The $622 Repair That Was Quietly Absorbing a $1,300 Sale

Two technicians, two visits to the same address, weeks apart, and one invoice that quietly tangled both of their paychecks together. That is the kind of commission problem that never shows up on a dashboard. It only shows up when a technician looks at their pay, knows they sold a real job, and cannot figure out why the number does not match what they remember doing.

At a multi-trade company running plumbing, HVAC, and electrical technicians under one roof, that is exactly what happened. One technician had gone out on a same-day repair call and billed $622.25 for the work. Weeks later, a different technician returned to the same customer, sold and completed a full install worth roughly $1,300, and closed the job. On paper, those should have been two separate, cleanly attributed transactions: one technician credited for a repair, a different technician credited for a sale. Instead, the invoicing ended up structured so that the earlier $622.25 repair got applied against the later $1,300 install, effectively netting the two jobs together and threatening to shortchange the technician who actually closed the bigger sale.

I can see where the confusion happened, one team member said while walking through it, because he only brought in the $6.80, the smaller remaining balance, because of how he set it up in his invoice. It wasn't the full estimate that he had. That is a precise, almost forensic description of how invoice-level commission tracking can quietly go wrong. The install technician had done nothing incorrect from a service standpoint. He sold the job, completed the job, and should have been credited for the full amount. But because the invoicing system applied an earlier repair credit against his sale before anyone caught it, his commission calculation was at risk of reflecting a fraction of what he actually earned.

Why This Kind of Mistake Is Easy to Miss and Hard to Generalize a Fix For

The reason this is worth writing about is not the specific dollar amounts. It is what the case reveals about how commission tends to be tracked at multi-trade home service companies by default. ServiceTitan's standard invoice splits and sold-by fields are built around the invoice as the unit of attribution, not the individual technician's actual work. That works fine when one technician handles a job start to finish on a single visit. It breaks down the moment two different technicians touch the same customer address on different visits, weeks apart, and the system has to decide how to net those transactions against each other.

The fix required moving away from invoice-level splits entirely and rebuilding commission logic around individual technician attribution using time-on-job splits instead. Revenue and commission logic will be changed to attribute commissions by individual employee and use time-on-job splits instead of relying on ServiceTitan invoice splits, was the core decision. In practice, that means every job gets evaluated by who actually spent time on it and how much, rather than by whatever the invoice's sold-by field happens to say, which can be wrong for reasons that have nothing to do with who did the work. A technician who does not clock time on a job, for instance, would receive zero percent of that job's time-based revenue attribution under the new logic, which sounds strict until you realize the alternative is a system that can be gamed or simply gotten wrong by an invoicing mistake nobody caught in time.

Getting the job corrected specifically meant modifying the invoice and estimate so the sold-by field and the split reflected reality: the install technician credited for the full sale, not a fraction of it netted against an unrelated repair from weeks earlier. That is a manual fix for one job. The real work was building the underlying logic so the same mistake could not quietly repeat itself on the next multi-visit job that came through.

It is worth pausing on why this particular failure mode is so easy for a business to miss entirely. Nobody was trying to shortchange anyone. The office staff processing invoices were following a reasonable, ordinary workflow: apply an existing credit to a new invoice so the customer is not double-charged for a small balance. That is correct behavior from a billing standpoint. The problem only appears once you ask a second question the billing workflow was never designed to answer, which technician actually gets commission credit when two separate jobs, done by two separate people, get netted together on paper. Billing accuracy and commission accuracy are not the same problem, and a system built only for the first one will quietly get the second one wrong on exactly this kind of multi-visit job.

That gap is precisely why a multi-trade home service business running plumbing, HVAC, and electrical technicians through shared customer accounts needs commission logic that is built and reviewed separately from the invoicing workflow, even when both draw from the same underlying ServiceTitan data. The two systems are solving different problems, and treating them as interchangeable is exactly how a technician ends up quietly credited for a fraction of a sale they closed in full.

Excavation Work Had the Same Problem, Just Bigger

The invoice attribution issue was not limited to one repair-then-install sequence. The same root cause, a price book that did not have a clean, specific line item for a type of work, showed up again with excavation jobs. Excavation had been getting logged under a generic special item placeholder rather than its own dedicated code, which created two separate problems. First, it made it hard to tell whether a given excavation job had been done in-house or by a subcontractor, a distinction that changes the commission rate significantly. Second, it opened the door to a job getting mapped to the wrong category entirely, the way one dig job initially landed under a general plumbing special item before getting corrected to its proper excavation code.

The resolution was to add a proper excavation price book item and map it explicitly to the plumbing plan, with commission set at nine percent, distinguishing clearly between excavation completed in-house versus subcontracted out. Once that distinction exists as a real, selectable option in the price book instead of a generic catch-all, the same kind of accidental misattribution that happened with the repair-and-install job becomes far less likely, because the data feeding commission calculations is specific enough to actually mean something.

Callback Jobs Needed Their Own Rule, Fault-Based and Fair

A third piece of the same puzzle involved callbacks: situations where a customer needs a follow-up visit because something from an earlier job did not hold. Callbacks were not standardized across the team. Many showed up as zero-dollar jobs with inconsistent tracking, and there was no clear method to determine whether the original technician should be disqualified from commission when a callback traced back to their own mistake.

The rule that emerged treats the question of fault as the deciding factor, which is the fairest way to handle it. If the same technician who did the original work goes back out to fix their own callback, they do not receive the additional non-fault recall pay, since the visit is effectively a self-correction on their own job. If a different technician gets sent out to handle someone else's callback, that technician receives a flat non-fault recall rate for the hours spent, currently set at twenty-five dollars an hour, since they are doing unplanned work to fix a problem that was not theirs to begin with. That distinction matters for the same reason the invoice attribution fix mattered: technicians should be paid based on what they actually did and whose mistake, if anyone's, created the extra work, not based on whatever a generic job type field happened to record.

One Rate Table, Applied Consistently Across Three Trades

Once the individual attribution logic, the excavation mapping, and the callback rule were all in place, the company ended up with a single, coherent commission structure spanning plumbing, HVAC, and electrical work under one roof. Plumbing runs at nine percent for sold and completed work. HVAC runs at ten percent for both sold and completed. Installers earn a three percent add-on tied to their hourly role. Excavation, once properly separated from the generic special item catch-all, runs at nine percent in-house, distinct from the subcontracted rate. Every one of those rates gets applied per technician based on actual time-on-job splits, which means a job with multiple technicians involved gets divided fairly by who actually did the work, rather than getting assigned wholesale to whoever's name happened to land on the invoice.

That structure is what let a multi-trade field service company stop relying on a system that occasionally, invisibly, credited one technician's work to someone else's paycheck. The specific $622 repair and $1,300 install that started this whole review got resolved with a manual correction. The lasting fix was making sure the next repair-then-install sequence, the next excavation job, and the next callback all get attributed correctly the first time, without requiring anyone to notice a discrepancy and dig through invoice history to find it. For a company running three trades and a rotating cast of technicians across shared customer addresses, that kind of consistent, individually attributed commission structure is not a nice-to-have. It is the only way to guarantee that every technician's paycheck actually reflects the job they did.

Rollout Came With Training, Not Just a New Formula

None of this logic matters if the technicians on the ground do not understand it. Part of the rollout plan included a live walkthrough for the field team at go-live, explaining the metrics, the formulas behind them, and how to check their own numbers inside the platform, backed by written documentation technicians could refer back to later. That step matters more than it might seem. A commission structure that is technically correct but never explained clearly to the people earning it will still generate confusion and mistrust, even if every calculation behind it is accurate. Technicians who understand exactly how their time-on-job split gets calculated, and why an excavation job pays differently depending on whether it was subcontracted, are far less likely to assume a paycheck is wrong the moment it looks unfamiliar.

The rollout was timed to start with the next full pay period, giving the office time to confirm the price book updates and communicate the new performance pay start date to the whole team before the first live cycle. Getting that timing right, launching cleanly at the start of a pay period rather than mid-cycle, avoided the kind of partial-period confusion that can make an already unfamiliar system feel even more chaotic in its first weeks.