Add a fair collaboration signal to reviews — without reading anyone's messages

PerfCopilot reads only your team's Slack activity metadata — who is active in which channels, how often, and when — never the words anyone typed. It turns that into a bias-checked collaboration and responsiveness signal that strengthens a review built on real work output from GitHub and Jira.

Connect Slack — free up to 5 seats → · read-only, metadata only, no credit card

✓ Metadata only, never message content · ✓ Bias-checked · ✓ Combines with GitHub/Jira

What PerfCopilot reads from Slack (and what it never reads)

This is the whole point. We are explicit about the boundary because trusting us with Slack means trusting that boundary.

What PerfCopilot reads (activity metadata, read-only):

• Channel activity counts — how active a person is across the team and project channels they belong to
• Participation breadth — which cross-functional channels someone shows up in, a proxy for collaboration reach
• Timestamps — when activity happens, the basis for a responsiveness pattern (not a surveillance clock)

What PerfCopilot never reads:

• Message bodies — we do not ingest, store, or summarize what anyone actually said
• Direct messages (DMs) — private conversations are never touched
• File contents, threads, or reactions as content — the substance of communication stays in Slack

Because we read metadata, a Slack line in a draft never quotes a person. It describes a pattern — "consistently active across three cross-team channels" — and is labeled as an activity signal, not a content citation. Want the full source list? See all integrations.

Why Slack is a supporting signal, not the whole review

Slack metadata tells you someone shows up and collaborates broadly. It cannot tell you whether the work was good — that requires content-bearing evidence. So PerfCopilot treats Slack as a complement, never a standalone basis for a rating:

• It complements content sources. Pair Slack participation with the GitHub PRs and Jira tickets that show what shipped. Slack adds the cross-functional, "good teammate" dimension that code and tickets miss.
• Activity counts alone are misleading. The most active person in Slack is not automatically the highest performer, and the quietest is not the lowest. Heads-down deep work often looks like Slack silence.
• The bias check guards against over-weighting raw activity. PerfCopilot's built-in bias check flags when a draft leans too hard on activity volume — the same recency-and-visibility bias that punishes focused contributors — so Slack signal informs the review without distorting it.

The honest framing: Slack answers "did they collaborate and stay reachable?" GitHub and Jira answer "what did they actually deliver?" You need both, weighted fairly.

Slack analytics tools vs. a fair, combined review draft

PerfCopilot folds a review cycle of Slack collaboration activity into a cited, bias-checked draft in under 5 minutes — as a labeled supporting signal (activity metadata), never a quote.

Why "metadata only" matters

Most Slack-analytics tools read message text to score "sentiment" or "engagement." That is a privacy line PerfCopilot will not cross, and it is the wrong input for a fair review anyway — message content is easy to misread and easy to weaponize. By reading only activity metadata and labeling it as a supporting collaboration signal, PerfCopilot gives you the cross-functional visibility you want without surveilling what your team says. Plainly: we read activity metadata, never your messages.

Run the performance review generator across Slack and every content source at once, then see pricing — free up to 5 seats.

Connect Slack — free up to 5 seats → · metadata only

FAQ

Does PerfCopilot read my Slack messages? No. PerfCopilot reads activity metadata only — channel activity counts, participation, and timestamps. It does not read message bodies, DMs, or file contents. Nothing anyone typed in Slack is ingested, stored, or summarized.

Then what does it actually read from Slack? Read-only activity metadata: how active each person is across the channels they belong to, which cross-functional channels they participate in, and the timestamps of that activity. That becomes a labeled collaboration and responsiveness signal — never a quote.

Can Slack alone justify a performance rating? No. Slack metadata is a supporting signal, not standalone evidence. Activity counts do not measure work quality. Combine Slack with GitHub and Jira so the review rests on what was actually delivered, with collaboration as the complement. PerfCopilot's bias check also flags drafts that over-weight raw activity.

Is the Slack connection private and secure? Yes. The connection is read-only and scoped to activity metadata, so message content is never exposed to PerfCopilot. This is documented on our security page and is the same boundary across every metadata-only source.

Do I have to leave Lattice or 15Five to use this? No. PerfCopilot is a writing layer. It drafts from your Slack activity (combined with other sources) and exports the finished review into the HR platform you already run. Nothing to migrate.