How CPA Firms Actually Reduce WIP Write-Offs Before the Bill Goes Out

When realization slips, partners blame rates and clients. Look at the calendar instead. Realization is billed dollars divided by standard dollars, and a write-off is the difference you handed back. The report that names the write-off runs weeks behind the decision that caused it: someone logged hours the client was not going to pay for, the hours sat on WIP, and the partner met the problem for the first time on a pre-bill with an invoice date on it.

Firms that hold realization keep a short gap between the day work happens and the day somebody rules on whether it bills.

The write-off is a lagging indicator

A write-off at billing time is the last event in a chain. The chain runs like this. You scoped the engagement loosely. A preparer put extra hours into a return that should have been simple. No one on the engagement looked at WIP until it had piled up. The partner, holding a bill higher than the number the client heard at the start, knocked it down rather than pick up the phone for an awkward call. You could have broken any link before the last one. The invoice records the loss.

That chain is why "bill more aggressively" fails as a directive. The partner writing down the bill is making a rational call in the moment, because the alternative is a fee dispute with a client the firm wants to keep. The place to intervene is upstream, in the hours that landed on WIP and should never have been logged there.

Where the hours leak

Four failure modes account for most of the leakage. All four are operational problems with operational fixes.

Unlogged scope creep. The client sends three amended documents and asks two "quick questions" that eat an hour each. Those hours go in under the same engagement code with no note, so at billing time the partner reads them as a slow preparer, and slow preparers get written down. Tag the same hours out-of-scope the week they happen and you have a change-order conversation instead of a write-off.

Junior time on senior-scoped work, and the reverse. A first-year spends twelve hours on a task a manager finishes in four. That is eight hours of WIP no partner will approve. The preparer did nothing wrong; the staffing was wrong on the day the work went out. You catch it by reading WIP composition before the bill, because the WIP total tells you nothing about who did the work.

Stale WIP that ages past defensibility. Time logged in February and billed in June is hard for the client to accept and hard for the partner to remember well enough to defend. Aged WIP writes off at a much higher rate than current WIP. The fix here is calendar work: bill on a cadence short enough that nobody has to reconstruct February in June.

The silent write-down at review. The expensive one, and the hardest to see. A partner opens the pre-bill, sees a number above what the client is expecting, and trims it. No note, no reason code. The team never learns what to do differently, so the same engagement bleeds the same way next year.

Treat the pre-bill as a control point

Change one thing and make it the step between "work is done" and "invoice is sent." At most firms the partner rubber-stamps the pre-bill during the last week of the month. That step is the last place anyone can stop a write-off instead of recording one.

In a working review you put three questions to each engagement, out loud, before anything goes to a client:

1. Does WIP match the scope we agreed to? If not, you either call the client about the overage or write it down with a reason code attached. A silent trim is not one of the options.
2. Is the WIP composition defensible? Right staff level, no rework, no prior-period time buried in the total.
3. Would we bill this if the client asked us to itemize it? Any line that fails that test gets handled this week, while somebody still remembers the work.

You will still write time off. The goal is that a named person decides each one on purpose and attaches a reason. A quarter of coded write-offs gives you something to fix. A quarter of silent trims gives you a worse realization number and no idea why.

Put the review on a weekly cadence

Volume is why pre-bills get rubber-stamped. Hand a partner eighty engagements in the last week of the month and they will approve eighty engagements. Move the review earlier and spread it across the month.

• Review WIP weekly, at the engagement level. A weekly look catches a bad line while the work is current and the preparer still remembers it.
• Set a WIP threshold per engagement type that triggers a look before it grows. When a "simple" return crosses the hours you budgeted for it, somebody reads the detail that week.
• Bill on a fixed cadence so WIP does not age past the point where you can defend it. Interim and progress billing on longer engagements keeps the client's expectations and your WIP in sync.
• Capture a reason code on every write-down. Four buckets is enough: scope, staffing, rework, client relationship. Without the codes you cannot separate a pricing problem from a staffing problem, and you will keep treating one as the other.

Practice management platforms already support this loop. In CCH Axcess Practice, time and expense feed WIP in real time, and partners generate and adjust pre-bills inside the billing and WIP-management workflow before any invoice is released. Whatever system you run, you need the same things out of it: WIP visible at the engagement level, visible early enough to act on, and write-downs stored as coded decisions instead of adjustments that disappear into a total.

Know your own number first

Benchmark before you start fixing. Realization moves with firm size and service mix. The annual Rosenberg MAP Survey and the biennial AICPA and CPA.com National MAP Survey both track it as a core practice-management metric, and partners use them to see where they sit against peers. The CPA Journal's "How Realization Negatively Impacts CPA Firms" covers how realization erodes profitability across a firm. Then pull your own realization by engagement type and by partner. Your number will not be uniform across the firm, and the outliers show you which process broke.

Write-offs are a visibility problem, and cadence is how you solve it. Look at WIP early, at the level where the work happens, and put a name and a reason on every write-down. Realization is the output of that habit.