You have the history. Ask it what happens next.
Five forecasting tools read a column of past periods and write the next ones into your file, with the error figure beside them so you know how much to trust the line.
Method guide for Google Sheets Five forecasting tools free, and no plan buys a bigger one
Forecasting in Google Sheets, five methods
Start with Autocorrelation. It asks your history whether it repeats, and tells you the length of the repeat if it does. That answer decides everything after it: a series with a season needs a method that carries one, and a series without one does not.
Moving Average is the plain one, and it is honest about itself: it smooths the noise, shows the trend and reports its own standard error. Exponential Smoothing weights recent periods more heavily, which is what you want when the level has shifted and the old months are no longer evidence.
Holt-Winters carries level, trend and season at once, which is the one to use when December is always December.
ETS & ARIMA is the self-tuning pair in one tool: ETS fits a family of state-space forecasters and keeps the one the evidence scores best, naming its choice, and ARIMA works out its own differencing and order. Both come back with exact variance bands. And every method on this page is deterministic arithmetic, not a model that guesses differently each time you ask.
The five, and when each fits
All five are in the Data Analysis panel, all five are free, and each writes a forecast tab with the chart already drawn.
Find the shape first
- Autocorrelation (ACF) does this series repeat, and every how many periods?
- Fourier Analysis the same question asked of a signal rather than a calendar: which frequencies are really in there.
Then forecast it
- Moving Average no trend, no season, and you want the noise gone.
- Exponential Smoothing the level moved recently and you want the forecast to believe it.
- Holt-Winters trend and a repeating season together.
- ETS & ARIMA you would rather the tool pick: it compares a family of models, names the winner, and bands the forecast.
When the future is a range, not a line
- Monte Carlo risk simulation a forecast with odds on it rather than a single line. That is a Pro engine, explained on the Monte Carlo page.
A forecast that reports its own error
Every forecast here comes back with the error figure attached, because a line without one is a guess wearing a suit. The report tab holds the fitted values next to the actuals, so you can see where the method missed before you build a plan on the part that comes after.
And when the honest answer is a range rather than a line, the range is a run away: put distributions on the inputs and simulate. A Holt-Winters or ETS forecast opens straight into that simulation from its own report tab; the forecast stays free, and the door says before you take it that a simulation spends a Monte Carlo run. The Monte Carlo guide is the next page along.
Free or Pro
All five forecasting tools are free. The classic four are uncapped: any history length, any number of periods ahead, no counter. ETS and ARIMA carry their own ceilings, the same on every plan: 50,000 and 20,000 readings, up to 120 periods ahead.
Pro covers the five simulation engines and the AI readings on top of every result. Pro is $199/year.
Worked examples to start from
Each loads with a real history in it, so you can see what the method does before you paste your own.
- Ask your revenue whether it repeats, and how oftenDoes Your Revenue Have a Season?
- Four quarters ahead, with the season carried throughSeasonal Demand Forecast
- How fast should a forecast believe a shift?A Forecast That Trusts Recent Weeks
The other 9 models in the library that use one of these:
- What is next month's demand, give or take?Smooth the Noise, See the Trend
- How many people will the peak month need?Staff for the Season, Not the Average
- Is intake growing, or does it just feel busy?How Many Matters Are Coming?
- Which forecast should you actually quote?What Will Next Quarter's Fees Be?
- Is the line running on a cycle nobody scheduled?Does Our Output Repeat on a Cycle?
- Is a flat daily plan good enough?The Season Repeats: Forecast It
- Are bookings growing, or is that the zigzag?Smooth the Booking Curve
- How many teachers will next summer actually need?How Many Students Next Term?
- What will a tonne of steel cost you next month?Where Is This Material Price Going?
Try it in your own sheet
- Open Sortia in Google Sheets and choose Start from a template.
- Pick one of the models above, and it loads with the inputs filled in.
- Change the assumptions to fit your situation and press Run.
Never used Google Sheets? Start here goes the whole way, in seven steps, and assumes nothing.
Other methods: Monte Carlo Decision trees Schedule risk Critical chain Optimization under uncertainty Statistics Machine learning Optimization What-if analysis