Case Study: UK Practice Manages 200+ Clients

Growth in accountancy is typically linear. More clients means more staff. More staff means more office space, more management overhead, and more risk. The conventional wisdom is that a sole practitioner can manage around fifty to seventy clients comfortably, and a small team of three to four can handle perhaps a hundred and fifty to two hundred before the next hire becomes necessary.

This case study examines how a UK accountancy practice broke that pattern. Starting with approximately eighty clients and a team of three, they scaled to over two hundred clients within eighteen months — without hiring additional staff, without compromising review quality, and without the team working unsustainable hours.

The practice has asked us to present this as an anonymised case study. We have changed the name and certain identifying details, but the numbers, timelines, and outcomes are accurate.

The Practice: Background

The practice, which we will call Harrison & Co, is based in the Midlands. It was founded in 2019 by Sarah, a former senior manager at a mid-tier firm who wanted to build a practice focused on sole traders and small businesses. By early 2024, the practice had grown to approximately eighty clients, with Sarah as the principal, one qualified accountant (James), and one bookkeeping assistant (Lisa).

The client base was predominantly sole traders: tradespeople, freelancers, landlords, and small online businesses. Services included bookkeeping oversight, Self Assessment returns, VAT returns, and basic advisory work. Fees ranged from £75 to £200 per month depending on complexity.

The practice used Xero for most clients, with a handful on QuickBooks. Practice management was handled through AccountancyManager. The workflow was conventional: clients maintained their own Xero records (with varying degrees of success), the team reviewed records periodically, and returns were prepared and filed on the clients' behalf.

The Challenge

By mid-2024, Harrison & Co was at capacity. The team was working full weeks with little buffer, and Sarah had stopped accepting new clients because there was no capacity to serve them properly.

The specific bottlenecks were:

Transaction review. The quarterly review process for each client took two to four hours, depending on the client's record-keeping quality. With eighty clients and a mix of quarterly and annual reviews, the team was spending approximately sixty to seventy hours per week on review work alone.

Client chasing. A significant portion of the team's time was spent chasing clients for missing receipts, incomplete records, and responses to queries. Lisa estimated she spent fifteen to twenty hours per week on this alone.

Workpaper preparation. James prepared workpapers for every review, a process that added roughly thirty minutes per client. Across eighty clients over four quarterly cycles, this was substantial.

Deadline management. With MTD for Income Tax approaching, the practice was facing the prospect of four hundred quarterly submissions per year (for affected clients) on top of existing annual filings. The maths did not work without additional staff.

Sarah considered hiring a fourth team member but was reluctant. The additional salary, NI, pension, and overhead would add approximately £35,000-40,000 per year to practice costs. And finding a qualified, experienced person in the current market was proving difficult — she had been trying to recruit for six months without finding the right candidate.

The Decision to Try Accounted

Sarah first heard about Accounted at a local ACCA networking event in late 2024, where another practice owner mentioned the AI-assisted review workflow. The concept of exception-based review — reviewing only flagged items rather than every transaction — was immediately appealing.

She was initially sceptical. The idea that AI could accurately categorise the majority of transactions without human oversight seemed optimistic. But the free practice portal meant there was no financial risk in trying it. She set up a portal account and moved five clients across as a pilot.

The Pilot (January-March 2025)

The five pilot clients were chosen deliberately: three straightforward sole traders with predictable transaction patterns, one freelance consultant with more complex expenses, and one landlord with two rental properties.

The results were immediate and measurable:

• Average review time per client dropped from 2.5 hours to 40 minutes for the three straightforward clients
• The consultant's review dropped from 3.5 hours to 1.25 hours — more flagged items due to expense complexity, but still a significant reduction
• The landlord's review dropped from 3 hours to 55 minutes — rental income and mortgage payments were patterns Penny learned quickly

For the three months of the pilot, Sarah tracked time carefully. The five pilot clients consumed approximately 12 hours of review time per quarter, compared to an estimated 35 hours under the previous workflow. A 65% reduction.

James, who conducted most of the reviews, noted two additional qualitative observations. First, the review quality was higher because he was focused on genuinely ambiguous items rather than scanning through hundreds of routine transactions. Second, the automated workpapers were more comprehensive than the ones he had been preparing manually, and they took zero additional time.

The Migration (April-September 2025)

Based on the pilot results, Sarah decided to migrate the entire client base to Accounted. The migration happened in phases over six months.

Phase 1: Straightforward Clients (April-May 2025)

The first thirty clients to migrate were those with simple transaction patterns: sole traders with regular income and predictable expenses. These were the "quick wins" — clients where the AI would learn fastest and the time savings would be most immediate.

The onboarding process for each client took approximately fifteen minutes of the team's time (sending the invitation, doing the initial setup review) and ten to fifteen minutes of the client's time (connecting their bank account and WhatsApp). Lisa handled most of the onboarding, batch-processing five to eight clients per day.

Client reception was overwhelmingly positive. The WhatsApp-based interface was immediately intuitive for most clients. Several commented that it was "the easiest bookkeeping has ever been." One plumber told Sarah he used to dread the monthly Xero login and now did not even think about bookkeeping because Penny handled it through messages he answered in moments.

Phase 2: More Complex Clients (June-August 2025)

The second phase covered clients with more complex circumstances: VAT-registered businesses, clients with multiple income sources, and those with higher transaction volumes. These clients had higher initial flag rates (Penny needed more time to learn their patterns), but by the end of the second quarter on the platform, review times had converged with the straightforward clients.

The team also began migrating clients who had been on QuickBooks, which required a clean transition. Historical data was imported, and Penny's categorisation engine learned from both the imported history and the new transactions flowing in via bank feeds.

Phase 3: Remaining Clients and New Growth (September 2025 Onwards)

By September 2025, seventy-five of the original eighty clients had been migrated. Five clients with particularly complex circumstances remained on Xero (two construction companies with CIS complications and three clients with international transactions), though Sarah intended to move them once she was fully comfortable with those specific workflows.

Critically, with the review workload dramatically reduced, the practice started accepting new clients again. Sarah began actively marketing for the first time in eighteen months.

The Results: 18 Months Later

By February 2026, Harrison & Co's client base stood at 208 clients — a 160% increase from the pre-Accounted figure of 80. The team remained at three people: Sarah, James, and Lisa. Here are the detailed outcomes.

Time Allocation

The team's weekly time allocation shifted dramatically:

| Activity | Before (80 clients) | After (208 clients) | |---|---|---| | Transaction review | 60-70 hours/week | 35-40 hours/week | | Client chasing | 15-20 hours/week | 3-5 hours/week | | Workpaper preparation | 8-10 hours/week | 0 hours/week (automated) | | Advisory and complex work | 5-8 hours/week | 20-25 hours/week | | Business development | 2-3 hours/week | 8-10 hours/week | | Administration | 8-10 hours/week | 5-7 hours/week | | Total | ~110 hours/week | ~80 hours/week |

The practice went from 110 hours per week (unsustainable for three people) managing 80 clients to 80 hours per week (sustainable) managing 208 clients. The per-client time investment dropped from approximately 1.4 hours per week to 0.4 hours per week.

Revenue Impact

With 208 clients at an average fee of £120 per month, the practice's monthly recurring revenue reached approximately £24,960 — up from approximately £9,600 with 80 clients. Annual revenue increased from approximately £115,000 to approximately £300,000.

The practice did not hire additional staff. The only additional cost was the clients' own Accounted subscriptions, which clients paid directly. The practice portal remained free. The volume pricing meant clients at the 200+ tier received a 30% discount on their subscriptions, which Sarah highlighted as a benefit of working with the practice.

The practice also participated in the partner programme, earning revenue share on referred clients. At the 20% tier with 200+ clients, this added approximately £750-800 per month to practice income.

Review Quality

Sarah commissioned an external quality assurance review in November 2025 to verify that the rapid growth had not compromised quality. The reviewer examined a sample of twenty client files across the full range of complexity.

The findings were positive:

• Workpapers were comprehensive and consistent across all files (a notable improvement from the variable quality of manually prepared workpapers)
• Categorisation accuracy was high, with a less than 1% error rate on auto-applied transactions
• The review process was well-documented and auditable
• The reviewer noted that the AI-assisted process provided a more robust audit trail than typical manual processes

Sarah also found that the shift clarified the distinction between bookkeeping and accountancy work within her practice. With Penny handling the bookkeeping layer, her qualified staff could focus entirely on the accountancy: review, advisory, and compliance. For practices evaluating platform options, the comparison between Accounted and Xero was one of the resources Sarah used during her decision-making process.

Client Satisfaction

Sarah surveys clients annually. The 2025 survey (the first full year on Accounted) showed:

• 92% rated the bookkeeping experience as "easy" or "very easy" (compared to 61% the previous year)
• 87% said they maintained records more consistently than before
• 94% were satisfied with the WhatsApp-based communication
• The most common positive comment was about simplicity: clients appreciated not needing to learn accounting software

Three clients left the practice during the eighteen-month period, all for reasons unrelated to the software change (two retired, one emigrated). No clients cited the switch to Accounted as a negative factor.

Key Lessons

Sarah identified several lessons from the experience that may be useful to other practices considering a similar path.

Start Small but Start Now

The three-month pilot was essential for building confidence, but Sarah wishes she had started it sooner. The six-month migration period was comfortable but could have been compressed to three or four months without difficulty. The sooner you begin, the sooner the benefits accumulate.

Let the AI Learn Before Judging It

The first month on Accounted shows the AI at its weakest — it has not yet learned client-specific patterns. Judging the system's accuracy based on the first month is misleading. By the third month, the flag rate drops significantly, and by the sixth month, it reaches a steady state that is remarkably accurate. Patience during the learning period is important.

Reinvest Time Savings Strategically

The time freed up by AI-assisted review is valuable. Sarah chose to reinvest it roughly equally in three areas: advisory services (higher revenue per hour), business development (client growth), and personal time (sustainable working hours). The balance will be different for every practice, but the key is to be intentional about how the time is used rather than letting it fill with low-value tasks.

The MTD Catalyst

The approaching MTD for Income Tax deadline was a significant catalyst for client migration. Many clients who might otherwise have been reluctant to change software were motivated by the need for MTD-compliant record-keeping. Sarah used the regulatory requirement as a natural conversation point: "You need to change something to comply with MTD. Let me recommend the option that makes it easiest for both of us."

Communicate the Change Positively

Clients respond well when the change is framed positively. Rather than "we are moving you to new software," Sarah framed it as "we are upgrading your bookkeeping to an AI-assisted system that makes your life easier and your records better." The WhatsApp demo — showing a client how they can photograph a receipt and have it instantly categorised — was particularly effective. For guidance on framing these conversations, our post on client communication tips offers practical approaches.

Is This Replicable?

The obvious question is whether Harrison & Co's experience is typical or exceptional. Based on the data from practices using Accounted's practice portal, the efficiency gains are consistent. Practices that adopt exception-based review routinely report 65-80% reductions in per-client review time, consistent with Harrison & Co's experience.

Whether a practice can triple its client base depends on factors beyond software: marketing capability, local market demand, the practice's service quality and reputation, and the principal's appetite for growth. Accounted provides the capacity for growth by removing the bottleneck. Whether the practice chooses to grow, and how aggressively, is a business decision.

What is clear from this and similar examples is that the conventional ceiling — fifty to seventy clients per practitioner — is an artefact of the traditional review workflow, not an inherent limitation of the profession. When the review workflow is transformed by AI assistance, the ceiling rises dramatically.

For practices interested in exploring this path, the starting point is straightforward. Set up a free practice portal account, pilot with a few clients, and measure the results. The AccountingWeb community also has active discussions from accountants at various stages of the same transition, offering peer perspectives on the experience.

The MTD for Income Tax mandate arriving in April 2026 will create similar pressures for practices that have not yet adapted their workflows. Harrison & Co's early adoption positioned them well ahead of this deadline.

Harrison & Co's journey from eighty clients to two hundred and eight is not a story about technology. It is a story about removing a bottleneck that was limiting a capable team's potential. The technology was the enabler. The ambition and execution were the practice's own.

Visit our page for accountants to learn how Accounted can support your practice's growth.

You may also find our The Accountant's Guide to Recommending AI Bookkeeping helpful.

See our detailed comparison: Accounted vs Traditional Practice Management.

Related reading: Why Accountants Should Recommend AI Bookkeeping.

Accounted gives accountants a free practice portal — manage all your clients, file to HMRC, and let Penny handle the routine work. See the accountant portal →