The practice that outgrew its timesheet

Time-based billing has an internal logic that held for a long time: the work took hours, the hours were the cost, the fee covered the hours plus a margin.

Automation breaks the logic in a specific way. The work that took the hours is the work that automates. A set of accounts that consumed twelve hours now takes three, and under time billing the fee falls by three quarters.

The practice has invested in software, improved turnaround, and reduced its own revenue. That is not a pricing inconvenience; it is a structural disincentive to improve.

The three responses

Bill the same hours anyway. Common, understandable, and untenable. It requires either recording hours that were not worked or maintaining a fiction about effort. It also fails the moment a client asks why their bill is unchanged when your marketing says the work is faster.

Bill the lower hours. Honest and unsustainable. Revenue falls in proportion to efficiency, so the practice funds an investment that reduces its own income.

Change the basis. The only stable answer, and the one that requires actual work.

What the fee is actually for

Worth being clear before repricing, because the answer determines the basis.

A client is not buying hours. They are buying:

• Accounts that are right, and the ability to rely on them
• Someone accountable if they are not
• Judgement on the things that are not obvious
• Not having to think about it

None of those scale with time taken. All of them are as valuable at three hours as at twelve — arguably more so, since the accounts arrive sooner.

That is the argument for fixed pricing, and it is not a trick. The hours were always a proxy for value, and it was a reasonable proxy while effort and value moved together. Automation broke the correlation, which means the proxy has to be replaced rather than defended.

Moving to fixed fees without losing money

The concern is pricing wrong and being locked into it. Four practical protections:

Price the scope, not the client. Define what is included: transaction volume band, number of bank accounts, which returns, how many meetings. Outside the scope is a separate fee. Most fixed-fee losses come from undefined scope rather than underpricing.

Set a volume band. "Up to 300 transactions a month" makes growth a repricing trigger rather than a silent margin erosion.

Include a review point. Annual, with the right to reprice. Clients accept this readily; practices frequently forget to put it in.

Price the difficult clients differently. Automation makes it visible which clients consume disproportionate effort. That information should reach the fee — see when a client refuses automation.

What to do with the time recording

A common overcorrection is abandoning time recording entirely on moving to fixed fees.

Keep recording, and stop billing on it. The data is how you know whether a fixed fee is profitable, which clients are consuming effort, and how much capacity a change actually released. Without it, fixed pricing becomes guesswork and you will not discover a loss-making client until the year end.

The change is what time recording is for: management information rather than a billing mechanism.

The conversation with existing clients

Moving established clients from hourly to fixed is easier than partners expect, because most clients prefer it. A predictable fee is worth something to a business owner, and an unpredictable one is a source of anxiety.

The framing that works: "You'll know what it costs before the year starts, and it won't change because something took longer than expected."

The one that does not: leading with the automation. It invites the response that if the work is faster the fee should be lower — a reasonable inference from a time-based mental model, and difficult to argue against once raised.

The measure

If a practice has automated and revenue per client is falling, the pricing model is the problem rather than the automation.

If revenue per client is stable and time per client has fallen, the transition has worked and the capacity is available for either more clients or advisory work.

Common questions

Why does automation break time-based billing?

Because the work that consumed the hours is precisely the work that automates. A set of accounts taking twelve hours may now take three, so a time-based fee falls by three quarters — meaning the practice has invested in software, improved turnaround and reduced its own revenue. That is a structural disincentive to improve rather than a pricing inconvenience.

How should a practice price once processing is automated?

On scope rather than time, since clients are buying accounts that are right, someone accountable for them, judgement on what is not obvious, and not having to think about it — none of which scale with hours taken. Fixed fees should define what is included, set a transaction volume band so growth triggers repricing, and include an annual review point.

Should we stop recording time under fixed fees?

No — stop billing on it and keep recording it. The data tells you whether a fixed fee is profitable, which clients consume disproportionate effort and how much capacity a change actually released. Without it, fixed pricing becomes guesswork and a loss-making client is not discovered until year end.

How do you move existing clients from hourly to fixed fees?

Lead with predictability, since most clients prefer knowing the cost in advance to an unpredictable bill. Avoid leading with the automation, because it invites the reasonable inference that faster work should cost less — an argument that follows naturally from a time-based mental model and is hard to counter once raised.

Related: the shift from compliance to advisory · taking on more clients without more staff · setting prices with confidence