Built for operators

Know what next month
actually looks like.

Paste your sales, bookings or order history and get a forecast with a realistic low and high case — so you can decide how much to order and how many people to schedule.

Paste a column · Get a range · Plan the low case

app.forecastion.com / forecast-lab
End of forecast
611
± 115 (90% CI)
Cumulative
35.7K
h = 60 periods
Forecast avg
596
per period
Backtest MAPE
2.2%
R² 0.97 · 12/13 in range
Holt-Winters forecast
HistForecast
The planning problem

You're making real decisions on a gut-feel number

Inventory buys and staffing schedules get locked in weeks ahead. Most of the time the input is last month plus a bit — and then you eat the overstock or the missed demand.

Ordering blind

Buy too much and cash is sitting in a warehouse. Buy too little and you're out of stock in your best week.

Seasonality gets guessed

You know December is different. Turning "different" into a number is the part that never happens.

Nobody has time to be the analyst

The forecast that would help requires skills and hours you don't have spare during a planning cycle.

◷ Forecast Lab

Paste a column. Get a forecast that fits your history.

Copy the sales or bookings column straight out of your spreadsheet or platform export. Forecastion tests the methods for you and tells you which one fits your data best — you don't need to know why.

  • Paste or upload, nothing to set upa column of numbers is enough to get started.
  • The tool picks the methodevery method is scored on accuracy against your recent history, with the best fit flagged.
  • Seasonality built inbusy months and slow months are learned from your history instead of guessed at.
  • Change the horizon instantlyforecast the next 4 weeks or the next 12 months with a slider.
See the Forecast Lab in action
ForecastionForecast Lab · Air Passengers
End of Forecast
897
± 118 (90%)
Cumulative
34.5K
± 680 (90%)
h = 56 periods
Forecast Avg
616
per period
Methods
Linear TrendLogarithmicHolt-WintersGrowth Rate
Forecast Lab chart showing a Holt-Winters forecast of the Air Passengers series with historical data, fitted values, and forecast confidence bands.
⌁ Monte Carlo Lab

Plan with the range, not in spite of it

A single number tells you nothing about what to do when demand comes in soft. Simulate thousands of possible months and see how bad the bad case really is before you commit the order.

  • A realistic low casethe P10 outcome — what you should be able to survive.
  • Up to 10,000 simulationsper run, so the range reflects real variation, not one scenario tab.
  • Model the levers you controlprice, conversion, capacity, lead time — each with a range instead of a fixed guess.
  • P10 / P50 / P90 outcomesorder for the middle, staff for the low, prepare for the high.
Explore the Monte Carlo Lab
ForecastionMonte Carlo Lab · 5,000 sims · h=36
Mean
2,014
P10 → P90
1,970–2,059
CV
1.7%
Distribution · Carrier Passengers
10,000
simulations per run
8
forecasting methods
<2 min
CSV to first forecast
90%
confidence intervals, always
⊞ Workspaces

Keep every location and product line in one place

Save a forecast per SKU, per store, per service line — grouped into workspaces so next month's planning starts from what you already built, and the rest of the team can see it.

⊟ Carrier Forecast 2 tabs
Air Passengers
Carrier Demand Model
⊟ SpaceX Valuation 1 tab
xAI Revenue Forecast
⊞ GDP Forecasts 2 tabs
How it works

From last year's numbers to next month's plan

Three steps. No setup, no scripting. Just forecasts.

1

Paste your history

Sales, orders, bookings, appointments — a column of numbers from any export.

2

Get the forecast and the range

See the expected path plus the low and high cases you should plan around.

3

Order and schedule with it

Export the numbers or share the link with whoever signs the PO.

Common questions

What operators ask before they start

“I'm not an analyst — can I use this?”

Yes. Paste a column of numbers. The tool reports which method fits your history best; you don't need to know how it works.

“Does it handle my busy season?”

Seasonality is learned from your own history and built into the forecast, so peak months aren't averaged away.

“How much should I order for a bad month?”

Run a simulation and plan against the P10 low case instead of a single point estimate.

Stop operating
on a gut-feel number.

Paste your history, get a forecast with a realistic range, and make the call with something better than last month plus a bit.

Paste a column · Get a range · Plan the low case