October 17, 2026

Connect Granola, review feedback classifications, and check metric history

In this edition

Choose which Granola meetings and content to sync

Connect Granola with an API key under Settings → Connections (Feedback & Inbound) and select folders to import into Signals. Unfiled personal meetings and private scratchpad notes are excluded; syncing pauses if all selected folders are removed in Granola.
Sync defaults to the speaker-attributed transcript, using the Granola summary only when transcript turns are absent. Choose Verbatim Transcript Only or AI Summary Only instead. The Granola card shows three-phase progress while Sync Now runs.

Filter the Feedback page by Granola. If a re-synced Granola note’s raw message changes, its feedback classification is rerun to regenerate the signal title, summary, and extracted evidence chunks; confirmed problem analysis sets are preserved.
When Contact & Company linking is enabled on the connection, external attendees are matched against existing workspace Contacts and Companies before new records are created. High-confidence name and domain matches enrich missing primary emails; matched records also have company domains enriched. Both the Contact and Company are linked to the imported Feedback signal.

Initiative KPI cards retain saved metric history after display-language changes, including metrics renamed in the workspace. They omit an invented zero-to-current growth trend when tracking is missing.
Individual Metrics graph nodes omit an invented zero-to-current sparkline when tracking and its baseline are missing or invalid. Recorded zero and existing history remain visible.

In the Feedback detail Technical tab, collapsible sections separate signal-level Nexus Intelligence from per-item Evidence Classification. Signal details include source, origin, sentiment, frustration intensity, commercial opportunity, and sales and discovery facets. Evidence details include type, domain, severity, sentiment, agent blocker, and enterprise risk.
In Signals Reporting, multi-evidence signals aggregate their evidence’s Issue Types and Layers, with multi-cluster badges in the Signal Table. Group or filter by issue type, layer, sentiment, or commercial opportunity.

In the Feedback detail Brief tab, Canonical and Proposed Problem cards show severity, title, and a structured summary by default. Multi-sentence summaries are presented as a lead sentence with bullet points.
Supporting and context evidence is collapsed behind Show evidence. Expand it for an individual problem or use the Problems detail toggle to expand evidence across the section.