Our forecast is wrong every quarter. What is actually causing that?

No gated pricing and no call required to see a number. If you want to talk it through, the first one is free and takes thirty minutes.

Sapwood is a revenue operations and GTM engineering shop in Denver. The diagnostic is free. Retainer $2,000 a month for up to 15 hours a week, Kaypo included. included in the retainer outside it.
See every price

The forecast is rarely the problem. The causes sit underneath it: stages that mean different things to different reps, deals aging in place, close dates that slip without consequence, no historical conversion data by stage, and managers adjusting numbers without a consistent method. Diagnose it by pulling the last four to eight quarters and comparing what was committed, when it was expected, and what actually happened.

Last updated September 2026

Published prices. The diagnostic is free. Retainer $2,000 a month for up to 15 hours a week, Kaypo included.

The usual causes

Stages mean different things to different repsWithout exit criteria, a stage three deal is a different animal depending on who owns it, and no aggregate number built on that is meaningful.
Deals sit in stages long after they should have movedAging pipeline inflates coverage and hides the fact that nothing is progressing.
Close dates slip without consequenceIf pushing a date costs nothing, dates stop carrying information.
No historical conversion data by stageWithout knowing what a stage three deal has historically converted at, weighting is guesswork wearing a percentage.
Reps forecast on optimism rather than evidencePredictable, human, and correctable with qualification structure rather than with pressure.
Managers adjust numbers without a methodJudgment applied inconsistently is worse than no adjustment, because it cannot be corrected over time.
Sales cycles are longer than leadership assumesEspecially by segment. Blended cycle length hides the segments that are actually slow.
Pipeline coverage is inadequate and nobody noticedA quarter can be mathematically impossible in week two, and often nobody checks.

How to find out which one you have

Pull the last four to eight quarters

Every deal, with its stage history, close date changes, and outcome.

Compare commit against actual

What did you think would close, when did you think it would close, and what happened.

Look at where slippage clusters

If it concentrates in one stage, that stage lacks exit criteria. If it spreads evenly, the problem is qualification.

Check conversion by stage and segment

This tells you whether the issue is pipeline creation, qualification, or execution.

Then build methodology on real rates

A weighted model built on your own history rather than a benchmark borrowed from another business.

Want a defensible forecast?

Forecast methodology, stage design with exit criteria, and opportunity lifecycle are covered in Revenue Architecture, part of the retainer over six to eight weeks.

Common questions

Every price we charge is on the site.

Let's chat See pricing

Reporting and Business Intelligence

What it covers and what it costs

What we built and what it changed

The definitions everything inherits, where revenue leaks between teams, and one number with one source.

  • The first 30 days What we do first
  • What an audit covers Scope and output
  • What RevOps is A plain definition

Questions

How accurate should a sales forecast be?

Within 10 percent of commit is a common target for a mature process. More important than the number is whether variance can be explained after the quarter.

Is forecasting a data problem or a people problem?

Almost always both, and the data half is cheaper to fix. Stage definitions and historical conversion rates remove most of the room for optimism.

How long does it take to fix forecasting?

Six to eight weeks for methodology, stage design, and lifecycle. You will need a quarter or two of running it before accuracy improves measurably.

Do we need a forecasting tool?

Usually not. Most forecasting tools produce a cleaner version of the same unreliable input. Fix the stages and the history first.

What improvement is realistic?

An 18 percent improvement in forecast accuracy is achievable when the underlying stages and methodology were never properly built.

Let us look under the hood.

Read-only access to your systems, and a written report you keep either way.

Start the free diagnostic

Let's chat.

Thirty minutes. Tell us what you are working on and we will tell you on the call whether we can help.

Add your name and a work email so we know who we are talking to.

Thank you!

We will reach out to you shortly. If you have not heard from us by tomorrow, email hello@sapwood.io.

While you wait, keep reading