Define the measure first, then leave the definition alone, because there's no independent published benchmark for lock-up or debtor days in UK accountancy practices to compare yourself against. Once your own series is consistent from one month to the next, the changes that tend to shorten it are billing sooner, agreeing how the fee will be collected at engagement, and separating the few clients who generate most of your delay from the many who don't.
In this article
The short version
- There's no independent, current benchmark for lock-up or debtor days in UK accountancy practices, so any figure offered to you as the practice average is worth tracing to its source.
- The nearest well-documented professional-services figure is for law firms: the Law Society's 2026 Financial Benchmarking Survey, written by Hazlewoods and sponsored by Lloyds Commercial Banking, reported median total lock-up of 134 days excluding unbilled disbursements, or 144 days including them, from 121 firms' 2025 figures.
- Ten days of difference from one drafting choice, inside a single survey, is the reason a benchmark computed one way can't be read against a benchmark computed another.
- The accountancy figure in circulation, lock-up at 32% of fees, traces through a 2010 trade article to a survey run in 1982.
- Two conventions exist: lock-up in days, and lock-up as a percentage of annual fees. Neither converts into the other.
- Where a direct debit mandate already existed when the invoice was raised, 96.98% of invoices were paid and 1.31% were still open six or more months later, against 90.25% and 5.36% with no mandate (Adfin platform data, 152,689 invoices created in the seven months to 31 January 2026).
Why there's no figure to compare yourself against
Go looking for the average lock-up in a practice like yours and you'll find what everyone finds. Nothing current and accountancy-specific, from a source with no product to sell, is published anywhere we could read. What circulates instead is borrowed from the legal sector, or comes from vendor surveys of a few hundred self-selected respondents, or was collected before you had a spreadsheet to put it in.
The oldest of those tends to be quoted with the most confidence. Lock-up at 32% of fees, with a recommended target of no more than that, comes from an AccountingWEB article published in February 2010, and the survey behind the 32% was run in 1982. Both dates belong with the number every time you meet it.
The best-documented professional-services figure in the UK covers solicitors. The Law Society's 2026 Financial Benchmarking Survey, written and produced by the legal team at Hazlewoods LLP and sponsored by Lloyds Commercial Banking, covers 121 law firms in England and Wales with combined fee income over £1.2 billion, an average of £9.9 million each, on 2025 data. It reports total year-end lock-up days, work in progress and debtors combined, two ways in the same edition: 134 days excluding unbilled disbursements, down from 146 in 2024, and 144 days including them, down from 156.
Those are solicitors, your sample is 121 self-selected firms, and the survey is written by a firm of accountants selling to the legal sector and sponsored by a bank that lends to it. So it's unlikely to work as your benchmark. What you can take from it is that one drafting decision moved the same firms' lock-up by ten days. Most published figures don't say which decision they made.
Define it once, and show the arithmetic
You'll see it measured two different ways, and there's no conversion between them. The legal sector puts lock-up in days. Accountancy trade content usually gives it as a percentage of annual fees. The arithmetic normally runs like this:
| Convention | Arithmetic | Reads as |
|---|---|---|
| Lock-up days | (work in progress + trade debtors) ÷ annual fee income × 365 | The number of days of fee income you're carrying |
| Lock-up as a percentage of fees | (work in progress + trade debtors) ÷ annual fee income | The share of a year's fees tied up at any moment |
| Debtor days alone | trade debtors ÷ annual fee income × 365 | The collection half only, with billing excluded |
No professional body publishes either formula for accountancy practices as far as we could establish, so treat both as the common convention with the arithmetic on display and not as a standard anybody set.
Your consistency comes from the choices underneath the formula. We're suggesting five of them below, and no professional body prescribes any of these. Write down whether unbilled disbursements are in or out, since the Law Society's two figures show you what that single decision does. Then write down whether your debtors are gross or net of VAT, whether a credit note raised after the month end goes back into the month it relates to, whether you value work in progress at charge-out rates or at cost, and whether your denominator is the last twelve months of fees or an annualised current month. Keep those five answers still for a year and your own figures will compare with each other, even though they still won't compare with anybody else's.
Split the number into the two halves somebody owns
There are two separate delays inside lock-up, and different things drive them. Work in progress is time recorded and not yet billed, and it moves when your billing routine moves. Trade debtors is fees billed and not yet collected, and it moves when your collection arrangement moves. The Law Society's write-up doesn't report the two separately in what we read, so a single combined figure won't tell you which of them is growing.
Report them separately and each one has somebody accountable for it. Somebody in your practice decides when a job gets billed, and somebody decides what happens between the fee note and the money. Where that's the same person, the billing usually slips first, since an unraised fee note is invisible to everyone while an unpaid one sits on a report.
The debtors side is the half you can see on a report already, so if you want to read yours column by column, how to read an aged debtor report goes through it, and the credit control metrics worth tracking sets debtor days alongside the other measures.
Measuring it monthly
Measure monthly, not annually. An annual figure smooths over the month that went wrong, and that month is where your explanation lives. Four habits keep a monthly series honest.
- Take your reading on the same day of the month, from the same report.
- Record your definition alongside the number in the same file, because a definition kept in somebody's head changes without anybody noticing.
- Keep your fee volume next to it. A quiet quarter can improve lock-up for reasons that have nothing to do with billing or collection, and a busy one can worsen it while everything you changed is working.
- Note anything one-off in the same row. One large fee note raised on the last day of your month can move a small practice's figure further than a quarter of genuine improvement.
If most of what you collect arrives by direct debit, treat any days-late measure carefully. Bank debit settles a few days after the due date through the Bacs cycle by design, so a days-late figure blended across direct debit and card or bank transfer is partly measuring the settlement cycle. Your on-time rate, or the share of your fee notes still unpaid after 60 days, avoids that.
What tends to shorten the wait
A small practice has four changes available to it, and the first two are billing decisions.
Bill on a routine your client already knows about. Work in progress that ages is usually work nobody has decided to bill yet, and a fixed billing date per client takes the decision away. The joint engagement letter guidance from ATT, CIOT, ACCA and AAT, updated 3 March 2026, names monthly fixed fees among the structures open to you and asks your covering letter to say when fees become due. Pricing and fee recovery for accountancy firms covers what each structure does to your recovery.
Check the due dates you're issuing. Across 59,777 paid customer-initiated invoices on Adfin, 3,271 of them, 5.5%, carried a due date that had already passed when the invoice was raised, and their median delay ran to 42 days (Adfin platform data). Some of that will be catch-up billing entered after the fact, so read it as invoices issued with a due date already gone. Checking your own due dates for the same thing takes a minute.
Agree the collection method while your client is agreeing everything else. Where a direct debit mandate already existed when the invoice was created, 96.98% of invoices were paid and 1.31% were still open six or more months later. With no mandate, 90.25% were paid and 5.36% were still open (Adfin platform data, 152,689 invoices created in the seven months to 31 January 2026). Mandated invoices in that cohort were smaller, a £126 median against £180, and went to clients who'd already agreed to a mandate, so some of that reliability was there before the mandate was. Where payment setup belongs in client onboarding covers your sequencing, and making direct debit part of signing the engagement letter the mechanics at signature.
Separate the persistent from the occasional. Among 1,359 client relationships with at least five customer-initiated payments each, 27.4% were never late once, 10.3% were late on every payment, and the quarter late 80% of the time or more accounted for 51.9% of all late payments (Adfin platform data, 12,863 payments). Clients reaching five customer-initiated payments are a selected group, since a client paying reliably by card five times running is the one a practice would have moved onto a mandate, so read the shape and not the levels. Your own delay probably has a short list of names on it, and how to set different chasing rules for different clients covers treating that list differently.
What one firm's numbers looked like
Clarity Counts, an accountancy practice using Adfin, brought its average days paid late down from 40 to 4 within seven months (Adfin platform data, one firm). Three things travel with that figure, and you'd want all three before you set a target from it.
One firm's result is not a platform benchmark, so nothing here says your practice would see the same movement. The 40 and the 4 are average days-late figures, the class of measure we don't publish across the platform for the Bacs reason above. And if that firm's book moved onto direct debit over the same seven months, the improvement and the settlement cycle are entangled in one number, because a mandate both collects the fee and changes when collection lands.
The external comparator belongs with it. Xero Small Business Insights put UK small businesses at 7.3 days late on average for April to June 2024, on a different population with a different definition, so treat it as context and not as the same measure.
What carrying it costs you
No source we could find quantifies what lock-up costs a UK accountancy practice, whether in interest paid, facility use or the effect on partner drawings. So arithmetic you can check yourself is worth more to you than any published figure.
Bank Rate was 3.75% on 19 August 2026. Carry £100,000 of lock-up on an overdraft priced at Bank Rate plus a margin and you're paying that margin plus 3.75% a year for the privilege, which is a number you can put in front of your partners with your own facility rate in it. Anyone quoting you a typical practice overdraft rate or a typical cost of lock-up is estimating.
For most small practices the interest isn't the expensive part. The drawing you didn't take costs more, along with the recruitment that waited a quarter and the fee note nobody wanted to chase turning into the one nobody expects to collect. Bad debt and write-offs: when to stop chasing covers the end of that road, and can accountants charge clients interest on unpaid fees what you can add to the fee while it's still open.
Common questions
What is a good lock-up figure for an accountancy practice? No independent, current benchmark for UK practices exists that we could find, so you have nothing defensible to compare yourself against. The nearest well-documented professional-services figure is the Law Society's, for law firms: median total lock-up of 134 days excluding unbilled disbursements or 144 including them, from 121 firms' 2025 data. Your own consistent series is more use to you than any borrowed average.
How do you calculate lock-up? Commonly as (work in progress + trade debtors) ÷ annual fee income × 365 for a figure in days, or the same fraction without the 365 for lock-up as a percentage of annual fees. No professional body publishes either formula for accountancy practices, so state which one you're using and whether unbilled disbursements are in or out.
What's the difference between lock-up and debtor days? Debtor days covers fees you've billed and not yet collected. Lock-up adds your unbilled work in progress to it, so it captures the wait that starts when the work is done instead of when your fee note goes out. A practice with tidy collection and slow billing can show good debtor days and poor lock-up.
Where does the 32% of fees figure come from? From an AccountingWEB article published in February 2010, reporting a survey run in 1982. Use it as an illustration of how old the circulating figures are, with both dates attached, and not as a target for your own practice in 2026.
How often is lock-up worth measuring? Monthly, on the same day from the same report, with your fee volume for the period beside it. An annual figure hides the month that went wrong, and that month is where your explanation lives.
Does moving clients to direct debit reduce debtor days? It changes what the number is measuring, so compare carefully. Bank debit settles a few days after the due date through the Bacs cycle by design, so a mandated book can look marginally late on a days measure while resolving more reliably. On Adfin, invoices raised with a mandate already in place were paid 96.98% of the time against 90.25% without one, on 152,689 invoices created in the seven months to 31 January 2026.
Sources
- Bank of England — official Bank Rate history (accurate as of August 2026)
- ATT — engagement letters guidance (accurate as of August 2026)
- Law Society — Financial Benchmarking Survey 2026 (accurate as of August 2026)
- Xero — UK Xero Small Business Insights, April to June 2024 (accurate as of August 2026)
- AccountingWEB — managing lock-up, part one (accurate as of August 2026)
This article is information about practice measurement and getting paid, and it isn't accounting, tax or legal advice. The lock-up conventions here are the ones we found in published use, not a standard set by any professional body, and no independent benchmark for UK accountancy practices was available to us. A practice making a decision about its own reporting, its client engagements or its professional obligations should take it up with its professional body or its own adviser. Bank Rate is as at 19 August 2026 and changes with the MPC. Last updated August 2026.
