What is cost per booked job, and how do you calculate it?
Cost per booked job is fully loaded marketing spend for a source divided by the jobs that source booked in the same window. It replaces cost per lead by moving the denominator out of the ad platform and into your field service system. The arithmetic is trivial. The work is deciding what spend counts, what a booked job is, and how spend in one month gets matched to jobs that book in the next.
The formula is trivial. The denominator is where it goes wrong
Take the spend attributable to a source, divide by the jobs that source produced, done. The division takes a second. Everything that makes the result trustworthy happens before it: which costs count as spend, which bookings count as jobs, and how you line the two up in time.
Cost per lead can be produced entirely inside an ad platform. Cost per booked job cannot, because the booking only exists in your operational system. Producing the metric at all requires joining the two systems and holding a stable identifier across them. That is the actual project.
Booked, completed, or invoiced — pick one and label it
There are three defensible versions of this metric and they answer different questions.
- Cost per booked job. Fastest signal, good for pacing. It includes jobs that later cancel, so it flatters sources with high cancellation rates.
- Cost per completed job. Slower and truer. You have to wait out your scheduling lag before the number stops moving.
- Cost per completed job with revenue attached. The version that supports budget decisions, because it can be compared against the gross profit that job type actually produces.
Spend happens now; the jobs show up later
Service work has lag between the lead and the money. If you divide this month's spend by this month's completed jobs, you are dividing one period's cause by another period's effect, and the metric will lie in both directions — flattering during a ramp-down, punishing during a ramp-up.
The fix is to cohort jobs by the date the lead arrived, not the date the invoice closed, and to let each cohort mature before you grade it. That single change reconciles most arguments between marketing and finance about whose number is right. See revenue intelligence for how cohorting is normally structured.
The number is meaningless without a ceiling next to it
Cost per booked job is not good or bad on its own. It is only good or bad relative to what a job of that type can afford to pay for acquisition. A drain clearing call and a system replacement are not the same product and should never share a target.
So report it segmented by job type, with the affordable ceiling on the same row. That turns a metric into a decision. Reporting it blended across every job type produces a number that is technically correct and operationally useless — the recurring theme of marketing intelligence work.
Topics: cost per booked job · marketing metrics · attribution · budget
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.