Forecasting · Clari

Make the forecast evidence-based, not a gut-feel fiction

Most teams forecast on rep optimism — which is why the number misses. Wire Clari into the loop and the forecast becomes what the data, the activity, and the historical pattern actually predict. Here's the loop, the build, and the metrics to track.

01 · Signal
Clari auto-captures CRM stage, amount, close date, and next-step hygiene, plus Gong activity and engagement. Its time-series database snapshots every opp daily — so you capture what changed, not just today's state.
02 · Reasoning
Clari computes an AI-projected forecast, a score per deal, and risk/sandbagging flags. Point-in-time analytics show what moved week-over-week: pushed, pulled, added, slipped.
03 · Action
Slack digests fire: deals needing attention, commit/upside changes since last week, and the weekly forecast-call agenda auto-built and ranked by risk and dollar impact.
04 · Writeback
Forecast category, AI score, and risk flags sync back to Salesforce — the CRM stays the one system of record.
Key Clari use-cases
Auto-capture + time-series (RevDB)AI projection vs. rep commitDeal & pipeline inspectionScenario / what-if roll-upsPipeline-created vs. coveragePushed/pulled time-series
Gong Forecast and Outreach Commit overlap here — pick one authoritative forecast tool or your numbers fork.
Build it
1Connect Clari to Salesforce; map opp fields + enable bi-directional sync.
2Define forecast categories (commit / best case / pipeline) and the roll-up hierarchy.
3Set the weekly cadence — snapshot day, submission deadline, lock time.
4Wire Slack digests for at-risk deals and week-over-week commit changes.
5Run the manager forecast call off Clari's auto-built, risk-ranked agenda.
TrackForecast accuracy %Commit coverage ratioSlippage ratePipeline-created vs. target

Revenue predictability

A number you can take to the board

The deliverable here is one thing: forecast accuracy you can defend quarter after quarter. Drive committed-vs-actual error from the ~15% most teams live with toward a sustained ±5–10% — by scoring every deal on evidence, not optimism.

The forecast operating cadence

A forecast is a ritual the VP runs, not a spreadsheet filled out Friday afternoon. Set a hard weekly submission deadline. Enforce three categories everyone shares — Commit (contractually likely, the rep will defend it), Best Case (upside if things break right), Pipeline (everything earlier, unweighted). AI flags every deal where the submitted category outranks evidence-based health — the commit-vs-evidence delta — and routes it to the manager's 1:1 as a coaching prompt, not a leaderboard. The weekly CRO pipeline-review agenda auto-builds (what moved, what slipped, the deltas to inspect, the one segment where coverage is thin). Monthly, the engine assembles the CFO revenue package. The number rolls cleanly rep → manager → region → board, with renewal commit feeding the same call as new business.

Pipeline-coverage gating

Coverage is a leading input, not a quarter-end discovery. Surface coverage vs. quota by segment (target 3–5×) and weeks-of-pipeline-left, so a thinning top-of-funnel is visible six weeks before it becomes a missed quarter. When a segment drops below coverage, that is the signal to build pipeline now.

What to measure

Forecast accuracy (committed vs. actual) — target ±5–10%, sustained across consecutive quarters. Commit coverage — 3–5× committed pipeline vs. quota, by segment. Slippage rate — trending down quarter over quarter; flag any segment above ~20%. Pipeline created vs. target — ≥100% of the build needed to sustain coverage. Proof frame: teams that run weekly pipeline-velocity discipline correlate with ~87% forecast accuracy versus ~52% for irregular tracking — the cadence is the accuracy.