Why Traditional Sales Forecasting Fails - and What Forward-Thinking Leaders Are Doing Instead

Sales forecasting traditionally relies on deal “rollups.” Deals are categorized, typically as Commit, Best Case, or Pipeline; amounts are totalled for each category; forecasts are debated; named deal lists are delivered, and leadership aligns to fulfil the Commit plus whatever they can from less-likely categories. Despite advances in CRM systems, deal inspection applications, and advanced analytics, this process remains fundamentally unchanged—and fundamentally flawed.

The problem isn’t effort, discipline, or lack of data. It’s the method.

The Limits of Rollups and Named Deal Lists

CRM systems provide three core opportunity fields: stage, amount, and close date. Each conveys independent information about deal progress, value, and anticipated timing. Rollups often add Forecast Category calls (Commit, Best Case…) to each deal. But Forecast Category overlaps and may contradict stage and close date—weakening those signals and muddying the picture rather than clarifying it.

Businesses use Forecast Category and other types of rollups for two primary reasons:

1. To produce accurate forecasts

2. To focus effort on deals that will maximize sales

In practice, neither objective is achieved and the use of rollups actually hinders these goals.

Across dozens of real-world companies and quarters, rollups rarely come close to actual sales. Management opinions are sometimes added in an attempt to impose order, but these are ad hoc and non-repeatable. Lists of the most certain deals are more reliable, but they steer execution toward the safest outcome rather than the maximum possible. Rollups don’t deliver accurate forecasts or maximize sales. 

Most revenue technology vendors don’t fix these problems—they automate it. Recognizing these limitations, they also provide “AI” forecasts—some claiming ±2% accuracy.

Why Accuracy Isn’t the Real Goal

There’s a deeper issue hiding beneath forecast rollups and AI forecasting: the objective is to maximize sales, not forecast accuracy. Accuracy is easy if you aim low.

Imagine a football team projected to lose a game by 3 points. Would they play expecting to deliver on that? No! They would do everything they could to change the outcome. Sales leaders have control over outcomes too. Headcount, pricing, demand generation, resource allocation, and deal focus all influence results. When a forecast predicts a shortfall, the question isn’t “Is the number accurate?” It’s “How can we improve this?”

Rollups will tell you to win more deals. Admittedly, at the end of a quarter that’s all you can do. But before you win a deal, you need to advance it through your sales process. To maximize future sales, you should be looking to advance the right deals now.

Forecast Distributions: A More Honest Conversation

Rollups reduce a complex system to a named deal list and a single Commit figure—implicitly treating it as certain. There’s a reason it’s called “Commit.” The result is surprise, disappointment, and eroding trust. Reporting a low/high range helps, but how certain are those figures? There’s a lot of room for interpretation.

Modern forecasting can and should reflect reality: many possible outcomes, each with a likelihood.

A forecast distribution does exactly that. Instead of declaring one number, it shows the relative chances of all possible results—for both existing pipeline and new pipeline creation.

This approach enables a very different executive conversation:

• “There’s a 90% chance we will beat this lower figure.”

• “We have an even chance of beating this midpoint.”

• “Here’s how those odds change if we advance these specific deals or increase demand generation in these segments.”

Uncertainty is explicit, not hidden, not implied. And leadership can see how resource allocation influences outcomes.

Why Rollups and Named Deal Lists Don’t Maximize Sales

Rollups don’t just hurt forecasting—they hurt sales.

Named deal lists (Commit, late stage… whatever) focus attention on deals most likely to close soon. You have to do that at times, but that’s myopic and rarely optimal. You can do better.

  ​When a forecast predicts a shortfall, the question isn’t “Is the number accurate?” It’s “How can we improve this?”   

Focus matters. No one can say specifically which deals will close. But, with a good model, you can estimate sales and risk for a basket of deals. Then it’s a matter of figuring out “which basket should we focus on to maximize sales for an acceptable level of risk?”

This is a portfolio optimization problem, not a deal rollup problem.

One effective approach borrows a Nobel Prize–winning technique from finance: efficient frontier optimization. Rather than rating deals by likelihood alone, this method identifies a portfolio of opportunities that maximizes expected sales for a given level of risk. The result is a “Focus Ranking”—a burn-down list that sellers work through each day, applying their own judgment to determine where additional effort can advance deals.

It’s systematic instead of ad hoc optimization. In extensive side-by-side tests across multiple companies and quarters, Focus rankings delivered about 60% more sales than same-sized company priority lists (Commit, late stage, biggest deals). 

Less Theater. More Selling.

Traditional forecasting consumes enormous resources: forecast calls, rollup submissions, management overrides... That’s time not spent selling.

A modern approach simplifies:

• Maintain disciplined use of stages, amounts, and close dates

• Automate forecasts instead of judgement rollups

• Report forecast distributions, not point estimates

• Focus on deal portfolios that maximize expected sales

This shift doesn’t abandon deal reviews. These are a critical part of figuring out if you are stuck and how to advance your cause. Nor does it abandon judgment—it puts judgment where it belongs: deciding how to change the outcome.

A Better Path Forward

Forecast rollups were once the only option. Today, they’re a legacy.

Forward-thinking leaders are moving beyond rollups to probabilistic forecasts and optimized focus. The payoff is substantial: automated realistic forecasts, more sales time, better board conversations, and more sales. 

At Funnelcast, we’ve built an application to support this approach—from forecast distributions to daily optimized Focus rankings. 

When you stop treating forecasting as reporting—and start treating it as informing your behavior—you don’t just report the past or predict the future.

You change it.

This article draws on research originally published in Funnelcast’s Forecast Category Considered Harmful blog.