Connect your data

Import from your accountant or sync a bank account so every forecast builds on real, booked numbers.

2 min read

A forecast is a projection of your actuals, so the actuals decide how good it is. Connecting a source is the only setup step that genuinely matters.

The two sources

DATEV import — booked entries from your accountant, mapped to canonical categories via SKR03/SKR04 presets. Complete and categorised, because someone has already done that work.
Bank sync — live transactions and a real balance. Immediate and exact about cash, with no view of what is owed.

Which to start with

They answer different questions, and the right first choice depends on which one you have.

Accounting data gives you a full picture of what the business earns and spends, categorised, going back as far as your books do — which is what a 12-month forecast needs. Bank data gives you today's cash position to the cent, which is what a 13-week view needs, but a bank feed alone cannot tell revenue from a loan or a supplier payment from a tax bill without categorisation.

Most founders start with accounting and add the bank feed once the forecast is running.

One source is enough

You do not need both to begin. Either produces a full forecast; the second sharpens it. Driver sources like Shopify plug in on top as you need them.

How much history

Foreqast uses up to 24 months and projects up to 24 months forward. Twelve months of history is comfortable — enough for the trailing averages to be stable and for a seasonal pattern to be visible. With less, the forecast still works, but treat the far end of it as a sketch until more actuals arrive.