Forecast Retrospective vs Win Loss Review: Which One Explains the Miss?

What Is the Difference Between a Forecast Retrospective and a Win Loss Review?

A forecast retrospective asks why the number was wrong. A win loss review asks why the deals went the way they did. Those are separate questions with separate answers, and a team that runs only one of them will misdiagnose half its quarters.

The distinction is easiest to see in the cases where the two disagree. A team can win almost every deal it committed and still miss badly, because the forecast assumed in-quarter creation that never happened. A team can lose several winnable deals and still land on the number, because the model already priced in a lower win rate. In the first case, deal execution was fine and the forecasting process failed. In the second, the forecast was sound and the selling was not.

Running only win loss reviews produces a permanent narrative that the number missed because reps lost deals. Running only forecast retrospectives produces an equally wrong narrative that the process needs another field.

What Does a Forecast Retrospective Examine?

The gap between what was predicted and what happened, decomposed into the reasons. The subject is the prediction, not the reps.

Start with the week-by-week trace. What did the roll-up say in week two, week six, and week ten, and where did it move. A forecast that was accurate in the final week and wrong in week two did not help anyone, because by the last week the quarter has already happened. Value comes from knowing the shape of the quarter early enough to act on it.

Then decompose the number into its sources. Carry-over deals that were in pipeline on day one and expected to close. In-quarter business that was created and closed inside the period. Deals pulled forward from future quarters, including the discounting that usually buys them. Most teams over-trust the deals they can see in the CRM and under-model the revenue that has to be created inside the period, so the retrospective almost always finds the miss in the second bucket.

Finally, test the assumptions. Average deal size assumed against average deal size realized. Win rate assumed against win rate delivered. Cycle length assumed against cycle length observed. Each gap points at a specific broken input.

What Does a Win Loss Review Examine?

Buyer behavior on individual deals, sourced from the buyer rather than the rep.

The core of a real win loss program is talking to people who bought and people who did not. Reps produce loss reasons that cluster around price, because price is the reason buyers give when they do not want to explain the real one. Buyers, asked by someone who was not on the deal, produce far more specific answers about evaluation criteria, internal politics, and the alternative they chose.

The output is a pattern rather than a verdict per deal. Losses concentrated in one competitor. Losses concentrated at one stage. A no-decision rate climbing in a segment. Wins that consistently included a specific proof step early.

That pattern is what changes the sales motion, which then changes win rate, which eventually changes the forecast. The path from a win loss review to a better number runs through behavior, and it takes at least a quarter to show up.

How Do the Two Compare?

One fixes the prediction. The other fixes the selling. The cadence row matters more than it looks. A win loss program that only runs at quarter end interviews buyers months after a decision, and buyers forget. Sample continuously and review the pattern monthly.

What Usually Turns Out to Be the Cause?

Most forecasts miss because something changed and the forecast was built on old assumptions. The mechanism is stale inputs rather than dishonest reps.

Four changes account for a large share of it. A new competitor enters and creates pricing pressure, so average deal size falls. Interest rates rise, private equity slows capital deployment, valuations drop, companies cut cost to protect earnings, and fewer of them buy, so win rates fall. Broad uncertainty, whether from a pandemic or an AI-driven repricing of technology budgets, produces fewer decisions and longer paths from qualified to closed. Territories get redrawn and reps are distracted, so execution suffers while the coverage ratio looks perfectly healthy.

The visible symptoms are the same in all four cases. Pipeline stagnates, deals close for less than their recorded value, and win rates decline. A retrospective that stops at those symptoms produces no fix. One that names which underlying change occurred produces a specific correction to the model.

Seasonality is the other assumption teams get wrong. Q2 and Q4 generally run stronger than Q1 and Q3, and the third month of a quarter runs stronger than the first two. A forecast that spreads a quarter evenly across three months will look wrong in weeks one through eight and fine at the end, which is the worst possible pattern for anyone trying to make decisions in month one.

Is the Data the Problem?

Almost never, and the belief that it is stalls more forecasting programs than any technical limitation.

Every revenue team believes its CRM data is uniquely bad and that this is why it cannot run the business the way it wants. It is a shared delusion. Everyone has messy data, and messy data still supports accurate prediction as long as it is consistently messy. Consistency is the requirement, not cleanliness.

What actually breaks a forecast is a model that cannot respond when conditions shift. Manual processes on new and expansion business generally reach around ninety percent accuracy, at a real cost in effort, and the number they produce is static from the moment it is built. ORM targets ninety-five percent without manual adjustment and holds it from day one to day ninety, which changes the retrospective from an autopsy into a tuning exercise. The practices that support that shift are covered in sales forecasting best practices.

How Should the Two Feed Each Other?

Run the retrospective first, then use it to choose which deals get interviewed.

The retrospective tells you where the number broke. If it broke on win rate in the enterprise segment, interview enterprise losses from that quarter. If it broke on average deal size, interview the wins and ask what happened to scope between proposal and signature. Sampling losses at random produces interesting anecdotes and no correction.

Then close the loop in the other direction. Every pattern a win loss review establishes should become an input the forecast model or the process actually uses, whether that is a stage exit rule, a revised assumption, or a change in how close dates are set. Otherwise both meetings become rituals, and forecast accuracy stays exactly where it was.