Most forecast reviews don’t produce a forecast

Last working day of the month. The forecast review is on the calendar. Everyone opens the same dashboard, and the deals get narrated one by one: this one’s looking good, that one went quiet but I’ll chase it, this should land if legal moves fast.

Forty minutes later you have a number. It’s roughly the number you walked in with, adjusted for the mood in the room. Nobody could tell you exactly how it was calculated — because it wasn’t. It was felt.

That’s not a forecast review. It’s a status meeting with a total at the bottom.

A real forecast review is a workflow. It has a trigger, a required input state, a fixed set of decisions made the same way every time, and named outputs. The meeting is just the room the workflow runs in. Get the workflow right and the meeting gets shorter — because there’s nothing left to argue about, only deals to act on.

Here’s the workflow, stage by stage.

Most teams think the trigger is the calendar: last Friday, month-end, whatever the cadence is. The calendar tells you when. It doesn’t tell you whether the review can run at all.

The real trigger is a precondition: every open deal in a forecast-eligible stage has been updated inside the cadence window. A deal nobody has touched in three weeks isn’t forecast data — it’s a memory. If it walks into the review unchanged, you’re reviewing fiction and calling it a number.

So the workflow starts before the meeting. Stale deals get routed back to their owner to update, or they drop to pipeline-only until they’re current. What reaches the review is a set of deals that are at least telling the truth about where they are.

For a deal to be reviewable, it has to carry five things: its stage, its amount, a close date, a next step with a date attached, and a forecast category the owner has set — commit, best case, or pipeline.

That last one is the field most teams skip, and it’s the one that decides whether the review is a workflow or a guessing game. If the owner hasn’t made the call on which deals they’re standing behind, the manager makes it for them in the room, by reading faces. A forecast built on a manager’s read of an owner’s optimism isn’t a forecast. It’s a rumor with a decimal point.

The category is the owner’s claim. The review is where that claim gets tested.

Once current deals arrive carrying their data, the review makes three decisions on each one. The same three, every deal, every month.

Reconcile. Is the stage honest? A deal sitting in a late stage with no scheduled next step isn’t a commit, however much anyone wants it to be. The stage has to match reality before anything downstream means anything.

Categorize. Does the forecast category hold up against an explicit test — a verbal yes, a mutual plan with dates, an order on the way — or is “commit” quietly doing the work of “I hope”? Best case is not a junk drawer for everything that isn’t dead yet. Each category has an entry test. A deal either passes it or moves down.

Act. Every deal that has slipped its close date or lost its next step leaves the review with exactly one thing: a single next action, one owner, one date. Not three ideas and a shrug. A review that surfaces problems but assigns no actions is the reason the same deals sit in the same place next month, getting narrated again.

A workflow is defined by what it hands back. This one hands back four things:

• A committed number — the sum of the commit category, defended deal by deal, not a top-line wish.
• An action list — every at-risk deal with its owner and its date.
• A slippage list — what pushed, what died, and why, so the pattern is visible instead of buried.
• The number, written down — so next month you can compare what you committed against what actually closed.

That last one is the step almost everyone skips, and it’s the one that makes the whole thing compound. A forecast you never check against reality is a number you will never get better at producing. Write it down, and the review stops being a monthly performance and starts being a system that learns.

Strip it back and the workflow is a contract running in two directions.

Every deal hands the review: stage, amount, close date, next step and date, forecast category, last activity. The review hands the business: a committed number, an action list with owners, a slippage list, and an accuracy delta against last month.

If a deal can’t meet its side of the contract, it doesn’t get to count. If the review doesn’t produce its side, it didn’t happen — a meeting took place, but the workflow didn’t run.

You have a forecast review, not a status meeting with a number, when all four of these hold:

• The review has a precondition, not just a time slot.
• Every deal carries a forecast category its owner set.
• Every at-risk deal leaves with one action and one owner.
• The committed number gets written down and checked next month.

Miss all four and the number on your slide is a story the team tells itself, retold every month with the same characters.

A forecast review isn’t where you find out the number. It’s where the system that produces the number gets tested — and the number is just the receipt.

Talk soon,

Marco

Founder / CEO, Novrith