Daily Prompt Pack: Revenue Forecasting – Turn Your Sales Pipeline Into a 12-Month Forecast
Most founders guess at next quarter. I get it — I’ve been in those budget meetings where the numbers are vibes and the slide deck gets built at 11pm the night before. These three prompts turn your real pipeline data into a forecast you can actually defend to investors, or at least use to decide whether to make that next hire.
One honest caveat: weighted pipeline forecasts are blunt. They assume the probability field in your CRM is accurate, and that assumption is shaky more often than not. Treat the aggressive and conservative columns as the real story, not the weighted middle.
[Claude]
You are a SaaS revenue forecaster. Given my pipeline export with columns: deal name, stage, amount, close date, probability, and win date, build a 12-month revenue forecast. Use a weighted pipeline model (amount times probability per month), then add a conservative scenario at 80% of weighted and an aggressive one at 120%. Output a markdown table with monthly rows for weighted, conservative, and aggressive totals, plus a one-line note flagging the three deals that move the forecast most.
[ChatGPT]
Act as a startup CFO. I run a subscription business with current MRR of $X, monthly churn of Y%, and new sales of $Z per month. Build a 12-month MRR projection using the formula: MRR_next = MRR_current x (1 – churn) + new_sales. Show month-by-month MRR, net new MRR, and cumulative ARR. Then tell me the month we cross $100k MRR and what new-sales rate we’d need to hit it two months sooner. Output as a table.
[Cursor]
Write a Python script that reads a CSV of my historical monthly revenue (columns: month, revenue) and produces a 12-month forecast. Use linear regression on the last 6 months, then apply a seasonality factor from the same months last year if available. Output a CSV with month, forecast, and lower/upper bounds at 90% confidence. Add a CLI flag –plot to render a matplotlib chart. Keep dependencies to pandas, numpy, scipy, matplotlib.
Full archive at aicofounderstack.com/tag/prompt-pack. Pick the prompt that matches the data you already have. If your CRM is a mess, run the Python script first — it’ll surface the data quality issues fast.
