HubSpot AI forecasting ignores recently closed deals in pipeline reports mainly because of close date mismatches, forecast snapshot timing, and deals falling outside the forecast period or pipeline filter. The AI model scores deals based on their close date and forecast category, so a deal closed after the report's date range or assigned to the wrong pipeline simply won't appear in projected numbers.
How HubSpot AI Forecasting Decides Which Deals Count
HubSpot's predictive forecasting (available in Sales Hub Professional and Enterprise) builds projections from three signals: the deal's close date, its forecast category, and the pipeline it's assigned to. The AI weights open deals by historical win patterns and rolls closed-won deals into actuals. If any of these fields drifts out of alignment, the deal drops out of the calculation.
Most teams get this wrong by assuming "recently closed" means "recently created or updated." The forecast doesn't care when you marked a deal closed — it cares about the close date you set on the record.
Close date sits outside the forecast period
This is the single most common cause. HubSpot forecasts are scoped to a time window — monthly, quarterly, or custom. A deal closed today but with a close date set to next quarter rolls into next quarter's forecast, not this one. Reps frequently leave the auto-populated close date untouched, so a deal won in March still shows a January close date and lands in the wrong bucket.
Check: Deal record > Close date field
If Close date < report start OR Close date > report end
→ Deal excluded from current forecast period

Forecast category never updated to Closed Won
HubSpot AI separates deals into categories: Pipeline, Best Case, Commit, Closed Won, and Closed Lost. When a deal stage changes to closed-won, the forecast category usually updates automatically — but custom pipelines or misconfigured stage-to-category mappings can break this. A deal stuck in the "Commit" category while technically closed won't roll into actuals correctly.
Wrong pipeline assignment
If your report filters on a single pipeline and a deal lives in a different one, it's invisible. Teams running separate pipelines for renewals, expansions, and new business hit this constantly. The deal closed — it's just closed in a pipeline the report doesn't query.
Snapshot Timing and Data Sync Lag
Forecast reports often run on scheduled snapshots rather than live data. HubSpot's AI forecast recalculates on a cadence, and pipeline snapshot reports capture deal states at a fixed point. A deal closed an hour ago may not appear until the next recalculation cycle.
- Live deal reports reflect changes near-instantly
- Forecast snapshots can lag by hours depending on portal settings
- Historical pipeline reports are frozen at their snapshot timestamp and never backfill
If you're comparing a live deal list against a snapshot-based forecast, the mismatch is expected behavior, not a bug.
Generate Proposals with AI in seconds.
Try now
Permissions and Team-Based Forecast Scoping
Forecasts can be scoped by team or owner. A recently closed deal owned by someone outside the report's selected team won't show. This trips up managers who switch between "My team's forecast" and "All deals" views without noticing the filter. The same logic that shapes whether HubSpot Sales Hub or Salesforce fits a B2B startup applies here — forecast scoping depends heavily on how you structure teams and ownership.
Step-by-Step Fix
- Open the missing deal and confirm the close date falls inside the forecast period.
- Check the forecast category — it should read Closed Won, not Commit or Best Case.
- Verify the pipeline matches the one your report filters on.
- Confirm deal ownership sits within the team or user scope of the report.
- Force a recalculation by reopening the forecast tool, or wait for the next snapshot cycle.
- Audit your stage-to-category mapping in Settings > Objects > Deals > Pipelines if categories aren't auto-updating.
HubSpot's own forecasting documentation covers category mapping and period configuration in detail, and it's worth re-reading whenever you add a new pipeline.
