Reading the Insights hub
This is the top altitude — patterns that only appear when you look across every account at once. The hub turns them into cards you can act on, refreshed automatically every night.
Individual accounts have their own timelines and health. The Insights hub sits above all of them: it clusters the evidence from every account and surfaces the themes — the risks repeating across customers, the capability everyone is asking for, the doc that's missing. Each theme becomes a card, and every card carries the evidence behind it.
Three views, one hub
The hub splits into three tabs. Alerts is what needs attention now, scoped to specific accounts. Voice of Customer is company-wide patterns from your conversations. Account Signals is the raw per-account feed when you want to drill in.
Alerts — what needs you now
Alerts are the account-level risks worth acting on this week: an account trending toward churn, a champion who's gone quiet, a support relationship slipping. They're generated by rule-based detectors that watch health, sentiment, and engagement — so they fire on real movement, not vibes.
Voice of Customer — patterns with receipts
This is where conversations become a counted pattern instead of an anecdote: how many accounts raised a theme, whether sentiment around it is sliding, and the conversations that prove it. “12 accounts asked for X this quarter” — with receipts. Product managers prioritize with it; ops and enablement use the same cards to spot doc and training needs.
Confidence & evidence
Every card shows a confidence score and opens into the exact conversations that produced it, with timestamps. Treat confidence as a sort order, not a verdict — a high-confidence churn alert is where to look first, but the evidence is what you act on.
This is the same principle that runs through the whole product: evidence, not insights. A card you can't trace is a card you shouldn't trust.
How fresh it is
The intelligence pipeline runs nightly, so the hub reflects everything through yesterday. Cards expire as the evidence behind them ages out — a churn alert that's no longer supported by recent conversations clears itself, so the hub stays a list of what's true now, not a graveyard of old warnings.