Why does our close rate change depending on who calculates it?
Because the denominator changes. Counting by estimate document divides sold estimates by all estimates, and good-better-best quoting inflates that denominator. Counting by opportunity divides won customers by customers who received any quote. Counting by visit divides sold visits by dispatched visits. All three are defensible, none are comparable to each other, and none are comparable to an industry number whose definition you do not know.
Three denominators, three different businesses described
Close rate by estimate answers: of the documents we produced, how many were signed. Close rate by opportunity answers: of the customers we quoted, how many bought something. Close rate by dispatched visit answers: of the trucks we sent, how many produced sold work. The third one is the closest thing to a business-level number, because it includes visits that never produced an estimate at all.
A shop can post a strong estimate-level close rate while a third of its visits end with no quote written. That failure is invisible in the first two numbers and expensive in the third.
Where multi-option quoting breaks the math
Present good, better and best as three estimate records and sell one, and the estimate-level close rate for that customer is one in three. Present the same three options as one estimate with three line-item choices and it is one in one. Same visit, same sale, same customer, and the reported close rate differs by a factor of three.
This is not a hypothetical rounding issue. It is the single most common reason two locations of the same company appear to have wildly different sales performance when nothing differs except how the office was trained to build quotes. Normalizing this is ordinary data integration work — collapse option sets to one opportunity before you count.
Match the metric to the decision
- Coaching an individual. Use opportunity-level close rate within job type. It is the fairest to the person and the least sensitive to quoting style.
- Judging a marketing source. Use visit-level or lead-level conversion. The source is responsible for what walks in the door, not for how the quote was formatted.
- Forecasting revenue. Use expected value on the open pipeline instead of a close rate at all. Close rate multiplied by total open value overstates almost every pipeline.
Fix the definition once, in the data layer
The durable fix is not a policy memo. It is a defined metric in the reporting layer, with the collapsing rules and exclusions written down, so every dashboard, brief and export derives from the same computation.
When people argue about a close rate in a meeting, they are almost never arguing about performance. They are arguing about two denominators nobody wrote down. See how we handle metric definitions in business intelligence work and why rep coaching depends on getting the denominator right before anyone is measured on it.
Topics: close rate · estimates · definitions · reporting · benchmarks
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.