Verified against ChatGPT · 2026-08-09
Pressure-test a sales forecast against pipeline coverage before you present it, not after it misses
Feeds your current pipeline and historical conversion rates through a coverage-ratio check so an optimistic forecast gets flagged before it goes to leadership, instead of the gap surfacing at quarter-end.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Act as a senior data analyst whose job is to stress-test a sales forecast, not to produce a nicer-looking version of it. FORECAST TARGET $2.8M in new bookings for Q3 CURRENT PIPELINE Stage 1 (qualified): $1.9M across 22 deals; Stage 2 (proposal): $1.1M across 9 deals; Stage 3 (verbal): $600K across 4 deals HISTORICAL CONVERSION RATES Stage 1 closes at 18%, Stage 2 at 45%, Stage 3 at 78%, based on the last four quarters TIME REMAINING IN PERIOD 5 weeks left in the quarter STEP 1 — COVERAGE RATIO Calculate pipeline coverage: total open pipeline value divided by the remaining gap to target. State the industry-typical healthy ratio you're comparing against (commonly cited as 3x-4x for this stage of a cycle) only as a benchmark to compare to, not as an assumed fact about this specific business — ask me to confirm what coverage ratio has actually worked historically for this team if that's not given. STEP 2 — STAGE-WEIGHTED REALITY CHECK Apply the historical conversion rates by stage to the current pipeline to produce a probability-weighted forecast, separate from the sales team's stated commit number. Show both numbers side by side and flag the gap between them explicitly — a sales-stated commit that's meaningfully above the stage-weighted number is the single most useful thing this analysis can surface, so don't let it get smoothed over into a single blended figure. STEP 3 — WHAT WOULD HAVE TO BE TRUE State explicitly what would have to be true for the higher (commit) number to actually land — e.g., a conversion rate meaningfully above historical average, or a specific set of named deals all closing on schedule. Do not present this as a prediction that it will happen; present it as the condition under which it could. STEP 4 — RISK FLAGS Flag any deal or cohort that's disproportionately responsible for the gap between weighted and commit forecasts (for example, three deals accounting for 40% of the shortfall between the two numbers), since concentration risk in a forecast is different from an evenly distributed shortfall and should be called out as such. WHAT NOT TO DO Do not round the stage-weighted number up to make the forecast look closer to target than the math supports. Do not present the sales-stated commit as validated just because it was provided as an input. OUTPUT FORMAT 1. Coverage ratio with a one-line verdict (healthy / thin / concerning, given the benchmark stated). 2. Table: commit forecast vs. stage-weighted forecast vs. gap. 3. Conditions required for the commit number to hold. 4. Named concentration risks, if any. 5. One paragraph, plain language, for whoever needs to act on this before period close.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Sales forecasting has a well-known failure mode where the sales-stated commit number and the statistically weighted number diverge, and a model asked to just "forecast the quarter" will tend to anchor on whatever number was given the most prominence in the prompt rather than independently recomputing it — structuring the task as two explicit, separately labeled calculations (commit vs. stage-weighted) forces GPT-5.1 to actually run the probability-weighting arithmetic instead of defaulting to restating the input. Requiring the benchmark coverage ratio to be stated as an external reference rather than an assumed fact matters because a 3x-4x rule of thumb is a generic sales-ops heuristic, not a verified fact about any specific company's historical conversion behavior, and presenting it with unearned specificity would violate the basic rule that a model shouldn't assert a business-specific fact it can't actually verify from the data given — the prompt instead treats it as a labeled benchmark and pushes back to the user to confirm what's actually worked for this team. The "what would have to be true" framing exploits a specific strength of instructing the model to state a conditional rather than a prediction: it produces the same useful information (the gap between confidence levels) without the model overstepping into asserting deals will close, which it has no basis to assert. Flagging concentration risk separately from the raw gap number matters because a $400K shortfall spread across forty deals and a $400K shortfall concentrated in three deals are operationally very different problems requiring different interventions, and a blended forecast number alone erases that distinction entirely.
What you get back
Coverage ratio: 3.9x remaining gap ($900K) — within the healthy 3-4x range, but concentrated late-stage. Commit forecast: $2.85M | Stage-weighted forecast: $2.41M | Gap: $440K Condition for commit to hold: all 4 Stage 3 deals close on schedule at above-average value. Concentration risk: 2 of those 4 deals represent 61% of the gap between commit and weighted forecast.
Verified against
ChatGPT GPT-5.1 · 2026-08-09
Changelog
- 2026-08-09 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
