Build a sales forecast in Excel from your real numbers
You need next year's numbers for a budget, a bank or a board, and a gut-feel guess will not hold up. This prompt has your todo.is agent read your past sales or CRM export, pick a sensible method, and build a 12-month sales forecast in Excel with three scenarios and every assumption written down.
The prompt
- Build a [FORECAST PERIOD] sales forecast for [BUSINESS AND WHAT YOU SELL]. I attached [YOUR SALES DATA]. Our known changes coming up: [PLANNED CHANGES]. Look for trends and seasonality in the history, then forecast monthly revenue and units by [BREAKDOWN]. If I gave you a pipeline, use stage-weighted probabilities; otherwise use the trend plus seasonality. Make three scenarios (worst, base, best) and put every assumption (growth rate, win rate, average deal size, churn) in one Assumptions tab I can edit, with formulas linked to it. Add a chart of actuals vs forecast. Deliver an Excel file and a short summary of the base case, the biggest risks and what would move the numbers most.
What to change
- [FORECAST PERIOD]: e.g. "12-month", "next quarter by week" or "FY2027".
- [BUSINESS AND WHAT YOU SELL]: e.g. "B2B HR software, annual and monthly plans" or "a bakery with 2 shops and catering".
- [YOUR SALES DATA]: Attach a CSV/Excel of monthly sales for 1–3 years, or a CRM pipeline export. You can also type rough monthly totals.
- [PLANNED CHANGES]: e.g. "price rise of 8% in March, 2 new sales reps from May, new shop in September". Write "none" if nothing.
- [BREAKDOWN]: e.g. "product line", "region", "new vs existing customers" or "sales rep".
Example result
- Sales forecast 2027: Northwind HR (base case)
- Illustrative example built from a fictional company's data.
- Method
- • 24 months of history: revenue grew about 3% per month on average
- • Clear seasonality: December and August are about 20% below average, January and September above
- • Open pipeline of 46 deals weighted by stage: Discovery 10%, Demo 25%, Proposal 50%, Negotiation 75%
- Key assumptions (Assumptions tab)
- • New deals per month: 9 (base), 7 (worst), 12 (best)
- • Average first-year deal: 6,800
- • Win rate from Demo: 22%
- • Monthly churn of existing revenue: 1.2%
- • Price rise of 8% on renewals from March
- • Two new reps from May, fully productive after 3 months
- Forecast summary
- • Q1: 412,000 (pipeline-heavy, most confident)
- • Q2: 455,000
- • Q3: 471,000 (August dip, new reps ramping)
- • Q4: 508,000
- • Full year base case: 1.85M (worst 1.58M, best 2.12M)
- What moves the number most
- • Win rate: each point up or down changes the year by about 60,000
- • Rep ramp: if the new reps take 5 months instead of 3, the year drops about 70,000
- • Churn: 1.2% vs 2% monthly is a 110,000 difference
- Biggest risks
- • 38% of Q1 depends on 5 large deals in Negotiation
- • No history yet for the higher price, so renewal churn may rise
- Workbook tabs
- • Assumptions, History, Pipeline, Forecast by month, Scenarios, Chart
How to do it with todo.is
- Copy the prompt and fill in the [brackets] with your business, the period and the changes you already know about.
- Paste it into todo.is on the Today screen and attach your sales history or CRM export.
- Your agent analyzes the data, builds the Excel model and explains the base case and the risks.
- Change any assumption in the workbook, or ask your agent "rerun it with a 25% win rate".
- Make it monthly: "On the 2nd of each month, update the forecast with last month's actuals I email you".
Tips for a better result
- Forecast from the bottom up when you can (deals x win rate x deal size). Top-down growth percentages are easy to defend badly.
- Keep at least 12 months of history so the agent can see seasonality. With less, ask it to use industry seasonality as an assumption and label it.
- Separate new and existing customers. Renewals are far more predictable than new sales and deserve their own line.
- Compare the forecast to actuals every month. The gap tells you which assumption was wrong.
- Ask your agent to remember your assumptions so next month's update starts from the same model.
sales forecast: FAQ
- What is the best method for a sales forecast? It depends on your data. With a CRM pipeline, a stage-weighted forecast works well for the next 1–2 quarters; for longer periods, trend plus seasonality from past sales is more reliable. Many teams combine both.
- How accurate is a sales forecast? Short-term forecasts based on a real pipeline are usually much closer than 12-month ones. That is why the model shows best and worst cases instead of one number.
- Can I forecast sales for a new business with no history? Yes. The agent builds it from drivers like traffic, conversion rate, average order and capacity, and can research typical figures for your industry. Every number is marked as an assumption to check.
- Does the Excel file keep working formulas? Yes. The forecast tabs use formulas linked to the Assumptions tab, so changing one input updates the whole workbook.
JavaScript is required to use the todo.is app.