Forecasting is the most visible output of your revenue system. It's the number your CEO takes to the board, the metric your investors use to assess predictability, and the signal your team uses to calibrate effort. When it's wrong consistently, it doesn't just create discomfort in QBRs — it erodes the trust that a high-performing revenue organisation is built on.
In my experience, forecast failure falls into one of three distinct patterns. Most companies have a dominant one — occasionally two. Knowing which one you're dealing with is the difference between fixing the right thing and wasting three months on the wrong lever.
The Three Failure Patterns
The Optimism Problem
You always forecast higher than you close
Reps consistently submit numbers that don't close. The forecast comes in at 110% of what actually lands. Leadership has learned to apply a "haircut" — silently discounting the number before presenting it upward. Everyone knows the forecast is inflated, but no one says it out loud.
This is a pipeline inspection problem. Deals are moving through stages based on rep optimism rather than verifiable buyer signals. Commit categories mean different things to different people. There's no shared language for what "likely to close" actually requires.
The Volatility Problem
The number changes dramatically week to week
Monday's forecast is $1.2M. By Thursday, it's $800K. A deal slipped. A champion left. An integration issue surfaced. The number isn't wrong — it's just impossible to rely on. Leadership can't plan, and board conversations become exercises in explaining volatility rather than demonstrating control.
This is a pipeline coverage and deal qualification problem. When your forecast depends heavily on a small number of large deals, any movement creates outsized impact. Combined with weak deal qualification, individual deal surprises cascade into forecast swings.
The Black Box Problem
Nobody knows how the number was built
The forecast exists. It gets submitted. But ask anyone how it was derived and you get a different answer. Some managers roll up rep numbers. Others apply their own adjustments. Someone factors in historical conversion rates. Someone else is going by gut. The number might even be accurate — but it's not reproducible and not trustworthy.
This is a methodology and governance problem. Without a documented, agreed-upon forecasting process, every manager is running their own model. When the quarter misses, there's no system to interrogate — only people to blame.
Important: Many companies have elements of all three. But there is almost always a dominant failure mode. Start there. Trying to fix all three at once is how forecast improvement projects stall — too much change, too fast, not enough visible progress to sustain momentum.
How to Diagnose Which One You Have
Pull your last eight quarters of forecast vs. actuals. Plot them on a simple chart. Then ask three questions:
- Direction: Do you consistently over-forecast or under-forecast? (Optimism Problem)
- Variance: Is the gap between forecast and actual getting larger or more unpredictable over time? (Volatility Problem)
- Explanation: When you ask your managers how the number was built, do they give consistent answers? (Black Box Problem)
The answers will point you to your dominant failure mode. From there, the fix is specific, measurable, and achievable within a single quarter — if you're willing to do the structural work, not just the cosmetic work.
What Forecast Excellence Actually Looks Like
A high-performing forecast doesn't mean you always hit the number. Markets change. Deals slip. That's the nature of sales. What it means is that when you miss, you understand exactly why — and you can point to the system signal that should have caught it earlier. That's the standard worth building toward.