The Human Tax: three ways we sabotage our own forecasts and how to stop

Most executive teams in B2B businesses have had the same experience. The pipeline looks healthy at the beginning of the quarter, the forecast has been reviewed, and the CRM is up to date. The dashboard is telling a reasonably coherent story. And yet, when a real decision needs to be made, confidence is still thinner than it should be.

The difference between success and failure is often a handful of opportunities that slip to the next quarter. That is why pipeline and forecasting remain such a point of frustration. They are among the most analyzed views in the business, but often among the least trusted. 

This is the first in a short series on forecasting. I want to start with the part that is easiest to understand and, in many cases, easiest to fix: the ways businesses unintentionally sabotage their own pipeline and forecast.

Pipelines and forecasts are about humans

Forecasting is often presented as a data problem—and it partly is. Better data matters and a messy CRM doesn't help. But anyone who has worked around sales forecasting for long knows it is not just math. It is a management process. It is a judgment process. And, more than anything, it is a human process. The pipeline you see in the CRM is only the visible version of what is happening. Underneath the headline number sit judgment, optimism, pressure, incentives, assumptions, politics, deal knowledge, buyer behavior, and sometimes hope—and it changes constantly.

That is why forecasting is difficult. The business is not just trying to calculate a number. It is trying to understand what people believe is likely to happen, based on incomplete information, changing buyer behavior, and the pressure of a target.

And this is where businesses can create their own problem.

1. Metrics replace purpose

When confidence in the forecast is low, businesses usually respond by adding more control. That is understandable. They introduce pipeline coverage targets to make sure there is enough future opportunity. They may introduce targets for forecast accuracy to improve planning discipline. They may introduce KPIs for CRM hygiene because poor data makes everything harder. They introduce stale-deal reports to clean up opportunities that should no longer be there.

None of this is irrational. In fact, most of it is sensible.

The problem starts when the metric becomes the thing people are managed to, rather than the purpose the metric was meant to serve. This is Goodhart’s Law in practical form: when a measure becomes a target, it stops being a good measure.

Take a 3x pipeline coverage target as a simple example. The original purpose is to give the business confidence that there is enough opportunity to support future revenue. But once 3x coverage becomes the target, rational people respond to the target.

• Weak opportunities stay open.  

• Early-stage deals enter the pipeline before they are properly qualified.  

• Stale deals (zombie deals) survive because removing them makes the coverage number look worse.

The business gets more pipeline, but less confidence. A few iterations later, when the conversion rate drops, leadership moves the KPI to 4x pipeline with no actual improvement in performance.

The issue is not that metrics are bad. The issue is that metrics can become detached from the decision they were meant to support. Before adding another pipeline or forecast metric, leadership should ask a simple question: What behavior will this create?

If the behavior damages the truthfulness of the pipeline, the business may improve the report while making the forecast worse.

2. When pressure makes truth difficult

Forecasting depends on truth arriving early enough to be useful. That sounds obvious, but many organizations make early truth difficult.

The business asks for realism and encourages sales teams to call out risk. But the way the business reacts can send a completely different message. A deal slips and the seller gets interrogated. A manager reduces the forecast and is asked why the quarter is going backward. A sales leader calls out risk early and the conversation becomes about whether the team is losing control.

And people learn from that.

Bad news starts to arrive later. Managers soften the message before it reaches executives. Salespeople keep upside alive because removing it creates scrutiny. Sales leaders protect the quarter narrative for as long as they can.

That is not always dishonesty. Often, it is simply risk management. People behave rationally inside the environment the organization creates. If telling the truth early creates pain, the truth will inevitably arrive late.

The fastest way to damage a forecast is to make the truth feel unsafe until it is too late to act on it. A better forecast requires an environment where uncomfortable information can surface early enough for the business to do something with it.

3. When judgment becomes invisible

Even in businesses that get the previous two things right, the forecast can still arrive broken. Not because anyone lied, but because everyone adjusted.

A rep enters a number they believe in, shaded slightly upward, because optimism is part of the job. Their manager reviews it and applies a discount based on knowing that this rep’s optimism tends to run hot. The regional lead rolls it up and applies a further haircut, because experience says the region always comes in soft against what gets submitted. By the time it reaches the executive team, the number has passed through multiple sets of hands, each applying a private adjustment based on a private model of who tends to be right.

None of these adjustments are dishonest; each is a reasonable response to experience. The problem is that nobody can see the other adjustments.

Every layer is just adding noise on top of someone else's noise. The business ends up with a number that nobody believes, arrived at by a process nobody designed.

This is what happens to a forecast in the absence of a single source of truth. Not a single system, most businesses have that, but a single agreed view of what the raw data means before anyone is allowed to adjust it. Without that, every layer of management quietly builds its own shadow model, and the official forecast becomes a kind of folklore: a number with no single author, shaped by everyone, owned by no one, and trusted least by the people who built it.

The fix is not to ban adjustment; judgment is part of the job. The fix is to make every adjustment visible. A manager who discounts a rep's number should be able to say by how much and why in the open, not as a private mental tax applied on the way up. The moment adjustment becomes invisible; it stops being judgment and starts being a layer of fiction indistinguishable from the truth it was supposed to refine.

Other Ways Businesses Weaken the Forecast

These are not the only ways businesses damage forecast confidence, though some are more structural:

• Process-Driven vs. Buyer-Driven: CRM stages may reflect what the seller has done (e.g., Proposal Sent) rather than where the buyer is in their actual decision process. 

• Definitional Loose Ends: Words like commit, upside, qualified, and close date sound precise while meaning vastly different things to different people.

• The Tech-Savior Delusion: A business may add a new forecasting tool, AI score, or deal review framework before it has defined what a real opportunity looks like in the first place, resulting in more reporting but no more clarity.

  
  

These issues matter. But the top three issues above are more universal because they sit closer to the culture and management system of the business. What gets measured, how truth is received, and how judgment is applied shape behavior every single week.

The Opportunity

The useful part is that many of these problems do not require expensive fixes. Before changing systems, buying tools, or redesigning the entire sales process, leadership can look at the forecast environment it has already created:

1. What behaviors are our metrics producing?

2. Is it safe for people to surface risk early?

3. Do we know exactly where judgment has been applied to the number?

4. Are we improving the forecast, or just adding pressure around it?

   
   

The forecast will never be perfect, and complex B2B sales will always involve uncertainty. But businesses can instantly improve performance simply by removing the friction, distortion, and noise they have unintentionally built into their own environment.