Most forecast reviews argue about which deals will close. Far fewer argue about how much those deals will close for. That second question moves the number just as much, and it fails silently because nobody has to admit anything went wrong.
One pattern makes it concrete: a pipeline carrying an average deal size of $80,000 while closed won deals average $40,000. Every forecast built on recorded amounts in that business runs high before a single deal slips.
Why do deals close for less than the amount in the CRM?
Recorded amounts capture the deal a seller hopes to sell, and closed amounts capture the deal a buyer agreed to buy.Four mechanisms produce the gap, and they are not equally common across businesses.
Optimistic entry. The amount is set at opportunity creation from a full-scope assumption, before anyone knows what the buyer will actually fund. Nobody revises it downward as the scope narrows, because a lower number attracts questions.
Scope reduction. The buyer buys two of the four modules, or 40 seats instead of 100, or a pilot instead of a rollout. The deal closes as a win at a fraction of its recorded value.
Term compression. An annual contract replaces the three-year deal in the record, or the start date moves so only part of the value lands in the period.
Competitive pricing pressure. A new competitor enters the category and average deal size falls across the board. That is a market change rather than a seller behavior change, and it shows up as a realization gap that widens across every rep at once.
How do I measure the realization gap?
Fix a reference point in the funnel and compare recorded amount at that point against final closed amount.Pick the stage where a real number should exist, usually proposal or its equivalent. Then compute realization for closed won deals across the last four to six quarters.
| Cut | Recorded amount at proposal | Closed won amount | Realization rate | What it suggests |
|---|---|---|---|---|
| Enterprise | Scope negotiation and multi-year term risk | |||
| Mid market | Standard packaging, watch for discount drift | |||
| SMB | Seat count assumptions at entry | |||
| Product A | ||||
| Product B |
Also compute the same rate for closed lost deals. If lost deals carried far higher recorded amounts than won deals, the gap is partly a qualification problem: your large opportunities are the ones that fail.
Is this a discounting problem or a scoping problem?
Compare unit price against units sold and the answer separates cleanly.If the price per seat or per unit fell while volume held, you have a discounting problem. That lives with deal desk, approval thresholds, and the competitive positioning that makes discounting feel necessary.
If unit price held and the number of units fell, you have a scoping problem. That lives earlier, in how the opportunity was sized at creation and whether the buying committee ever agreed to the full scope.
If both fell together, check what changed in the market. Pricing pressure from a new entrant tends to hit price first and scope second, as buyers use the alternative quote to renegotiate both.
The distinction matters because the two fixes point in opposite directions. Tightening discount approval on a scoping problem slows deals down and changes nothing about the outcome.
What does the gap do to the forecast and to coverage?
It biases both high, and neither one will report the bias to you. Pipeline coverage is computed on recorded amounts. A team running 3.5x coverage with a 50% realization rate is functionally covered at a much lower multiple against revenue that will actually book. The ratio looks fine right up until the quarter closes.Weighted forecasting inherits the same problem. Applying a stage probability to an inflated amount produces a confidently wrong number, which is why weighted pipeline should be built on realization-adjusted amounts rather than raw ones.
The failure mode is specific. Deals convert at expected rates, close dates hold, the commit list performs, and the quarter still misses by 10 or 15 percent. Value, not volume, produced the miss, and no deal-level review will surface it because every deal individually looks like a win.
How do I correct for it without punishing reps?
Adjust the model, not the rep's number.Leave the recorded amount as the seller's view of full scope. Then apply a realization factor derived from your own history, cut by segment, product, and stage, inside the forecast model. The forecast reflects what your business actually books while the pipeline record still reflects what the team is pursuing.
Forcing reps to enter conservative amounts creates a second bias, and a bias that lives in human judgment is far harder to measure and correct than one that lives in a model coefficient. It also destroys the signal you need, because you can no longer tell the difference between a genuinely smaller deal and a rep hedging.
Two operational habits keep the correction honest. Require an amount update at proposal generation so the number reflects a real quote rather than a creation-time guess. And treat any amount change as meaningful activity on the opportunity, alongside stage and close date changes, because a shrinking amount late in a cycle is one of the clearest risk signals available.
What realization rate should I expect?
Use your own trailing baseline. There is no useful cross-company benchmark for this number.Realization depends on how your business defines an opportunity amount, when it gets entered, whether it includes services and multi-year value, and how your packaging works. Two companies with identical sales performance can post very different realization rates purely on data entry conventions.
What matters is the trend and the spread. A stable rate is forecastable at any level, even a low one, because a consistent input produces an accurate model. Everyone believes their data is uniquely bad, and it almost never is. Garbage in does not have to mean garbage out, as long as the garbage is consistent.
The number to worry about is a rate that is moving. A realization rate falling quarter over quarter means the market or your packaging changed, and every forecast built before that shift is running on old assumptions. Track it next to forecast accuracy so you can tell which one caused the other.
Frequently Asked Questions
What is a deal amount realization rate?
It is closed won revenue divided by the amount those same opportunities carried in the CRM at a fixed reference point, such as entry into the proposal stage. It tells you how much of your recorded pipeline value survives to become revenue.
Is a realization gap a discounting problem?
Not usually. Discounting is one input. Scope reduction, shorter contract terms, removed products, and optimistic amounts entered at opportunity creation all move the same number, and each requires a different fix.
What does a realization gap do to pipeline coverage?
It inflates it. Coverage is calculated on recorded amounts, so if deals routinely close for a fraction of what they carry, the ratio overstates how much revenue the pipeline can produce. A team can look adequately covered and still miss on value rather than volume.
How large can the gap get in practice?
One pattern makes it concrete: pipeline with an average deal size of $80,000 against closed won deals averaging $40,000. When the gap runs that wide, every forecast built on recorded amounts is structurally high before any deal slips.
Should I make reps lower their opportunity amounts?
No. Adjust the model instead. Apply a realization factor derived from your own history by segment and stage so the forecast reflects reality, and keep the recorded amount as the rep's view of full scope. Forcing reps to discount their own numbers creates a second bias that is harder to correct.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo