"We can't roll this out to the team without single sign-on."
⚙ SAML/SSO Fernwood & Co · $5,000/mo
Linked to 1 feature · 1 account · counted in Demand below · Example data
Product · Insights & Prioritization
Feedback in. Priorities out.
Every request, review, survey answer, and support thread lands in one feed — linked to the feature it's about and the customer who said it. Then the roadmap argument ends: features rank by requests and the revenue behind them, scored in the framework your team already speaks.
Every source named on this page is live today, with real sync.
One feed
All of it. One place. No tabs.
Feature requests, app-store reviews, survey answers, research notes, design comments, community threads, error reports, support chats — unified into one feed the moment a source is connected.
-
Canny · Productboard · Featurebase
-
Typeform · Dovetail · Notion
-
App Store · Google Play
-
Figma · Discord
-
Sentry
-
Support chat · Manual capture
Capture anything by hand. Heard it in a call? Add it as an insight in seconds — same feed, same links, same scoring.
Chat becomes feedback. Threads from the support widget land on the feed as insights — not in a separate inbox silo.
Webhooks out. Every captured insight can fire an outbound webhook, so the rest of your stack reacts to new signals.
The feed fills per connector you switch on. No email-forwarding channel yet — feedback arrives via connectors, the chat widget, or manual capture.
Linked, not piled
Every insight knows its feature — and its customer.
A feed you can't act on is a junk drawer. Here, each insight links to the feature it's about and the account that said it — so feedback carries revenue, and revenue carries into every ranking on this page.
- ✓Tag any insight with a customer account — their plan and revenue come with it.
- ✓Link insights to features, and demand accumulates where decisions happen.
- ✓The same links surface on the customer's 360 page — feedback next to what they pay.
Demand
Ranked by who's asking — and what they pay.
The Demand view lists every feature with two numbers next to it: how many customers asked, and the revenue those customers represent. The loudest request and the most valuable request are rarely the same row.
-
Requests, counted
Every linked insight increments the feature it points at. No more re-reading threads to remember who wanted what.
-
Revenue, summed
Linked accounts bring their subscriptions. A feature's demand line reads in dollars, not just votes.
-
ROI, per feature
Demand over effort — a clear return figure on each row, so the trade-off is visible before the meeting starts.
Score it your way
Five frameworks. Your call, per product.
Prioritization stops being a vibe when the scoring is explicit. Pick the framework each product runs on — the demand and revenue data underneath stays the same.
- RICE
Reach × Impact × Confidence ÷ Effort. Reach defaults from revenue-weighted demand, so the math starts from real asks.
- WSJF
Weighted Shortest Job First — cost of delay over job size, for teams that ship in flow.
- Value-Effort
The classic 2×2, with scores instead of sticky notes. Quick wins surface themselves.
- MoSCoW
Must / Should / Could / Won't — when the conversation is scope, not sequence.
- Kano
Basics, performance, delighters — score how a feature lands, not just how loud the ask is.
- PER PRODUCT
Set in product settings, next to your methodology. A B2B product can run WSJF while the consumer app runs Kano — same workspace.
AI on the feed
AI that tidies the data and argues the call.
Two real capabilities, grounded in your records — no theatre. One keeps your feature taxonomy clean; the other gives every feature a recommendation you can interrogate.
Three names, one feature. Merged.
When your event stream reveals the same feature under different keys, the AI suggests a canonical name and merges them — so demand and usage count toward one row, not three.
export-csv · csv_export · exportCSV
↳ merge suggested
CSV export — one feature, one history
From your product events · Example data
An invest / fix / deprecate / watch verdict on every feature.
The AI reads each feature's usage, signals, and open work, then returns a confidence-scored recommendation — argued from the numbers, not from whoever spoke last.
SAML/SSOINVEST · 86%
Session replayWATCH · 71%
Legacy importerDEPRECATE · 74%
From usage + signals + open work · Example data
- The feed fills per connector — connect a source and its history syncs in; nothing pretends to be there before that.
- No email channel yet. Feedback arrives through connectors, the chat widget, and manual capture — forwarding an email into the feed isn't built.
- Support chat is human-answered. Threads land as insights automatically, but no bot replies to your customers.
Early access
Stop arguing in roadmap meetings.
Connect a feedback source and Stripe, and within the walkthrough you'll see your own demand ranked by revenue — scored in your framework.