Why your pipeline forecast is always wrong

Published . Updated .

Stage definitions, deal hygiene, and a forecast you can plan around.

Your forecast is wrong because the pipeline underneath it is. Stages describe what your team did, not what the buyer committed to. Close dates are hopes that slide a month at a time. Deals that died weeks ago are still open. Multiply that by a percentage per stage and you get a precise-looking number built on guesses. To fix it, define each stage by something the buyer has done, set rules for close dates and stale deals, and call the forecast deal by deal every week.

If you plan hiring and spending around a forecast that keeps missing, start here. The pipeline report says one thing, the bank says another, and you've started quietly discounting every number your team gives you.

The forecast is only as good as the pipeline

Most attempts to fix a forecast start with the math: better weightings, a new report, a different formula. None of it helps if the deals going in are wrong. A forecast is the sum of many small judgments about individual deals. When those judgments are loose, the total is loose, however it's calculated.

So start with the pipeline itself. The data problem is common: in the same Gartner survey, only 47% of respondents said their organization had high-quality data, and 13% said it was poor. The usual problems come in four kinds.

Why forecasts miss

Stages describe your activity, not the buyer's

"Demo done" and "proposal sent" are things your team did. They say nothing about whether the buyer is any closer to buying. A deal can sit at "proposal sent" for months, because sending the proposal was the last thing anyone did. As Gartner analyst Craig Riley put it, "customers simply don't follow a linear progression of steps to arrive at a buying decision" (Gartner, 2020).

Close dates are set by hope

The salesperson picks a date that sounds reasonable, usually the end of this month or next. When it passes, the date moves forward a month. Nobody asks what would actually have to happen for the deal to close by then.

Dead deals stay open

Nobody likes marking a deal lost, so deals that went quiet stay in the pipeline and inflate every total. The longer they sit there, the less anyone trusts the report, and the more the real forecast moves into the founder's head. Many of those deals were never lost to a competitor. In a study of more than 2.5 million recorded sales conversations, Matthew Dixon and Ted McKenna found that 40% to 60% of deals end with a buyer who intended to buy but never acted (Harvard Business Review, 2022).

Percentages create false precision

Weighting each stage by a fixed percentage turns a list of guesses into a confident-looking number. If the stages are loose, the percentages are fiction, and a total down to the last dollar hides how uncertain it really is.

Fix 1: define stages by what the buyer has done

Rewrite each stage as an exit test: something the buyer did that anyone can check. For example:

Your stages will be different. What matters is the test: could someone else look at the record and agree the deal belongs in that stage? If not, the stage is a feeling.

Two columns of four pipeline stages. Left, seller activity: first call held, demo given, proposal sent, negotiating. Right, the buyer exit test that replaces each: confirmed the problem, decider and reason to act; decider agreed the approach solves the problem; walked through price and terms; decider said yes with dated steps to signature.Two columns of four pipeline stages. Left, seller activity: first call held, demo given, proposal sent, negotiating. Right, the buyer exit test that replaces each: confirmed the problem, decider and reason to act; decider agreed the approach solves the problem; walked through price and terms; decider said yes with dated steps to signature.
Each stage becomes something the buyer did, which anyone can check on the record. Example stages; yours will differ. Diagram: Taut Tech.

Fix 2: set rules for close dates

Want to know how much of your pipeline is real? The free audit takes four questions.

Start Audit

Fix 3: keep the pipeline clean

Deal hygiene sounds dull. It is the forecast.

Fix 4: call the forecast deal by deal

Instead of trusting a weighted total, sort the open deals expected this period into three groups:

Each week, every deal owner calls their commit and best-case deals out loud, and someone asks the same three questions every time. What did the buyer do this week? What has to happen next, and when? What would make it slip? Then write the call down.

Three stacked bands: commit, deals the owner would bet on with a verbal yes; best case, deals that could close if one or two named things go right; pipeline, everything else. Below, the three weekly questions: what did the buyer do this week, what happens next and when, and what would make it slip.Three stacked bands: commit, deals the owner would bet on with a verbal yes; best case, deals that could close if one or two named things go right; pipeline, everything else. Below, the three weekly questions: what did the buyer do this week, what happens next and when, and what would make it slip.
Sort deals into three groups instead of trusting a weighted total, then ask the same three questions every week. Diagram: Taut Tech.

Then check the forecast against what happened

At the end of each month or quarter, compare what was called as commit with what actually closed. Note which deals slipped and why. Over a few periods you'll see whether your team calls high or low, which stages leak, and which kinds of deal always take longer. That's how forecasts improve: not with a new formula, but by checking last period's call against what really happened.

A quick self-check

When the forecast isn't the real problem

If the pipeline is clean and full and still little closes, the cause is likely elsewhere in how the business runs. Pipeline looks full and nothing closes walks through that. To watch new pipeline and closed deals every week, read The 5 numbers every founder should see every Monday.

Want an outside read on how much of your pipeline is real? Start with the free audit. If you then want to build the fix yourself alongside other owners, there's the Cohort, starting mid-January 2027. If you'd rather have it built and run with you, there's the Executive engagement.

Works Cited

Find out how much of your pipeline is real. The free audit takes four questions and sends back a written read within a business day.

Start AuditBook a call

FAQ

Why is my sales forecast always wrong?

Usually because the pipeline underneath it is unreliable: stages describe your team's activity instead of the buyer's commitments, close dates are guesses, and dead deals stay open. Fix the pipeline first, then the forecast.

How should I define sales pipeline stages?

Define each stage by something the buyer has done that anyone can check, such as confirming the problem, agreeing on the solution or reviewing the proposal. Write the tests down so everyone uses the same ones.

What is deal hygiene?

The habits that keep the pipeline honest: every open deal has an owner, a next step and a date, close dates are tied to real events, stale deals are reviewed on a set schedule, and lost deals are closed with a reason.

Should I use a weighted pipeline to forecast?

It can be a useful cross-check once the stages are well defined. On its own, with loose stages, it gives a precise-looking number built on guesses. A deal-by-deal call in commit, best case and pipeline groups is more honest.

How often should I review the forecast?

Every week, deal by deal, for anything expected to close this period. Then compare each period's call with what actually closed.

How do I get a free diagnosis?

Take the free audit. Four questions, about 90 seconds, and within a business day you get a score, your top three to five findings, and one of them written out as a complete fix.