How does Pilea turn customer feedback into a scored backlog?

Pilea ingests feedback you already have — a CSV export, interview recordings, call notes, or messages from Slack, Intercom, and HubSpot — and returns a scored backlog with the source evidence attached. It's built for product managers and UX researchers, and per the site you can start without admin access or integrations: upload a file and get a scored backlog in minutes, then connect Slack, Intercom, and HubSpot "when you're ready."

What you can feed it

Two paths, and you don't have to pick one permanently:

• Upload path (no integrations needed): CSV files, interview recordings, or call notes. This is the fastest way to see output, since it skips connecting any tool.
• Connected path: Slack, Intercom, and HubSpot. The site frames these as pulling in "the feedback already sitting" in those tools, so the value is coverage over time rather than a one-off import.

The site also lists Surveys as a data source in the product navigation, alongside Integrations.

What comes out

The output is a backlog table plus a per-item detail view. From the product screenshots on the site, each backlog row carries:

A detail panel for one item shows the same fields plus an assignee, a "Summarise Discussion" action, and the underlying customer quotes with source, author, company, and timestamp — e.g. a Slack message from Nora K. at Meridian, and an Intercom message from Jonas F. at Fieldset. That's the "evidence attached" part: the score isn't a black box, you can read the quotes behind it.

The dashboard also surfaces aggregate metrics: Items Crunched, Time to Triage, Time to Complete, and Customer Sentiment (shown as a percentage with a trend, e.g. "72% ↑ 4 pts").

How the scoring is framed

The site positions the score as the thing that replaces debate in roadmap meetings — "Roadmap meetings, minus the arm-wrestling" — and the backlog columns suggest the inputs are volume (mentions), breadth (customers), and revenue weight (deal size), with type and status as separate classifiers. The site does not publish the scoring formula, so treat the score as a ranking aid you can audit via the attached quotes rather than a number you can reproduce by hand.

Where it fits and where it doesn't

Reasonable fit if: you're a PM or UX researcher with feedback scattered across Slack, Intercom, and HubSpot, and your bottleneck is synthesis time — the site claims "Synthesis in an afternoon, not three weeks."

Weaker fit if: you need the scoring methodology documented, or you want to skip the upload step entirely — the connected sources are Slack, Intercom, and HubSpot specifically, so feedback living elsewhere still needs the CSV/notes path.

Practical notes before you start

• No admin access required to begin, per the site — useful if you can't get IT to approve a workspace install.
• Integrations are optional and deferred, so you can validate output quality on a CSV before asking anyone for OAuth.
• EU-hosted, stated on the site.
• Free to start, no credit card — the site says "Start for free" and "Free to start · No credit card · No integrations needed." What the free tier includes beyond starting is not specified on the page, so check the pricing page before assuming a limit.
• A live walkthrough is offered ("Book a live walkthrough with the team" / "Join weekly demo") if you'd rather see it driven than read about it.

A sensible first test: export one CSV of recent support tickets or call notes, upload it, and check whether the top-scored items match what you'd have prioritized manually. If they do, the connected sources are worth wiring up; if not, the gap tells you which fields your input data is missing.