The Sales Mastery Podcast · Episode 15

The Forecasting Lie

With James Denny · Sales Geek

Why your numbers say one thing and your gut says another — and how to build a forecast you’d bet your own money on.

Here is a statistic that should bother you more than it probably does. Gartner studied forecasting accuracy across 245,000 businesses. Only 7 percent of sales teams forecast within 90 percent accuracy. Which means the other 93 percent — almost everyone — are running their businesses on a number that is wrong, often by double digits, quarter after quarter.

It is the last week of the quarter and you are in the pipeline review. The CRM says £4.2 million. The board deck on the screen says £3.8 million. The finance model says £3.1 million. Somebody in that room is wrong. And here is the uncomfortable truth we uncover almost every time: it is probably everyone.

Most companies treat the forecast like the weather. It is something that happens to them. They look through an optimistic window, peer a few days ahead and hope the sun shines. But a forecast is not a prediction. In the 93 percent, it is a wish with a date on it — and you cannot run a business on a wish.

Last time we talked about pipeline velocity: the engine. This is the dashboard you steer that engine by. And most dashboards are quietly lying to the driver.

A forecast is not a prediction. In most companies, it is a wish with a date on it.

Sit with a leadership team and ask them to present their last forecast, and the same thing happens almost every time. It is not accurate. It is a tidy visual representation of some gut feels. Because for most sales teams, the forecast was never built to be accurate. It was built to be presented.

Think about how the number actually travels. A rep looks at their deals, picks a percentage that feels about right and sends it up. Their manager decides it looks a bit high or a bit low against target, nudges it, and sends it up again. By the time it reaches a board it has been massaged by four or five people who all had a reason to move it — and not one of those reasons was accuracy.

So the single biggest source of forecast error is not the maths. It is human judgement. Reps are optimistic by nature and by training — we hire them for it. They overweight the brilliant conversation they had on Tuesday and underweight the boring structural signals: how old the deal is, how many people have to say yes, whether anyone has actually committed to anything. We do not want to kill the optimism. We want to calibrate it.

The single biggest source of forecast error is not the maths. It is human judgement.

Every dishonest forecast is telling you one of three lies, if not all of them. The skill is learning to spot which one you are hearing, because each has a different fix.

Notice what all three have in common. The number you are staring at is a measure of what someone wanted you to believe, not a measure of what will close. It is a presentation, not an observation.

You beat all three lies with one rule: a deal earns its place in the forecast by what the buyer has done, never by what the rep feels. Feelings go up and down. Evidence does not. Build the forecast on evidence using three rungs, with two exceptions that keep it pragmatic.

The buyer’s actions decide the category, not the rep’s mood and not the manager’s target. “They seemed positive” does not qualify. “They sent the contract to legal and we have a signed date on the 14th” does. If the evidence is not there, it drops a rung. No argument, no negotiation.

Two exceptions that keep it honest

Omit. A deal is genuinely live, but a specific issue must be resolved before you can trust it — a contractual question, a change on the buyer side. You don’t want it inflating the forecast, and you don’t want it dragging your velocity while you sort it. So it comes out temporarily, with a clear action and a hard cap (say 60 days). Past that, it is not a special case — it is the decision you are avoiding, and it faces the usual discipline.

Bluebird. Every so often something wonderful and completely abnormal lands — the £500k deal in a business where the average is £50k. Leave it in your numbers like any other deal and it warps everything: average deal value balloons, velocity flattens, the forecast tells a story that won’t repeat. By management exception, it stays visible — it can even sit in the forecast — but it is held out of your velocity metrics so it doesn’t distort the averages you steer by.

A small discipline with a big payoff: never forecast a close to the 31st or the last working day of the month. If anything slips — even something you couldn’t control — the deal rolls into next month and dents your accuracy. Aim closes at the first ten days of the month, and a slip still lands in-month with room to recover.

Here is a bit of fun that gets under the hood fast — making it fun is one of our values, after all. For every deal a rep calls commit, run three steps.

It works because it changes what the rep is defending. They are no longer protecting their optimism, their pride or someone’s target. They are being asked to back a judgement with something that costs them — and the moment money is on the line, even hypothetical money, people get honest fast. The deals that were 80% because they felt good quietly become best case. That is not a downgrade. That is the truth arriving early.

Hesitation is the forecast telling you the truth.

The first time you run this, your forecast will shrink. It might shrink a lot, and there is a temptation to panic. Don’t. You haven’t lost a single deal — they were never really real. You have swapped a comfortable lie for an uncomfortable truth, and you can run a business on an uncomfortable truth.

Do it every week and something quietly powerful happens. Your reps stop performing the forecast and start inspecting the deals. Your number starts landing where you said it would. And your word — to your board, your bank, your team — starts to mean something again. Because in the end, a forecast is not a number. It is a promise.

You can run a business on an uncomfortable truth. You cannot run one on a comfortable lie.

1. A forecast is a promise, not a prediction. It is your word to the board and the bank. If you wouldn’t stake your credibility on it, don’t submit it.
2. The biggest source of error is human judgement. Reps are paid to be optimistic — don’t fight it, just stop letting feeling set the number.
3. Learn which of the three lies you’re hearing. Optimism, inertia and politics each need a different fix, but all three die under evidence.
4. Categories are earned by the buyer, not the rep. A deal’s place is decided by what the buyer has done, never by how the rep feels or what the target needs.
5. A stalled deal is not a forecast deal. If it hasn’t moved, it has no recent buyer action to point to — so it drops down the ladder on its own.
6. Make it fun — ask for the bet. Would you put £100 of your own money, or your bonus, on this date? The hesitation alone tells you the truth.
7. Honest and smaller beats inflated and wrong. A forecast that shrinks when you inspect it didn’t lose deals — it lost lies. That is a win.