An inbound call that could never have resulted in a booked job regardless of how it was handled.
A call is a non-opportunity when no handling could have produced a booked job: a wrong number, an existing customer confirming an arrival window, a supplier, a job applicant, a caller chasing an invoice. The test is counterfactual: could this call have booked, not did it, which is what separates it from a declined estimate, which was an opportunity that did not convert.
Non-opportunity calls belong in an answer-rate denominator (the phone should still be answered) and distort a booking-rate denominator if left in. "Bookings ÷ answered calls" and "bookings ÷ answered calls excluding non-opportunities" are both defensible and can sit thirteen points apart on the same data. Quoting either without saying which is how phone-performance numbers stop being comparable.
Write the exclusion list down and audit it: the category attracts wishful classification, because every call moved into it raises the booking rate. A declined quote is not a non-opportunity. A price-shopper is not a non-opportunity. If the caller could have booked and did not, the call stays in the denominator, that discipline is most of what makes the resulting rate worth quoting.
Related terms: Booking rate · Call mix · First-call resolution · the full glossary
The six-denominator problem in full, and which to standardise on.
How the major platforms define these metrics in their own documentation.
Six figures you already know gives you a screening range in about a minute. The full diagnostic prices it from your own call records in two to three weeks, for a fixed $4,500, and costs you about an hour of someone's time to pull the exports.
If your data cannot support a defensible figure, you get a readiness report naming exactly what to start capturing, at no charge, and no number is published.