AI Prompts for Sales Forecasting

A sales forecast is a commitment built from deal-level judgments, and it is only as good as the rules behind those judgments. 'Commit' has to mean something specific — evidenced, dated, with the buyer's process accounted for — or the number is a mood. The forecast improves when categories have evidence criteria, when the bottom-up roll-up is cross-checked against a weighted or historical view, and when accuracy is reviewed by rep and by category so the systematic optimists and pessimists are visible.

These prompts define the categories, build and cross-check the forecast, and run the call and the accuracy review. They complement demand forecasting on the Operations hub, which uses the same accuracy-review logic on a different problem.

Before you use these

Have these ready to replace the highlighted [variables]:

• Pipeline with stage, amount, close date, category, qualification evidence, last activity
• Historical: forecast vs actual by rep and category for past periods; win rates by stage
• Target for the period
• Deals with known risks (mutual action plan status, stakeholder changes)

The prompts

• 1. Define forecast categories with evidence criteria
• 2. Build the bottom-up forecast with a weighted cross-check
• 3. Run the forecast call and the accuracy review

1. Define forecast categories with evidence criteria

Act as a revenue operations lead defining forecast categories for [team].

Process and framework: [stages, qualification fields]
Current practice: [how reps categorize; where forecasts miss — sandbagging, optimism, late slips]
Deal profile: [size, cycle, paper process length]

Define each category — commit, best case (or upside), pipeline, omitted — with:
1. Required evidence: e.g. commit requires economic buyer confirmed with a stated decision, paper process started with a date, mutual action plan with buyer-owned milestones, no open competitive evaluation, close date within the buyer's stated process. Best case: what may still be open.
2. Movement rules: what evidence moves a deal up; what event moves it down immediately (close date slip, stakeholder change, missed milestone).
3. Manager verification: the two questions per category that test the evidence in the forecast call.
4. Timing rule: how close to period end a deal can enter commit; treatment of deals with close dates in the last week.
5. Anti-patterns each criterion prevents (from our misses) and the metric that would show it is working (commit conversion rate by rep).
6. A one-page category card for reps.

Criteria must be checkable in the CRM or in a sentence from the buyer. 'Rep is confident' is not evidence.

2. Build the bottom-up forecast with a weighted cross-check

You are building the forecast for [team / period].

Pipeline: [opportunity, amount, stage, category, close date, evidence notes]
Historical: [win rate by stage; commit and best-case conversion rates by rep and overall; typical slip rate]
Target: [number]

1. Bottom-up: sum by category (commit, best case, pipeline) with the deal list per category.
2. Weighted view: amount × stage win rate, summed; note where stage win rates are unreliable.
3. Historical view: commit × historical commit conversion + best case × its conversion — by rep where history exists.
4. Reconcile: the three numbers side by side; the deals that account for the gaps (large deals in commit with thin evidence; best-case deals with strong evidence); the rep whose history suggests their commit is over- or under-stated.
5. Risk-adjusted forecast: the number to submit, the range (low/likely/high), and the top five deals whose outcome swings it.
6. Coverage to target: the gap and whether it can be closed in the period from best case, or requires pull-forward or new pipeline.
7. Assumptions stated in one list.

Show the arithmetic. Do not submit the bottom-up number if the historical view disagrees by more than [x]% without explaining why this period is different.

3. Run the forecast call and the accuracy review

Act as a sales leader running the forecast call for [period] and the accuracy review afterward.

Part 1 — Forecast call:
Forecast: [deals by category with evidence notes, close dates, rep]
Category criteria: [summary]
For each commit and best-case deal: the two verification questions from the criteria; the evidence the rep must state; the decision (confirm category, move up, move down) and who takes the follow-up action. Agenda ordered by deal value at risk. Close with the submitted number, its range, and the deals it depends on.

Part 2 — Accuracy review (after period end):
Results: [forecast by rep and category vs actual; deals that closed as forecast, slipped, lost, or closed unforecast]
1. Accuracy by rep and by category: commit conversion, best-case conversion, slip rate, unforecast wins.
2. Bias: which reps or segments systematically over- or under-forecast, with the evidence, and the likely mechanism (criteria not applied, incentives, buyer process misjudged).
3. Deal-level lessons: the three misses with the largest effect and what evidence was missing at forecast time.
4. Changes: to category criteria, to the verification questions, to coaching for specific reps.
5. The metric to watch next period to confirm the changes worked.

Keep both parts factual. Name the mechanism, not the person, when diagnosing bias.

Evidence criteria by category, illustrated

What the first prompt produces for an enterprise motion with a 90-day paper process.

Related prompts

• Pipeline Review and Analysis
• Deal Reviews
• Mutual Action Plans
• Sales Qualification
• Sales Territory Planning
• CRM Data Analysis
• Demand forecasting — Operations & Supply Chain
• Forecast accuracy review — Operations & Supply Chain

Logical next step

After this, most sales teams move on to Pipeline Review and Analysis.

All Sales prompts · Search the full library