Accounting firm staff productivity: what AP benchmarking data implies for a multi-client practice

Accounts-payable shared-services benchmarking is usually framed around large internal finance departments, not accounting practices serving external clients. The underlying metric — how many documents a person can process accurately per year, and what separates the best-performing operations from the average — applies just as directly to a multi-client practice, where the same document-processing work happens across many client files instead of one company's own books.

What the benchmarking data actually shows

Shared-services research from IOFM puts average accounts-payable throughput at roughly 4,200 invoices per FTE per year, against 6,900 for best-in-class operations — a 64% gap driven largely by how much of each employee's time goes to manual keying and template maintenance versus reviewing genuine exceptions. Ardent Partners' AP Metrics research puts the cost gap in similar terms: $12.88-19.83 per invoice for manual processing, versus roughly $2.78 for best-in-class automated operations.

Why this gap applies directly to a multi-client practice

A practice's staff time is subject to the exact same split the benchmarking data describes: time spent on mechanical document processing (keying, checking arithmetic, coding to the right account) versus time spent on the judgment work only a qualified professional can do (advisory conversations, tax positions, reviewing genuinely ambiguous items). The 64% throughput gap and the roughly sevenfold cost gap both trace back to the same root cause the benchmarking data identifies: how much of the mechanical work still requires a person's direct attention.

What closing the gap looks like inside a practice

Closing that gap does not require working staff harder on the same tasks — it requires removing the mechanical share of the work from staff time entirely, so the hours a practice already has go further per client file. A practice sitting closer to the average end of the benchmarking range is not necessarily understaffed relative to its client count; it may simply be spending a larger share of each staff-hour on work that does not require a qualified professional's judgment at all.

What to actually measure inside a practice

• Client files processed per staff-hour, tracked over time — a number that should move as process changes are made, giving a concrete before-and-after rather than a qualitative sense that things feel faster.
• Share of staff time on mechanical processing versus advisory work — even a rough estimate, tracked consistently, shows whether capacity gains are actually being redirected toward the work that drives client value and retention.
• Exception rate — what share of documents require genuine human judgment versus what share are routine and could, in principle, clear without a person touching them at all.

Where practices typically lose the most time, and why it is rarely obvious

Asked to identify their biggest time drain, most practices point to whichever task feels most tedious in the moment — often data entry itself. The less visible but often larger drain is the correction and rework cycle: a document keyed once, found to have an error during a later review pass, corrected, and re-checked, sometimes more than once before it is considered final. Each pass through that cycle consumes staff time that the original benchmarking numbers, framed around a single "processing" step, do not fully capture — the real comparison is not manual keying versus automated extraction alone, it is the full cycle including rework, versus a process where the first pass is accurate enough that a second and third pass are rarely needed.

This is also where the benchmarking gap between average and best-in-class operations tends to concentrate: an operation closer to average is not necessarily slower at the initial keying step, it is more likely to be re-doing work because the initial pass missed something a later reviewer caught. Reducing rework, not just speeding up the first pass, is often the larger lever.

A simple diagnostic a practice can run without new tooling

Over a two-week period, ask staff to flag any document that required a second look or a correction after initial processing, and roughly why. A pattern in the "why" column — the same type of error recurring across different client files, the same category of document consistently needing a second pass — points directly at where a process change would have the most leverage, rather than guessing which part of the workflow to improve first based on impression alone.

Related reading

• Accounting firm client retention: it is a capacity problem before it is a pricing problem
• Accounting firm compliance deadlines 2026-2027: the DGFiP and HMRC dates that set the calendar

FAQ

Do these AP benchmarking numbers translate exactly to accounting-practice work?

Not exactly — a practice's work mix includes advisory and compliance tasks beyond pure AP processing that the benchmarking data does not cover. The directional lesson transfers cleanly even where the precise numbers do not: the gap between average and best-in-class throughput is driven by how much mechanical work still needs a person, and that mechanism applies to practice work generally, not just AP specifically.

Is a 64% throughput improvement realistic for a typical practice?

The benchmarking gap describes the spread between average and best-in-class operations broadly, not a guaranteed outcome for any specific practice — the achievable improvement depends on how much of a given practice's current work is genuinely mechanical versus already judgment-heavy, which varies by client mix.

How does a practice track this without expensive new reporting infrastructure?

Simple, consistent metrics — files processed per staff-hour, exception rate — tracked in a shared spreadsheet updated monthly are enough to show a trend; the value is in consistency over time, not in the sophistication of the measurement tool.