Forecast in Google Sheets with FORECAST, FORECAST.LINEAR and TREND, where they break on seasonality, and an AI approach with confidence ranges and goals.
Google Sheets ships with forecasting functions that work for simple trends:
FORECAST.LINEAR(x, known_ys, known_xs) — fits a straight line through your history and projects the next value. Good when the metric grows at a steady rate.
TREND(known_ys, known_xs, new_xs) — same linear idea, but returns a whole range of projected values at once.
FORECAST — the older alias of FORECAST.LINEAR.
For a quick "if this line continues, where do we land?" these are fine, and they are free and already in your sheet.
The native functions assume your data moves in a straight line. Real business metrics rarely do:
Seasonality — Revenue that peaks every December, or signups that dip every weekend, are invisible to a linear forecast. It will smooth right through the pattern and mislead you.
Curved growth — A metric compounding at 10% per period is a curve, not a line; a linear fit under-projects it.
No uncertainty — FORECAST returns a single number with no sense of how confident it is. "Revenue will be 142" reads as fact, when the honest answer is "somewhere between 128 and 156."
No goal context — The function tells you a projected value, not whether you are on pace to hit a target by a deadline.
You can hand-build seasonal models with more formulas, but it gets fragile fast.
Instead of wiring up formulas, you can connect the sheet to a tool that forecasts properly. VedaX Studio's forecasting engine automatically picks the right method for your series:
- Trend models (Holt) for steadily rising or falling metrics.
- Seasonal models (Holt-Winters) when it detects a repeating weekly or monthly pattern.
- Confidence ranges on every forecast, so you see the likely band, not a single false-precision number. Values that can never go negative (counts, revenue) are floored at zero.
It reads your spreadsheet's date column — even when dates are stored as text — and projects each metric forward, all without you writing FORECAST or TREND.
A forecast is most useful when it answers a decision. VedaX layers goal tracking on top of the forecast: set a target and a deadline, and it tells you the projected value at that deadline, the gap, and a status — on track, at risk, or off track — for both "higher is better" metrics (revenue, signups) and "lower is better" ones (churn, cost).
It also explains movements: when a number jumps, it can break the change down by segment so you see *which* product or region drove it. That is the difference between a projection and an answer you can act on.
Learn the full picture in Google Sheets Analytics: Get AI Insights From Your Spreadsheets, or analyze your sheet with AI first.
FORECAST.LINEAR (or TREND for a range of values) is the standard built-in choice. Both fit a straight line through your history, so they work well for steady linear trends but not for seasonal or curved data.
Not with the built-in FORECAST functions — they are linear and ignore repeating patterns. You can build a seasonal model manually with extra formulas, or use an AI tool that applies a seasonal method (like Holt-Winters) automatically.
For a quick projection, FORECAST.LINEAR on your monthly totals works. For a forecast with confidence ranges, seasonality, and goal tracking, connect the sheet to a forecasting tool that reads your date and value columns and projects each metric forward.
Usually because the data is seasonal or curved and a linear function is forcing a straight line through it — or because there is not enough history. A method that detects trend and seasonality, and shows a confidence range instead of one number, gives a more honest forecast.
See it in VedaX: AI dashboards from Google Sheets