Understand not only what the forecast says, but what could change it.
Understand not only what the forecast says, but what could change it.
Why this is hard.
A forecast is a set of assumptions about the future, and almost none of those assumptions live in the finance system. They live in the contracts sales signed, the pipeline that has not closed, the hiring that has not happened, the capacity the plant does not have, and the customer whose payment behavior changed last month.
When those assumptions are held in a spreadsheet, two things follow. Nobody can say which assumption a number depends on, and nobody finds out when one of them stops being true. The forecast is not wrong because the maths is wrong. It is wrong because the world moved and the model did not.
Improving forecast confidence is therefore not a modeling exercise. It is a matter of connecting the forecast to the places its assumptions actually come from, and then watching those places.
The signals this solution acts on.
Committed versus expected revenue
What is contractually owed, what is contracted but at risk, and what is still pipeline — separated rather than blended into one line.
Driver breakage
When an upstream driver moves — volume, headcount, price — every line that depends on it is recomputed and the change is attributed.
Assumption drift
Conditions the plan assumed are monitored continuously, and a material change is raised against the forecast that depends on it.
Backtested error
Each model is scored against a holdout of your own history, so the forecast carries a measured error rate rather than a claim of accuracy.
Variance with a cause
Actuals are compared to the published plan and the responsible driver is named, not just the account.
How the pieces combine.
Connect the sources
ERP, CRM, HRIS and the contract repository, so the forecast is built from records rather than re-keyed.
Model the real drivers
Declare the relationships that exist in the business — customers to headcount, headcount to payroll, volume to materials — and let them resolve.
Separate commitment from expectation
Contracted revenue and pipeline revenue are forecast differently because they behave differently.
Monitor the assumptions
Every material assumption becomes a watched condition with an owner.
Publish and measure
Publish the plan, then measure actuals against it continuously and report what moved.
Products involved.
You do not need all of them. Each is useful on its own; this is the combination that addresses the problem completely.
Systems it reads from
Where it fits
What this looks like when it works.
Services that help you get there
Frequently asked questions.
Why can’t our current forecast say what changed?
Because the assumptions live in a spreadsheet, apart from the records they depend on. Built from the ledger, contracts, pipeline and headcount, the forecast can attribute every movement to the driver and record that caused it.
How is contracted revenue treated differently from pipeline?
They are forecast separately because they behave differently: what is contractually owed, what is contracted but at risk, and what is still pipeline.
Do we need every product listed on this page?
No. Each product is useful on its own. A solution describes how several of them work together on one outcome; you can start with one and add the others as they create value.
Does Farsight DS replace our existing systems?
No. It works on top of the systems you already run. Nothing is replaced.
Work through improve forecast confidence against your own systems.
Thirty minutes. We map what you already have and where this would start paying.