The arithmetic people usually do is missed calls multiplied by average job value. That number is always wrong and always too big. Here is the version that survives being checked.
The temptation is understandable. You missed 120 calls, your average job is worth $520, therefore you lost $62,400. It is a big number, it is easy to compute, and it will not survive ten seconds of scrutiny from anyone with a finance background — which is a problem, because that is usually the person who has to approve doing something about it.
Three things are wrong with it, and each one makes it smaller.
A customer with a burst pipe does not give up because you did not answer. They call again in four minutes, or they call you back after trying someone else, or they were already in your queue twice that morning. A meaningful share of abandoned calls are recovered without you doing anything at all.
This is the single most important number in the whole calculation and it is the one almost nobody measures. If your call records carry caller ID you can measure it directly: take every abandoned call, look for a later inbound call from the same number within seventy-two hours, and count the matches. That fraction is your recovery rate. Whatever it is, it is not zero, and assuming zero is what produces headline figures nobody believes.
Of the calls that genuinely never came back, only some would have booked. The right multiplier is your own booking rate on answered calls — not one, and not the rate on your best day. An abandoned call is, on average, a call like your other calls.
And your booking rate has to be defined consistently, or this step imports whatever distortion is in the denominator — which is a bigger problem than it sounds.
Even with perfect staffing you would not capture every one of those calls. Some intervals spike beyond any sensible roster. Applying an attainability factor — the share of the modelled gap that could realistically be closed — keeps the figure honest.
Then convert to contribution margin. Owners think in revenue and revenue is the bigger number, but a controller will immediately ask what it costs to deliver that work, and a figure quoted in margin has already answered them.
| Abandoned calls | measured |
| × (1 − recovery rate) | measured from caller ID, or a conservative default |
| × booking rate on answered calls | measured, one fixed definition |
| × attainability | stated assumption |
| × average completed job value | measured |
| × contribution margin | stated by you |
Two of those six are assumptions rather than measurements, which is why the output should be a range with a most-likely value rather than a single figure. A point estimate implies a precision the inputs do not support, and claiming it is the fastest way to lose the room.
One more thing the simple version hides. Abandoned calls are not spread evenly across the month. They concentrate in the half-hour intervals where demand exceeded staffing, and those intervals are usually the same ones every week. A monthly total tells you the size of the problem; interval resolution tells you where it is, which is the part you can actually schedule against.