Ask a sales leader for a forecast and you get a number that feels precise and turns out to be fiction. It is not because the leader is careless. It is because a forecast is asked to do something it cannot do, predict the future, when its real job is something else entirely. Change what you ask of it and the number starts to mean something.
Most forecasts miss for a reason that has nothing to do with the market. They miss because of what a forecast is asked to be. A rep is asked to predict the future, so the rep guesses. Under pressure to look busy, the guess drifts up. Under pressure to sandbag, it drifts down. Either way the number is a mood, not a measurement.
There is a better frame, and it changes everything downstream. A forecast is not a prediction about what will happen. It is a set of commitments about what people are willing to stand behind. Predictions invite padding and daydreaming. Commitments invite evidence. The moment you stop asking a team what it thinks will close and start asking what it is willing to commit to, the number gets honest, and it gets smaller.
When you ask someone to predict a deal, you are asking a question with no cost attached. Being wrong about a prediction feels like being wrong about the weather. Nobody owns it. So the number fills up with deals that might happen, deals the rep hopes will happen, and deals that are really just conversations wearing a suit.
A commitment carries weight. To commit a deal is to say, out loud, that you will stake your name on it closing this period. That single change in wording filters the pipeline hard. The deal with no next step falls out. The deal where the buyer has gone quiet falls out. The deal you called a layup because the champion loves you, but the person who signs has never been in the room, falls out. What remains is smaller, and you can actually plan around it.
So what separates a committed deal from a hopeful one? Evidence. Not feeling, not rapport, not how good the last call felt. Evidence is a record of things that have already happened, not things you expect to happen.
Ask of every deal in the forecast one question: what has been proven true? A confirmed budget is evidence. A signed order form is evidence. A written agreement on timeline, a named signer who has said yes on a call, a security review that has cleared, these are evidence. A friendly champion is not evidence. A great demo is not evidence. A verbal maybe from someone who cannot sign is not evidence.
The test is simple and unforgiving. For each deal, name the specific proof it will close on the date you claim. If the only proof is that the rep feels good about it, it does not belong in the forecast. It belongs in the pipeline, which is a different thing.
This is where the number shrinks, and the shrinking is the point.
The pipeline is everything you are working. The forecast is the subset of the pipeline that has earned the right to be counted. Most of the pipeline has not earned it. That is not a failure. It is what a pipeline is for. A pipeline is a wide funnel of maybes. A forecast is the narrow set of deals with proof behind them.
When a forecast comes in close to the size of the pipeline, that is not confidence. It is a warning. It means the team is counting hope as though it were evidence. A trustworthy forecast is almost always a fraction of the pipeline, because most deals in any pipeline are early, unproven, or quietly dead. A forecast that shrinks the number is not a pessimistic forecast. It is an honest one, and honesty is the only quality a forecast is actually for.
A forecast is only as honest as the meetings behind it. We book qualified meetings with buyers who have a real reason to move, so the deals you count have proof, not hope. One call is enough to see if it fits.
It is tempting to blame the rep who inflates. That misreads the problem. Padding is rarely dishonesty. It is a rational answer to how the questions are asked.
Ask a rep what will close and reward optimism, and you will get optimism. Ask what might close with no penalty for being wrong, and every warm conversation goes on the board. The rep is not gaming you. The rep is answering the question you actually asked, which was a question about hope.
Change the question and you change the answer. Ask what the rep will commit to, put that commitment on the record, and review it against what actually happened. Now overcommitting has a cost, because next period the rep has to explain the gap. The number tightens on its own, without a lecture about sandbagging, because the system finally rewards being right instead of being upbeat.
Every sales team has one. Call it the deal that lived in the forecast for six months. Quarter after quarter it was going to close, and quarter after quarter it slipped. The rep believed it. The manager wanted to believe it. The champion kept saying good things.
Look at that deal through the evidence test and it was never real. There was warmth, and there were meetings, and there was a champion who genuinely liked the product. There was never a confirmed budget, and the person who controlled the budget had never once been on a call. The proof that mattered was missing the entire time. The forecast counted the warmth because warmth feels like progress.
A commitment forecast would have caught it in week one. The question, what has been proven true, has no good answer for that deal, so it never gets committed. It sits in the pipeline, worked but not counted, until the proof finally arrives or until everyone admits it was never going to.
Reframing the forecast is not a speech. It is a weekly habit that shows up in how you run the review.
Go deal by deal and ask one question of each. What has been proven true here, and what has to be proven true next for this to close on the date claimed? A deal with a clear answer stays. A deal with a vague answer moves out of the forecast and back into the pipeline, where it can be worked without polluting the number. You are not killing the deal. You are refusing to count it until it earns the count.
Do this every week and two things happen. The forecast stops swinging, because it is built on evidence that does not evaporate overnight. And the team starts pursuing evidence on purpose, because they know each deal will be asked for its proof. The review stops being a status update and starts being the thing that manufactures a reliable number.
Once the number is honest, it stops being a report card and becomes a tool. You can hire against it, because it will not collapse the month you sign the offer. You can promise a delivery date, because the deals behind it have proof, not hope. You can tell your own board a number and hold it, which is the entire point of having one.
There is a further payoff most teams never reach. A forecast built on evidence tells you exactly what to fix. If deals keep failing at the same missing proof, budget never confirmed, or signer never in the room, that gap is your whole problem, in writing. The forecast stops being a guess about the future and becomes a map of what your pipeline is missing.
Which raises the next question. If the forecast is only as good as the evidence behind it, what should you track to know a deal is real long before it is time to commit it? That is a question about the pipeline itself, and it is where this gets useful.
How to Run a Sales Pipeline Review
The weekly habit that turns a forecast into a reliable number.
What Is a Sales Pipeline?
The wide funnel of maybes your forecast is drawn from.
Sales KPIs to Track
The evidence that tells you a deal is real before you commit it.
B2B Sales Funnel Stages
Where in the funnel a deal earns the proof a forecast needs.
The pipeline is every deal you are working, warm and cold, early and late. The forecast is the smaller set of those deals that have proof behind them and that someone is willing to commit to closing this period. A pipeline holds maybes. A forecast should hold only deals with evidence. If the two numbers look alike, the forecast is counting hope.
Usually because you are asking the team to predict rather than commit. A prediction has no cost when it is wrong, so optimism drifts into it and every friendly conversation gets counted. Ask instead what each rep will stake a name on, and require evidence for it, and the number falls to something honest. High forecasts are a sign of how the question is framed, not of a dishonest team.
Evidence is a record of something that has already happened. A confirmed budget, a signed order form, a written agreement on timeline, a named signer who has said yes on a call, a cleared security review. A friendly champion, a strong demo, and a verbal maybe from someone who cannot sign are not evidence. If the only reason to count a deal is that it feels good, it is not evidence.
Weekly, deal by deal. Ask of each deal what has been proven true and what must be proven true next. Deals with clear answers stay in the forecast. Deals with vague answers move back to the pipeline until they earn their place. A weekly rhythm keeps the number from swinging and trains the team to chase evidence on purpose.
The rep owns the commitment, the manager owns the standard. A rep decides which deals to commit and stakes a name on them. The manager holds the line on what counts as evidence, so a warm feeling never gets promoted to a committed deal. Split that way, the forecast stays honest without turning every review into an argument.
A fortnightly note on one shift in global business and what it means for your revenue. One idea, one move, no filler. Leave any time.
One call is enough to work out whether outsourced sales is the right fit for where your business is right now.