for every committed and considered item. We do not pre-announce features we are not actively scoping, and we move items off this page only when they ship to thechangelog.

1. In reviewnext release

   Real LLM-backed retention analysis

   Replace the deterministic heuristics behind Sigmora Score and the teardown breakdown with native multimodal analysis (audio + visual) via Gemini 3 Flash, wired through the Vercel AI SDK.

   • Single API call per video — no separate transcription, no two-model handoff.
   • YouTube and Shorts URLs are passed directly; Files-API upload only for non-URL content.
   • Output validated against the existing AnalysisResult Zod schema, so the UI does not change.
   • Charge-after-success credit accounting — failed runs cost users nothing.
   • 90-day cache with a "Re-analyze with latest model" button.

1. Plannedsoon

   Hard cap on daily watching spend

   Today the billing page warns when projected month-spend exceeds 50% of allocation. The cap will be enforceable, not just visible.

   • Per-workspace daily and monthly ceilings configured from the billing page.
   • Watcher refreshes that would breach the cap are skipped, not partially-charged.
   • Logged so you can see exactly which refreshes were paused and why.
2. Planned

   Automated test suite

   Add Vitest + React Testing Library + Playwright coverage so we catch runtime regressions, not just type and build errors.

   • Unit tests for the analysis runner, credits, and watcher pipelines.
   • Component tests for the Studio chat surface and the Library tabs.
   • End-to-end tests for the import → score → remix golden path.
3. Plannedongoing

   More vertical packs

   Five packs ship today (SaaS, DTC beauty, fitness, fintech, B2B services). Real estate, gaming, education, restaurants, food creators, music are the next batch — pure curation, no schema changes.

   • Each pack adds pre-tuned hook patterns and CTA templates for one niche.
   • Existing pack architecture supports any vertical — adding more is content work.
   • Suggest a pack via Cmd+J inside the app and we will prioritize the most-requested verticals.

1. Considering

   Server-side TikTok, Instagram, and X watchers

   The Chrome extension already covers these surfaces from the user side via on-demand probes. Server-side scheduled watchers would let your team monitor without a browser open.

   • Requires a paid third-party scraper (Apify or similar) — adds a fixed monthly cost.
   • Watcher pipeline today supports YouTube, Reddit, and Mastodon RSS natively.
   • On the fence because the Chrome extension already handles 80% of the workflow.
2. Considering

   Affiliate / referral program

   A real distribution lever for creator-tools — TubeMagic runs a 50% recurring affiliate program. We have the partner / signup foundations; the missing piece is tracking links and payouts.

   • Sign-up captures attribution; commissions accrue against the referrer.
   • Monthly payout flow with tax-form handling.
   • Will only ship if we see organic demand from existing users.
3. Considering

   Mobile-friendly workspace

   Marketing pages are responsive. The signed-in workspace assumes desktop. Making the core flows (Library, Studio, score review) usable on mobile is in scope.

   • Tablet support first — phone-first is a deeper redesign.
   • Studio chat and asset detail are the most-used surfaces and the priority.
4. Considering

   Scheduled posting to TikTok / Instagram / YouTube

   Publishing queue exists today as a UI; it does not actually post on a schedule. Adding the worker plus per-platform OAuth would close the loop from idea to scheduled drop.

   • OAuth integrations per platform — fiddly and account-bound.
   • Each platform changes its publishing API roughly every two quarters; ongoing maintenance is real.