How is this page different from /features/forecasting?
Forecasting is the whole practice: the cadence, the categories, the permissions, the history. Sales Forecast is specifically about the forecast surface itself, the page a rep opens on Friday to submit, the page a manager opens Monday to roll up, the page a leader opens before the standup. The two pages are complementary. If you want the big-picture overview of how Strkr handles forecasting end-to-end, read /features/forecasting. If you want the detail on the submit lock, variance tracking, rep accuracy scoring, AI risk flags on the surface, and the Monday morning forecast email, this is that page.
What is the difference between commit, best-case, and pipe?
Commit is the number a rep bets their quarter on, the deals they believe will close this quarter with high confidence. Best-case is the stretch, the number if the deals break the rep's way. Pipe is the broader universe of deals that are still open and could still land. Strkr's forecast surface treats all three as distinct columns, each with its own roll-up, so the leader sees the confidence band on the forecast rather than a single blended point estimate. Reps categorize their deals into one of the three (or into Won / Omit for the exceptions) weekly on the forecast page.
What does Strkr AI actually look at when it flags a deal?
Strkr AI reads every open deal in a rep's commit column and scores it across six signal classes: activity recency (how recently anything happened on the deal), decision-maker engagement (whether the primary contact and other key contacts have opened or replied to recent emails), stage velocity (how long the deal has been in its current stage versus the pipeline average), close-date slippage history (how many times the close date has been pushed), email reply cadence (whether the back-and-forth is still happening at a healthy frequency), and meeting attendance (whether scheduled meetings are being kept). The score produces a Green / Yellow / Red grade, and clicking the grade shows the top-three signals that drove it. The rep can mark the flag as disagreed-with and leave a comment; the disagreement is audit-logged for the manager.
How does variance tracking work across the quarter?
Every weekly submit is persisted, the commit number, the best-case number, the pipe number, and the per-deal categorization. At the end of each week the deals that actually closed are reconciled against the submitted commit. Over the quarter the forecast surface draws a line chart of committed vs actual per rep. The rolling accuracy score is computed as the absolute percentage error between committed and actual, averaged over the last eight weeks. A score of 95 means the rep hit within 5 percent most weeks; a score of 70 means they were routinely 30 percent off. The score lives on the rep's profile and surfaces on the manager's forecast page. Slips, deals that fell out of commit week over week, are tracked separately as a pipeline health signal.
Can a rep resubmit a locked forecast?
Yes, but every resubmit is audit-logged with timestamp, user, reason, and the delta from the prior submit. The forecast call runs against the locked number regardless of whether a resubmit happened mid-week, the number the team sees in the Monday meeting is the Friday EOD number. Resubmits upward (raising the commit mid-week) are counted per rep per quarter and flagged in the "commit creep" surface so managers can coach the pattern. Resubmits downward are rarer and usually represent a late-breaking slip.
What does the Monday morning forecast email contain?
The email lands in the leader's inbox at 7am local time every Monday (timezone-aware per tenant). It contains: a row per direct report with commit, best-case, pipe, accuracy score, slip count, AI risk count, and a delta sparkline against last week; a sorted list of every Red-flag deal across the team with a one-line reason; an inline SVG trend chart of team commit over the last eight weeks against the quarter target; and links back into the deal records and the forecast surface. Managers get their report-scoped version; the VP gets the full-org version. The hierarchy walks automatically via the manager_id self-reference. Delivery cadence is per-user: weekly, biweekly, or off.
How does Strkr handle new reps who join mid-quarter?
A rep is assigned a quota start date in their profile. Variance tracking only kicks in from the start date, and the rolling accuracy score ignores weeks before it. If the team runs a ramp, 50 percent of full quota in month one, 75 percent in month two, 100 percent in month three, the ramp lives on the rep's quota schedule and the forecast surface shows the ramped target line rather than the full target. New hires are never punished for submits they never had the chance to make. The same mechanism handles PTO weeks and leaves of absence, with the surface recording the reason the week was excluded for the audit trail.
Do deals re-assigned mid-quarter break the forecast or accuracy numbers?
No. When a deal is transferred between reps, Strkr records the transfer date on the deal. On the forecast surface and in the variance history the deal appears under the new owner from the transfer date forward and under the prior owner before. Both reps' commit and accuracy numbers track against the deals they actually owned during each window. At quarter close the reconciliation attributes the win to the rep who closed it, with the full ownership history visible for comp or crediting conversations. Territory shifts, maternity-leave coverage, and reassignment after a departure all work the same way.
