Sales Forecasting Made Simple

Quick Answer: sales forecasting: A reliable forecast comes from structure, not judgement: a small number of pipeline stages with written exit criteria, a clear rule for what “commit” means, and a weekly review where every deal is judged against evidence rather than optimism. The method is to separate what is committed from what is possible, weight opportunities by stage and history, and never let a forecast be built by anyone who owns the revenue target.

What this guide covers

• Why most forecasts are wrong and what causes the error
• The four components of a working forecast
• Writing a commit definition your team will actually use
• Weighting pipeline by stage and history
• Running the weekly forecast call
• Turning misses into better numbers

Why the forecast is usually wrong

Forecasts fail for a predictable reason: they are opinions about the future dressed as arithmetic. A rep says “60 percent”, the manager says “70 percent”, the owner says “let’s call it 65”, and the resulting number is nobody’s actual judgement — it is the average of three different people’s guesses.

Then two things happen. The number misses, and the miss is blamed on the market. Nobody checks what the forecast assumed, because the forecast was never a set of assumptions — it was a number. And the process repeats, with everyone slightly more careful about appearing pessimistic and slightly less willing to raise problems early.

The alternative is not more sophistication. It is structure: a small number of stages, a written rule for commitment, historical conversion rates, and a weekly review that records what was said so it can be compared against what happened.

The four components

1. Stages with exit criteria. Four to six stages, each with a written, testable exit condition. The condition matters more than the name. “Proposal sent” is an activity; “scope agreed, price accepted in principle, and a decision date booked” is evidence. When a deal sits in a stage, its probability is a known quantity rather than a feeling.

2. A commit definition. The single most valuable line in any sales organisation. A deal is committed only when all three are true: the person with authority to approve has said yes, a date exists, and the paperwork is genuinely moving. Everything else is best case. This definition should be written down, agreed by the team, and applied without exception — the value comes from it being strict.

3. Historical conversion by stage. The ratio that turns stage into probability. Multiply the value of every opportunity in a stage by the rate at which deals in that stage have historically converted, and you have a bottom-up forecast that reflects your business rather than an industry benchmark. Recalculate quarterly; the rates shift as your mix changes.

4. A weekly review with a recorded output. Every deal is reviewed against the same questions, and the forecast as stated is written down with a date. The recording is what allows learning — without it, every forecast is a fresh guess and no pattern ever emerges.

Writing a commit definition your team will use

The test of a commit definition is that it survives the moment when a rep wants a number to land. Three conditions, all required:

Authority. The person who can approve has agreed — not the champion, not the evaluator, not the person who said “we’re really keen”. This is the condition most often skipped, and skipping it is what turns a forecast into a hope.

A date. Not a target date (“end of quarter”) but an actual date on which a signature happens. A specific date is a commitment; a window is an intention.

Movement. The paperwork is in progress — the contract is with legal, the PO has been raised, the proposal is with the procurement team. If nothing has moved, nothing is committed.

Apply it strictly at first, even if the forecast looks low. The first honest forecast is lower than the team expects, and that gap is the point: it exposes either a broken process or a broken pipeline, both of which are cheaper to fix now than in the final week of the quarter.

One guardrail matters: the person who owns the revenue target should not be the person who writes the forecast. Forecasts produced under target pressure drift upward every week, and the drift is invisible unless someone independent of the target is asking the questions. A founder, an operations lead or an external advisor asking “what has the approver actually said” is enough.

Weighting the pipeline

Once stages have conversion rates, the forecast is arithmetic:

Weighted forecast = sum over opportunities of (value × stage conversion rate)
Committed revenue = sum of deals meeting the commit definition
Best case        = weighted forecast + unweighted upside from early-stage deals

Three practical notes on the arithmetic. First, use value weighted by probability rather than multiplying at the end — the difference matters when deal sizes vary widely, because a single large early-stage deal should not dominate the month. Second, discount the forecast slightly for historical slippage: if deals in your business routinely close two weeks later than the date given, adjust the weighting rather than pretending the dates are accurate. Third, keep early-stage upside out of the main number. The purpose of the forecast is to plan cash and capacity, and including speculative pipeline makes both plans wrong.

Running the weekly forecast call

The meeting should take under thirty minutes and follow the same format every week. For each deal, four questions, in order:

1. What stage is this in, and what is the specific evidence for it?
2. What has changed since last week?
3. What is the next action, and what is its date?
4. What is the probability, and why that number?

Deal review should be in stage order, not in value order, so the discussion naturally moves from early-stage risk to late-stage certainty. Reps present their own deals — a forecast meeting where the manager narrates the pipeline is an inspection, and it produces exactly the defensive behaviour that makes forecasts unreliable.

The output is three numbers: committed, weighted, and best case. And a written note of which deals are driving the change, because the interesting information is rarely in the total.

Turning misses into better numbers

The mechanism that makes forecasting improve is boring and essential: record what was forecast, then compare it to what happened.

Keep two columns per deal: forecast at week one, and actual outcome. After a quarter, the pattern shows up clearly. Deals that were 80 percent and closed late. Deals marked 30 percent that closed. Stage rates that are wrong in a specific direction — perhaps your “proposal” stage converts at 60 percent, not the 40 percent you assumed. It also exposes the most common and most costly pattern: deals that were committed and then slipped, which usually means the commit definition was applied loosely rather than that the buyer changed their mind.

Three habits make this work: never change a forecast after the fact without noting the change and why; keep the missed forecasts visible to the person who produced them, without blame; and recalculate stage rates quarterly from the accumulated history rather than from opinion.

What a good forecast is actually for

A forecast is not a prediction exercise. It is a planning instrument, and its value is that it tells you three things in advance: whether you have enough revenue to cover the plan, whether you have enough pipeline to cover a bad week, and where the gaps are.

That last use is the one teams undervalue. A forecast that says “we will land at 70 percent of target” is not useful on its own — but a forecast that says “we will land at 70 percent, and the 30 percent gap is entirely in deals that have not reached proposal” tells you exactly which part of the process needs attention. That is the same diagnostic value the stage design provides, applied to time.

If the forecast is being used to shame people, it will be wrong in a way nobody can fix, because the people producing it will manage the number rather than the pipeline. The fix is structural: strict commit definitions, independent review, recorded history.

Related reading

• The 5-Step Sales Process Every Business Should Use
• Why CRM Implementations Fail
• What Every CEO Should Measure Weekly

Frequently asked questions

How do I forecast sales accurately in a small business?

Four things, in order. First, define pipeline stages with written exit criteria so that stage membership reflects evidence. Second, write a strict commit definition — approver has agreed, a date exists, paperwork is moving — and apply it without exception. Third, weight each stage by its own historical conversion rate rather than by anyone’s confidence. Fourth, review weekly in the same format and record what was forecast so it can be compared with outcomes later. Accuracy comes from that feedback loop, not from a better guess.

What is a commit in sales forecasting?

A committed deal is one where all three are true: the person with authority to approve has explicitly agreed, a specific signature date exists, and the paperwork is genuinely in motion. Anything less — a keen champion, a verbal “we’re close”, a target quarter rather than a date — is best case, not commit. The definition should be written down and applied strictly, because the entire usefulness of the forecast depends on the commit line meaning the same thing to everyone including the person who owns the target.

Why do salespeople under-forecast?

Mostly because they are judged on it and punished for missing, which teaches them to set the number low enough to be safe. Sometimes it is avoidance: a low forecast lets a missed target be blamed on the market. Both are rational responses to how forecast accuracy is used. The structural fix is to separate the two jobs — the forecast drives planning and coaching, while the deal review looks at behaviour — so that honesty costs the rep nothing.

Should I use weighted pipeline or a bottom-up forecast?

Bottom-up, which usually means weighted pipeline, calculated from stage conversion rates rather than from a fixed probability per stage. Use the team’s own historical rates, recalculated quarterly, because industry defaults will not match your cycle length or your average deal size. Keep early-stage deals out of the headline number and show them separately as upside, since including speculative pipeline makes cash and capacity plans wrong.

How often should we review the sales forecast?

Weekly, in a fixed slot, with the same format and the same questions each time. Weekly is frequent enough to catch a slipping deal while something can still be done about it, and regular enough that the forecast becomes a routine rather than an event. Monthly reviews compound errors because each month starts from the last month’s assumptions. The other rule is that the forecast as stated gets written down — without a record, there is no way to learn which stage rates are wrong.

What should I do when our forecast is chronically wrong?

Check the direction first, because it determines the cause. If forecasts are consistently too high, the commit definition is being applied loosely or stages have no real exit criteria. If forecasts are consistently too low, reps are protecting themselves from judgement. If they are just noisy, the issue is usually that deals sit in stages for months without movement, which means the pipeline needs a review rhythm and a rule for closing dead opportunities. The fix depends on the pattern, which is why recording forecast-versus-actual is the first step rather than a guessing game.

Written by SkilBrill Training Institute

November 5, 2026 · 11 min read