The share of calls that result in a booked job. The formula is bookings divided by calls — and everything contested about the number lives in which calls you count.
Booking rate = booked jobs ÷ a stated population of calls, over a stated period. The numerator is mostly settled: a job with a scheduled date counts, a "we'll call you back" does not. The denominator is where the number is won or lost. The same month of call records supports at least six defensible denominators — all inbound calls, answered calls, answered calls excluding non-opportunities, unique callers, and so on — and on real data the spread between the honest versions runs thirteen points apart.
Pick one denominator, write it down, and never quote the rate without it. The defensible default for CSR performance is bookings ÷ answered calls excluding non-opportunity calls: it isolates what the person on the phone could actually influence. For a whole-operation view, keep a second rate over all inbound calls — the gap between the two is your capacity problem, not your conversion problem. Both need the same period, the same call-record export, and a stated rule for callbacks from the same customer.
A booking rate says nothing about job value, and comparing rates across operations without segmenting by call mix is how vendor benchmarks mislead: an operation fielding mostly emergency calls will out-book a maintenance-heavy one with an identical team. What a benchmark can and cannot tell you covers the comparison problem in full.
Related terms: Call mix · Non-opportunity call · First-call resolution · the full glossary
Six defensible ways to compute it from the same month of calls, and which to standardise on.
The full six-denominator breakdown, and how to build a comparison you can defend.