The Flat-Rate Tech Who Made Less in a Busy Month
Why this matters
On flat rate, a full board is supposed to mean a fat check. When a tech runs more tickets than ever and gets paid less, the first reaction is that somebody is cheating, and the second is to assume the busy month was an illusion. Usually neither is true. Flag pay is the product of several separate rules, any one of which can move money out of the month you earned it, and only a few of which actually take it away. This case follows one tech through four suspects, using the tech's own log and the shop's pay records, to the explanation that held up, and to the real problems hiding behind it.
The flag-hour arithmetic itself (efficiency, the minimum-wage floor, piece-rate overtime) is covered in How Flat-Rate Pay Works From the Tech's Side, and is only used here, not re-derived.
The signal
Dana, a service tech three years into a flat-rate plan, is paid a flag rate F for each flag hour (the book time a job pays, regardless of how long it takes). The shop pays weekly. Dana compared the four pay weeks of June with the four pay weeks of July, using the flag-pay line on each stub. The overtime premium lines, figured weekly off each week's flag earnings, are set aside below to keep the flag pay visible; they move with the flags.
- June: 52 tickets on Dana's board, 196 clock hours, 172 flag hours paid. Flag pay 172 F.
- July: 68 tickets on the board, 224 clock hours, 158 flag hours paid. Flag pay 158 F.
Sixteen more tickets and 28 more clock hours, for 14 F less, about 8 percent below June (14 divided by 172). Dana keeps a daily log of start and stop times, tickets and flag hours, which is the only reason any of what follows could be tested. See related: Tracking Your Own Hours and Pay.
Dana walked in with four suspects: recalls, the long drives, comebacks, and something about when tickets get counted.
First suspect: recalls
July had three recalls on Dana's own work, each a return trip that flags zero under the shop's plan. Together they took 5.0 clock hours.
Why it does not explain the drop. A recall costs time, not paid flag hours on other jobs. At June's paid pace of 0.88 flag hours per clock hour (172 paid on 196 clock), 5.0 clock hours is worth about 4.4 flag hours of work Dana could otherwise have done. That is real, but it cannot produce a 14 F drop in a month with more tickets on the board, and it is normal under almost any flat-rate plan. Ruled out as the cause; kept as a small drag on efficiency.
Second suspect: the long drives
In July the dispatcher moved Dana into the outer service zone after another tech left. Dana's log shows drive time between jobs rising from 22 hours in June to 34 hours in July, 12 more clock hours that flag nothing.
This one looked promising, and it was the dispatcher's own first guess. Drive time is clock time, it counts for the minimum-wage floor, and it lowers the share of each hour that pays. But it lowers what Dana can produce; it cannot lower pay on work Dana did produce. That suggested the test that split the whole problem.
The test that split the problem in two
Dana added up the flag hours booked on July's tickets, from the log, ticket by ticket: 201 flag hours on 63 flagged tickets. The other 5 of the 68 were zero-flag return trips (the three recalls, plus two more covered below).
Dana produced 201 flag hours of work and was paid for 158. Producing more than June's 172 means the drives and recalls did not cause the drop; whatever happened, happened between the work and the paycheck. The gap is 201 minus 158, or 43 flag hours, and every remaining suspect has to be measured against that number.
Ruled out as the cause of the drop: the drives. They still cost something, which comes back at the end.
Third suspect: other people's comebacks
Two of July's zero-flag tickets were not Dana's work. They were comebacks on installs another tech had done before leaving, assigned to Dana because Dana now covered that zone, and coded "recall" by dispatch. Together they took 4.5 clock hours.
Why it does not explain the gap. Those tickets were zero in Dana's own log too, so they are not part of the 201, and they cannot account for any of the 43. They are a separate question: a plan that zeroes every return trip regardless of whose work caused it makes the tech who inherits a territory pay for the tech who left it. Not the cause; flagged as a policy problem to raise. See related: How Callbacks and Warranty Work Affect Your Pay.
Fourth suspect: the cutoff
The office manager pulled the flag report and the answer was in the dates. The shop credits flag hours in the pay week the ticket is closed (invoiced and signed off by the office), not the week the work was done. Dana's last July week was the busiest of the year, the office fell behind on invoicing, and a stack of tickets crossed the cutoff open.
OPEN AT CUTOFF SAT 7/26 23:59 TECH 14
TICKET WORKED STATUS FLAG CLOSED
5121 7/20 invoice draft 3.5 7/28
5126 7/21 parts on order 2.5 8/04
5130 7/22 invoice draft 6.0 7/28
5133 7/23 no signature 4.0 7/29
(9 more, worked 7/21 to 7/26) 31.0 7/28-8/05
TOTAL 13 tickets 47.0 -> Aug pay weeks
47 flag hours moved into August. That is more than the whole 43-hour gap, which looked like a contradiction until the June side was checked the same way.
Squaring the month
A pay week credits the tickets closed in it, so July's paid flags include June tickets that closed in early July, and the comparison has to be corrected on both sides, not just July's.
June: Dana's log shows 175 flag hours worked in June. Of those, 9 closed after June's cutoff and were paid in July. June's paid 172 was therefore 166 flag hours of June work plus 6 carried in from May. So June worked 175, June paid 172.
July: paid 158 is 9 carried in from June plus 149 of July's own work. July's 201 booked flag hours minus 47 carried out to August leaves 154, not 149. The last 5 flag hours turned up in one ticket: a compressor job Dana had booked at 8.0 flag hours that the office repriced to 3.0 after the manufacturer accepted a warranty claim, because the plan flags warranty work at the manufacturer's labor allowance. 154 minus 5 is 149. The month closes.
On the same worked basis, June produced 175 flag hours and July produced 196 (201 less the 5-hour warranty repricing), 21 flag hours or 12 percent more. Efficiency still slipped a little, from 0.89 flag hours per clock hour (175 on 196) to 0.88 (196 on 224), which is where the drives and recalls show up. The busy month was real. The difference was that June pushed 9 flag hours forward and July pushed 47, because the office's invoicing backlog grew with the board.
The week the floor was missed
Moving flags between months is mostly a timing problem: the 47 hours arrived in Dana's August checks. But the flag report also showed how July's 158 split by week: 46, 52, 48 and 12 flag hours, against 54, 56, 58 and 56 clock hours. Week 4 is the busiest week of the year, and it credited 12 flag hours on 56 clock hours, because nearly everything worked in it closed after the cutoff.
Take, for illustration, a minimum wage of 0.40 F per hour where Dana works (the flag rate is two and a half times the minimum). Week 4's floor under the federal workweek test is 56 times 0.40, or 22.4 F, and the week credited 12 F, short by 10.4 F. Under the Fair Labor Standards Act the minimum wage is owed for each workweek (29 CFR 776.4 and 778.104 make the workweek the unit), on that week's regular payday; a catch-up weeks later does not obviously cure a payday that came in short, and states such as California, which do not allow averaging at all, are stricter still. The stub for week 4 showed no make-up line.
What Dana raised, and what came back
Dana went to the office manager with three specific items, in writing, and kept the cutoff itself off the list because it was a timing rule, not a loss. See related: Raising a Paycheck Error.
- Week 4's minimum-wage make-up, 10.4 F. The owner's accountant agreed and paid it on the next check. The shop also changed payroll to compare every week's credited pay against clock hours before running checks.
- The two comebacks on the departed tech's installs, 4.5 clock hours at zero. The owner agreed that a return on someone else's work was not a recall and flagged both at book time, 4.0 hours.
- How the deferred 47 hours were treated for overtime. Dana's July weeks ran 54 to 58 clock hours, so each carried an overtime premium figured on that week's flag earnings. Federal rules on pay that is figured after the week it was earned (29 CFR 778.119 for deferred commissions and 778.209 for bonuses) require it to be allocated back to the workweeks it was earned in when the regular rate is calculated. The accountant took that question away; it is the item an attorney or the state labor agency would settle if the shop had refused.
The warranty repricing stayed as it was, because the written plan said warranty work flags at the allowance. Dana now invoices on the truck before leaving each job, which cut the carried-over tickets to almost nothing, and checks one number every Friday: flag hours worked from the log against flag hours credited on the stub, with the open-ticket list explaining any difference.
References
- 29 CFR 776.4 and 778.104 (the workweek as the unit for minimum wage and overtime); 29 CFR 778.111 (piece-rate overtime)
- 29 CFR 778.119 and 778.209 (deferred commission and bonus pay allocated to the workweeks earned)
- See related: How Flat-Rate Pay Works From the Tech's Side; How Callbacks and Warranty Work Affect Your Pay; Tracking Your Own Hours and Pay; Raising a Paycheck Error