The Deduction Nobody Recorded
A firm counted its annual discounts, split them by who decided, and found that a fifth of the total had never been decided by anybody.
- Entry
- 05 of 07
- Section
- What comes off
- Kind
- Case
A quarter's deductions on one practice, split by who decided
Agreed in engagement letters; not a write-down at all, though it was counted as one.
Concentrated in nine matters out of ninety-one.
Recorded as courtesy; the word covered both this row and the one above.
Consistent, defensible, and the client was never told it happened.
Second attendees, wrong matter codes, drafts left unapproved.
Of £203,974 the firm counted as discounts, £41,200 was its actual price and £38,934 was nobody's decision at all. Only the middle two rows were a judgement anybody made. One firm's own quarter, not a benchmark.
A professional firm counted a quarter's reductions and found £203,974 against £1.9m of recorded value. It had always called this figure its discount rate and treated it as a measure of generosity.
The record described in “The Deduction Nobody Recorded” should be created while the work is still clear rather than reconstructed at the end of the week. Used for 7 minute rule payroll, explore this Monitask resource can supply project-linked time context, provided people can review and correct entries before those records support billing or management decisions.
Split by who decided, a fifth of it turned out to be hours that had simply never reached a bill, and a further fifth was the firm's agreed prices being counted as concessions.
Teams documenting “The Deduction Nobody Recorded” may find Fair Work record-keeping guidance useful as an external benchmark. Record which parts apply to the real workflow and which do not.
Why the exercise was done
A new finance director asked what the firm gave away in a year. Nobody could answer beyond a percentage, and the percentage came from one field in the billing system called "write-off".
That field had no reason code and had never been split.
The first finding
Negotiated rate concessions — £41,200 — were in the figure. These are the firm's prices, agreed in engagement letters, and had been counted as discounts every year.
Removing them cut the headline figure by a fifth before anything else was examined.
The second finding
The overrun absorbed at the draft, £77,220, was concentrated in nine matters out of ninety-one. Seven of the nine were the same type of work.
That is an estimating problem with a name and a size, and it had been invisible inside an aggregate.
The third finding
Rework, £25,380, had been recorded under the same heading as overrun because the only word available was courtesy.
Rework is a supervision question and overrun is an estimating one. The firm had been averaging two unrelated problems.
The fourth finding
Hours that never reached a draft: £38,934 of recorded time, from second attendees on different matter codes, work before matters were opened, and timesheets left unapproved.
Nobody had decided any of it. It was the second largest line in the table.
How that was measured
Hours recorded against each matter in the time system, compared with hours appearing on drafts for the same period.
One query, run once. The firm had the data throughout.
What was done about each
Rate concessions were moved out of the discount figure entirely and reported as pricing. Overrun went to the practice group that generated it. Rework went to supervision. The unapproved time went to a weekly chase.
Four owners instead of one aggregate.
The result after two quarters
The nobody column fell by about two-thirds, which was recovered revenue requiring no price change and no conversation with any client.
The overrun column did not move, because estimating takes longer than two quarters to change.
What did not happen
Nobody's discretion was removed. Partners still cut bills at the draft; they now attach one of seven reasons to each cut.
The resistance anticipated before the change did not materialise once it was clear nothing was being taken away.
The reporting that followed
A quarterly table, five rows, hours and value, by practice group. Produced by one person in an hour.
It replaced a single percentage that had been reported for years and had never caused anybody to do anything.
What the case shows
That the word covering a reduction determines whether anybody can act on it, and that the largest category in a firm's write-downs may be one nobody chose.
What the exercise cost
Two days: one to build the comparison between recorded and drafted hours, one to design and apply seven reason codes.
No new system, no change to anybody's recording, and no new report beyond one quarterly table.
The objection raised beforehand
That coding reductions would be read as challenging partners' judgement, and that people would resist being asked why.
The framing that carried it was that the code attaches to the deduction rather than to the person, and that the report would never be broken down by individual.
Both commitments were kept, which is why the data stayed honest.
What would have broken it
Reporting write-downs per partner. The firm considered it and did not do it, on the reasoning set out in the previous notes.
A partner measured on write-downs reduces fewer bills or recodes everything as client-agreed within two quarters, and the data stops being worth collecting.
The figure the firm had been reporting
"Write-offs: 10.7 per cent." One number, in the management pack for four years, which had never caused any action.
It was not wrong. It was an average of five unrelated things, and averages of unrelated things are not actionable by anybody.
What replaced it
Five rows, quarterly, by practice group, with the owner of each row named.
The headline percentage was retained for continuity and nobody has looked at it since.
The checkable claim
Any firm can run the same comparison — recorded hours against drafted hours, one quarter, by matter — in an afternoon.
The answer is specific to that firm, and in most cases nobody has ever looked.