The True Cost of a Return Trip
Why this matters
Ask a tech what a return trip cost and you will get the time he was on site. Forty minutes, maybe an hour. That answer is not dishonest, it is just the only part he can see. The rest of it is spread across four other people's days and one empty slot on a schedule, and because no single piece is large, nobody ever adds them up. Shops that have never built this number consistently under-invest in prevention, because every prevention step has a visible cost and the thing it prevents does not. Put a real multiplier on it and the decisions get easy.
The seven things a return trip consumes
- On-site labor on the return. The visible piece. Usually the smallest.
- Round-trip drive. Paid time either way, and it does not shrink with the size of the fix. A five-minute repair carries the same drive as a two-hour one.
- Vehicle cost for the trip. Fuel, tires, brakes, depreciation, all of which accrue per mile whether or not the mile was billable.
- Restock and prep. Pulling the part, staging the truck, and where the return exists because a part was missing, a supply-house run.
- Office touch. The inbound call, the scheduling, the dispatch, the notes, the close-out, and the follow-up call afterwards. Small per event, and it is several people's small.
- Displacement. The slot the return trip occupies could have held billable work. On a full board this is the largest component of all.
- Customer standing. A return trip resets the customer's confidence and measurably affects whether they call you next time and whether they mention you to anyone. Real, and not something to fake a number for. Name it, hold it qualitatively, and do not let its unquantifiability keep it out of the decision.
The visible hour and the loaded hour
Items 1 through 5 are paid time you can count directly. Add them up in hours and divide by the on-site hour, and you have the loaded multiplier for a return trip before displacement.
Most shops that measure this for the first time land somewhere around 2.5 to 3.5 times the visible on-site hour on a typical short return, and the ratio gets worse as the fix gets shorter, because drive and office touch are close to fixed while on-site time is not. The quickest return trips are the worst deals in your entire operation.
Displacement is the largest component and the one nobody counts
The other six are costs you incur. Displacement is revenue you never earn, and it only exists when the schedule was actually full. That condition matters and it is the reason displacement gets argued about, so state it explicitly: count displacement at full value when the board is booked and you turned work away or pushed it out, count it at zero in a genuinely slow week, and do not split the difference to feel balanced.
The conversion is your billable recovery ratio, which is the share of paid tech hours that actually get invoiced. If the return trip consumes 2.0 hours of a tech's day and your recovery ratio is 60%, then those 2.0 hours of capacity would have produced about 1.2 billed hours had they been available.
Building your own return-trip multiplier
Work a typical short return, in hours, all in the same unit:
- On-site labor: 0.8 hours
- Round-trip drive: 0.9 hours
- Restock and prep: 0.3 hours
- Office touch across the call, schedule, dispatch and close: 0.4 hours
Paid time consumed: 0.8 + 0.9 + 0.3 + 0.4 = 2.4 hours, against 0.8 hours of visible on-site work. That is a 3.0x loaded multiplier before displacement, on a fix a tech would describe as "about forty-five minutes."
Now displacement. The field portion is 0.8 + 0.9 + 0.3 = 2.0 tech hours. At a 60% billable recovery ratio, those 2.0 hours of capacity would have yielded about 1.2 billed hours.
So on a full board, one return trip costs 2.4 hours of paid time and forgoes about 1.2 billable hours, for 3.6 hours of total impact against 0.8 hours of visible work: a 4.5x multiplier. In a slow week, displacement is zero and the multiplier is back to 3.0x. Both numbers are correct in their own condition, and a shop should carry both rather than averaging them into one that is true in neither.
Vehicle cost per mile is deliberately left out of the hour arithmetic above rather than converted, because mixing a per-mile cost into an hours model is where these calculations usually go wrong. Track it separately and add it as its own line when you compare against a prevention step that also has a per-mile component.
What the multiplier actually buys you
It sizes prevention. A pre-departure verification step that takes 0.15 hours, which is 9 minutes, on every job costs 12 x 0.15 = 1.8 hours across twelve jobs. One prevented return trip saves 3.6 hours of impact on a full board. So that step breaks even if it prevents one return in every 24 jobs, and returns two-for-one if it prevents one in every 12. Most shops asked to spend 9 minutes on every job flinch, and most shops asked whether they could prevent one return trip in twenty-four say yes without hesitating. Same question.
It settles the stay-late decision. A tech is 1.5 hours from finishing at the end of the day. Staying costs 1.5 hours at an overtime premium of 1.5 times, so 2.25 hour-equivalents. Coming back tomorrow costs 3.6 hours of impact. Staying wins, and it is not close.
That comparison has a hard limit. Do not push a tired crew through a task that carries a live hazard at the end of a long day: anything on a gas-carrying line, anything requiring work on energized equipment, anything on a pressurized or stored-energy system, and anything at height. Those get de-energized or shut down, locked and tagged where applicable, left in a documented safe state, and scheduled fresh. The multiplier is a scheduling tool, not an override on fatigue and hazard.
It sets the truck-stock threshold. Take a part that has caused one return trip per quarter for the last year. At 3.6 hours of impact per trip, that part has cost about 4 x 3.6 = 14.4 hours over the year. Weigh that against what carrying it actually costs: the shelf space, the capital tied up, and the risk it becomes obsolete on the truck. Both of those are real and neither is usually worth 14 hours. Run the same comparison on the five parts your techs most often go back for, and the stocking list writes itself.
It prices structurally return-prone work. Some job types genuinely carry a higher return rate for reasons that are not fixable: unknown equipment, restricted access, third-party coordination. If a type produces one return trip for every four jobs, each costing 3.6 hours of impact, that is 0.9 hours per job of embedded return cost, which belongs in the estimate.
What one week looks like
A shop running five short return trips a week at the numbers above spends 5 x 2.4 = 12.0 hours of paid time and forgoes 5 x 1.2 = 6.0 billable hours. That is 18.0 hours of weekly impact, which against a 40-hour tech week is 45% of one technician. Five short trips nobody logged, and the shop has quietly assigned nearly half a tech to them.
Run that against your headcount before deciding the return rate is acceptable. A four-tech shop losing 45% of a tech to return trips is operating at roughly 89% of the field capacity it is paying for, and it will experience the shortfall as a hiring problem.
Why the multiplier differs, and when to rebuild it
Route density. A dense urban route with 0.4 hours of round-trip drive produces a very different multiplier than a rural route at 1.6 hours. Rural shops should expect a materially higher number and should be correspondingly more willing to spend field time on prevention.
Job type. Long install returns dilute the fixed components and land closer to 2x. Short service returns concentrate them and can exceed 5x on a full board.
Season. Displacement is real in peak and zero in the slow stretch, which means the same return trip genuinely costs different amounts in March and in October. If your board is full nine months of the year, use the full-board number as your default and note the exception rather than the reverse.
Cause. A return caused by a missing part carries an extra supply-house leg. A return caused by a workmanship error usually does not, but it carries the customer-standing cost more heavily.
Rebuild the multiplier annually, or whenever your service area, fleet, or recovery ratio moves noticeably.
How to verify your multiplier is not fiction
The failure mode here is a number built entirely from assumptions and then quoted for years as if it were measured. Three checks:
Time a real one, end to end. For two weeks, have the office start a clock at the inbound call and stop it after the close-out note, and have the tech log drive and on-site separately. Compare against your assumed numbers. Office touch is the component most often underestimated, because it is four people spending five minutes each.
Check drive against vehicle records. Assumed drive time is the single largest error source in the model, and dispatch timestamps or vehicle records will settle it in an afternoon.
Check whether the board was actually full. Pull the days on which your sampled return trips occurred and confirm there was demand you deferred or declined. If there was not, you counted displacement that did not happen, and your multiplier is overstated. An overstated multiplier is not a harmless bias toward prevention: it will justify prevention steps that cost more than they save.
References
- U.S. Small Business Administration (SBA), operational cost and productivity guidance for small business
- Trade-standard practice for first-time-fix and truck-stock measurement
- See related: The Pre-Job Checklist That Prevents the Return Trip, How to Cost Callbacks and Rework Honestly, How to Tell If Your Labor Rate Is Covering Its Cost, The Real Cost of an Inefficient Schedule