Sales forecasting that survives the quarter: a method built on inspection, not optimism

Every sales leader has sat in a forecast call where a rep says "commit" and everyone in the room quietly knows the deal is not real. The rep is not lying. They are hopeful, and the process gave them no reason to be anything else.

Forecasting is not prediction. It is the disciplined inspection of what you know about each deal, applied consistently, on a schedule. This article lays out a method that works for teams from five reps to fifty, and that gets more accurate every quarter you run it.

Start with coverage, but calibrate it

The classic heuristic says you need about three times your target in open pipeline at the start of a period. It is a decent default because many B2B teams win somewhere between a quarter and a third of their qualified opportunities. It is a poor rule if you apply it blindly.

Compute your own ratio. Take the last four quarters, and for each one look at the pipeline that was open on day one and how much of it closed by the end. That gives you a realised conversion rate per quarter. The inverse is the coverage you actually needed. If it is 2.4x, planning for 3x is conservative. If it is 4.1x, planning for 3x guarantees a miss.

Then do the same by stage. Pipeline sitting in early stages converts far less than pipeline in negotiation. Coverage that is 3x but concentrated in discovery is not the same as 3x with half of it in late stages.

Inspect deals with a framework, every time

The forecast is a sum of deal-level judgments. Making those judgments consistent is the whole game. A qualification framework is the tool for that; MEDDPICC is the most widely used in enterprise software, and the questions it asks are the ones that matter regardless of the acronym you prefer.

The forecast call should not ask "is this deal going to close". It should ask "which of these elements do we have evidence for". A deal with metrics, an engaged economic buyer, a defined paper process and a champion is a commit. A deal where the rep has met one person who likes the demo is pipeline, whatever the close date says.

Define categories in writing

Most teams use some version of commit, best case and pipeline. The categories are useless unless they are defined by criteria a manager can check.

A workable set of definitions:

• Commit: economic buyer engaged, decision process and paper process mapped with dates inside the period, champion confirmed, no unresolved competitive or technical blocker. The rep would be surprised if it slipped.
• Best case: most of the above, with one or two identified gaps that have a plan and a date. Realistic upside, not a wish.
• Pipeline: qualified opportunity with a real pain and a plausible close date, but material gaps in the framework.
• Omitted: open in the CRM but not forecast for the period. Being honest here is what makes the other categories meaningful.

Publish these definitions. Hold reps to them. Hold managers to them too: a manager's roll-up that equals the sum of rep commits with no adjustment is not a forecast, it is a spreadsheet.

Run the cadence

Forecasting is a weekly habit, not a quarter-end scramble.

Weekly, per rep (30 minutes): review every commit and best-case deal against the framework. What changed since last week. What is the next step, who owns it, and when. Move deals between categories based on evidence. Record the reason.

Weekly, per manager (30 minutes): roll up the team, apply judgment, and submit a number. The manager's number may differ from the sum of the reps. That difference, and the reasoning behind it, is the most valuable information in the process.

Monthly: compare the calls made four weeks ago with what happened. Which deals slipped and why. Which category was most wrong. This is the feedback loop that improves accuracy, and most teams skip it.

The signals that a forecast is soft

Certain patterns show up before a miss, and they are visible weeks in advance if anyone is looking.

• Close dates that move by exactly one quarter, repeatedly.
• Commit deals where the rep has not spoken to anyone in the account for two weeks.
• A large share of the number concentrated in one or two deals.
• Deals with no next step scheduled in the calendar.
• A "decision process" field that says "they will get back to us".
• Paper process not started with fewer than three weeks in the period.

A forecast review that surfaces these each week turns them into actions. A review that only looks at the total turns them into surprises.

What accuracy looks like

A well-run team lands within roughly 10% of its commit number most quarters, and knows by the midpoint of the quarter whether it is on track. That is achievable for almost any team, and it has little to do with tooling. It comes from writing down what commit means, inspecting every deal against evidence, and having the discipline to move a deal out of commit when the evidence says so, even in week twelve.