Which channels produced paid collision jobs?

Direct repair programs bring volume at a negotiated rate. Search brings fewer jobs at full retail. Shops rarely put the two on the same page, so the trade-off between volume and margin is argued rather than measured.

Added to the lead file, each job tagged with the referring insurer, the DRP list ranks alongside advertising on the same revenue axis — and, if you export margin instead of revenue as the amount, on profit rather than sales.

Referral partners send a substantial share of jobs, and the relationship is usually maintained on instinct rather than evidence about which partners actually produce.

A partner's referred jobs can go into the lead file, tagged with the partner. Ranked on completed job revenue, they show which relationships are worth the effort and which have quietly stopped producing.

Nothing in a body shop is worth what the first estimate said. The vehicle comes apart, hidden damage appears, the supplement goes back to the adjuster, and the final invoice lands somewhere else entirely — often a long way above the figure written in the parking lot on day one. If the file you export is estimates, every channel is being ranked on money the business never collected. Export closed, invoiced jobs at the final agreed amount, and the ranking is built on what actually arrived in the bank.

That is more than a tidiness point, because the gap between estimate and final is not spread evenly. Newer vehicles with calibration requirements, aluminium panels and structural work supplement quite differently from a bumper cover and a blend. A source of work that brings heavier damage looks ordinary on estimate value and substantial on final invoice, and a shop deciding where to advertise on the strength of the first number is deciding on the wrong one entirely. The correction costs nothing; it is a different column in the same export.

The date deserves a decision as well. A job opened in one month and invoiced six weeks later can be booked on either, and period comparisons only mean something when the convention is identical in every import. Most shops use the invoice or closed date, because that is when the work is recognized and it keeps the report agreeing with the accounts. That matters more here than in most trades, since the two dates routinely fall in different months and a shop with long cycle times feels it in every comparison.

Parts on back order, a rental authorization that takes three days, a calibration that has to go out to somebody else, a total-loss decision that sits for a week. The interval between the tow-in and the invoice is measured in weeks and it is not stable from job to job. Any measurement that closes its books on a monthly window captures the vehicles that moved quickly and loses the ones that did not, which under-counts precisely the heavy jobs a shop wants more of.

Because there is no window here, the length of that interval is irrelevant. The vehicle that arrived in February and was collected in April matches the February inquiry, and the April revenue lands against whatever produced it. For a shop trying to decide whether a search campaign earns its keep, that one behavior routinely separates a channel that looked break-even from one that was profitable and being measured against the wrong month. It also removes the temptation to judge a campaign before its work has finished.

Total losses need handling of their own. A vehicle towed in, assessed and written off produces teardown and storage revenue rather than a repair, and it occupies the shop either way. Leaving those jobs out understates what a towing relationship or an emergency search campaign genuinely produces; putting them in flatters the average job value. A job-type column mapped as Product / service shows them as their own line, while the channel ranking counts them with everything else, which is the closest the report comes to answering both versions of the question.

Two things come as a surprise almost every time. The unattributed share is large — often close to half — because vehicles arrive on the back of a truck at three in the morning, because an adjuster steered them, or because the owner's father-in-law has used the shop for twenty years. None of that is advertising, and the report says so rather than quietly dividing it among whichever campaigns were live. That honesty is exactly why the rest of the numbers are worth arguing about.

The second surprise can be how far apart the sources sit on average job value. One relationship may produce a steady flow of small, fast, low-margin work while another produces far fewer jobs worth several times as much. Both look like a job on a monthly summary. On invoiced revenue they are not comparable at all, and the capacity question — which work to take when the shop is already full — becomes a different conversation entirely.

What changes after the first report is usually not the advertising. It is the referral effort. Once tow operators, dealers and insurers sit on one axis with search and social, the relationships that have quietly stopped producing become visible, and so do the ones nobody has maintained since the last manager left. That is a week of phone calls rather than a budget change, and in this trade it is frequently worth more than anything you could do to a campaign.