Optimized Sales Optimized Marketing Target Accounts For CROs For CFOs For CMOs Blog News Glossary Compare Tools About Schedule a Demo
Pipeline Analytics

Why Your Pipeline Average Deal Size Lies

Pete Furseth 6 min read
average deal sizepipeline qualityforecast accuracypipeline coverage
Why Your Pipeline Average Deal Size Lies
Home/ Blog/ Why Your Pipeline Average Deal Size Lies

There is a number sitting in almost every CRM that quietly invalidates the coverage ratio built on top of it, and it takes about ten minutes to check.

Most deals close for less than the value they carry in Salesforce.

The measurement

Take the average deal size of your open pipeline. Take the average deal size of deals you closed and won over a comparable period. Compare them.

A pipeline with an average deal size of $80,000 against closed-won deals averaging $40,000 is not a modeling curiosity. It means every deal in that pipeline is carrying roughly double the value it will actually produce, and every metric derived from pipeline value inherits the error.

The coverage ratio is the obvious casualty. Coverage is pipeline value over goal. If the numerator is inflated twofold, a reported 4x is a real 2x, and the team believing it is in good shape is running at half the cover it thinks.

Put this to work on your numbers
Run your own numbers with the free Forecast Accuracy Scorecard, then see how ORM builds it into a custom model.

Why the gap exists

The gap is structural rather than dishonest. Opportunity amounts get entered early, when the deal is a hypothesis and the number is the most optimistic reasonable scoping. Then several things happen that only push in one direction.

Scope narrows during technical evaluation. Procurement removes line items. Negotiation applies a discount. Multi-year terms get shortened. Each of these reduces the real value, and almost no sales process includes a step that requires revising the recorded amount downward when they do.

So the amount field records the largest version of the deal that anyone ever believed in, and it stays there until close.

What it does downstream

MetricHow the inflation reaches it
Pipeline coverageInflated numerator, proportionally inflated ratio
Weighted forecastStage weight applied to an overstated amount
Rep capacity planningTerritory potential overstated, headcount modeled against it
Win rate by valueDenominators overstated, apparent value-weighted win rate depressed
The weighted forecast case is the most damaging, because the error compounds. Applying a stage weight to an inflated amount produces a number that looks methodical and is wrong in two dimensions at once. That interaction is covered in why stage-weighted forecasting misses.

Fixing it without a process overhaul

The instinct is to demand better data entry. That rarely works, and it is not necessary.

The better move is to apply an empirical adjustment. If your closed-won average is consistently around half of your pipeline average, then a realization factor of roughly 0.5 applied to open pipeline value produces a materially more honest coverage number without asking a single rep to change behavior.

Three practical steps:

1. Measure the ratio by segment. Enterprise and commercial motions rarely share a realization factor, and a blended number hides both. 2. Apply it at the reporting layer, not the record layer. Leave the CRM amount alone and adjust in the model, so nobody is arguing about whose deal got marked down. 3. Re-measure quarterly. The factor moves when pricing changes or when competitive pressure increases, and pressure on deal size is usually a signal of competition in the market rather than a data problem.

The wider point about data quality

Teams often treat a gap like this as evidence that their data is uniquely bad and that forecasting is therefore impossible for them. It is not, and they are not unusual.

Garbage in does not have to equal garbage out. As long as the distortion is consistent, it is measurable, and anything measurable can be corrected for in the model. A pipeline that consistently overstates by a factor of two is far easier to forecast against than a pipeline that overstates by an unpredictable amount. That argument is made in full in your data is not uniquely bad.

For the underlying definitions see average deal size and pipeline coverage.

Frequently Asked Questions

Why do deals close for less than their CRM value?

Recorded opportunity amounts are usually entered early, before scoping, procurement and negotiation have reduced them, and there is rarely a process that revises them downward as the deal progresses. The result is a systematic upward bias in pipeline value that only resolves at close.

How do I measure the gap?

Compare the average deal size of open pipeline against the average deal size of closed-won deals over the same period. If pipeline averages $80,000 and closed-won averages $40,000, every coverage number you report is overstated by a factor of two.

Does this affect pipeline coverage?

Directly. Coverage is pipeline value divided by goal, so if pipeline value is systematically inflated then coverage is systematically inflated by the same proportion. A reported 4x built on doubled deal values is a real 2x.

Why do recorded opportunity amounts only move in one direction?

Because the pressures on them are one-directional. Scope narrows during technical evaluation, procurement removes line items, negotiation applies discounts, and multi-year terms get shortened. Almost no sales process includes a step requiring the recorded amount to be revised down when those happen.

Should I fix this by enforcing better data entry?

Usually not. Enforcing entry rarely works and is not necessary. Apply an empirical realization factor at the reporting layer instead, leaving the CRM record alone, so nobody is arguing about whose deal got marked down.

Does this problem mean my data is unusable?

No. A consistent distortion is a coefficient rather than a blocker. A pipeline that reliably overstates by a factor of two is easier to forecast against than one that overstates by an unpredictable amount.

Frequently Asked Questions

Why do deals close for less than their CRM value?

Recorded opportunity amounts are usually entered early, before scoping, procurement and negotiation have reduced them, and there is rarely a process that revises them downward as the deal progresses. The result is a systematic upward bias in pipeline value that only resolves at close.

How do I measure the gap?

Compare the average deal size of open pipeline against the average deal size of closed-won deals over the same period. If pipeline averages $80,000 and closed-won averages $40,000, every coverage number you report is overstated by a factor of two.

Does this affect pipeline coverage?

Directly. Coverage is pipeline value divided by goal, so if pipeline value is systematically inflated then coverage is systematically inflated by the same proportion. A reported 4x built on doubled deal values is a real 2x.

Why do recorded opportunity amounts only move in one direction?

Because the pressures on them are one-directional. Scope narrows during technical evaluation, procurement removes line items, negotiation applies discounts, and multi-year terms get shortened. Almost no sales process includes a step requiring the recorded amount to be revised down when those happen.

Should I fix this by enforcing better data entry?

Usually not. Enforcing entry rarely works and is not necessary. Apply an empirical realization factor at the reporting layer instead, leaving the CRM record alone, so nobody is arguing about whose deal got marked down.

Does this problem mean my data is unusable?

No. A consistent distortion is a coefficient rather than a blocker. A pipeline that reliably overstates by a factor of two is easier to forecast against than one that overstates by an unpredictable amount.

PF
Pete Furseth
ORM Technologies
Pete has built custom revenue forecast models for B2B SaaS companies for over a decade.

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