A growing accountancy firm typically moves through three shapes: flat, where everyone reports to the owner directly; pods or team-lead groups, where a named senior owns review and day-to-day decisions for five to seven people; and a graded, service-line structure, where department heads sit between the owner and several team leads. Move when the owner’s review queue or interruption count becomes the bottleneck, not on a fixed headcount, and change the shape once, deliberately, rather than letting it drift one ad-hoc promotion at a time.
Why the shape that worked at eight people fails at twenty
Every firm starts flat, and for good reason — with five or six people, the owner can see every file, answer every question and review every job, and doing so is genuinely the fastest way to get the work out. The trouble is that this shape has a hard ceiling built into it, and the ceiling is not headcount. It is the owner’s attention, which does not get bigger as the firm does.
A flat structure scales in exactly one dimension: how many people can queue behind one reviewer before the queue itself becomes the constraint. At six people that queue is short and mostly invisible. At sixteen, the same structure means sixteen people bringing questions, files and exceptions to the same desk, and the owner spends the day processing other people’s decisions rather than making the ones only they can make. Nothing about the firm has gone wrong. The structure that was correct for its size has simply stayed in place past the size it was correct for.
The signal is rarely headcount on its own. It is the combination of headcount and what is now routing through one person: a review queue that never empties, interruptions that arrive all day rather than at natural breaks, and decisions that wait not because they are hard but because there is only one person who can make them. That is the same bottleneck described in how to build a management layer in your practice, but structure and authority are two different fixes. A management layer decides who can make a decision. Team structure decides how many people sit between that decision and the owner, and how they are grouped — by client type, by service line, or not grouped at all. Most firms need both, but they are separate pieces of work, and firms that only do the first often find the second still stalling them.
The three shapes a growing firm moves through
There is no universally correct structure, but there is a fairly reliable sequence, and skipping a stage is where most of the pain comes from. Reorganising once, deliberately, at the right point beats a structure that has drifted into its current shape through a series of individually sensible promotions.
| Roughly this size | Shape | How review and decisions flow | What breaks if you stay here too long |
|---|---|---|---|
| 1–6 people | Flat | Everyone reports to, and is reviewed by, the owner directly | Nothing yet — this is the right shape at this size |
| 7–15 people | Emerging pods | One or two senior staff take informal responsibility for reviewing routine work in their area; the owner still sees anything non-standard | The “informal” part — nobody has actually been told they can decide, so questions still default to the owner out of habit |
| 16–30 people | Named pods or service lines | Three or four groups of five to seven, each with a named team lead who reviews and allocates within their group; the owner sees exceptions and the team leads directly | A team lead with no reduction in their own chargeable work, so leading happens after hours or not at all |
| 30+ people | Graded structure | Department or service-line heads sit between the owner and several team leads; a career ladder gives people somewhere to grow into other than “partner” | Team leads reporting straight to the owner again, because nobody removed that link when the layer above them was added |
The two ways to group people into pods are worth naming, because firms often default to one without considering the other. By service line — a compliance pod, an advisory pod, a payroll pod — concentrates expertise and makes technical review straightforward, but it means most clients touch two or three pods and nobody owns the whole relationship. By client type or portfolio — each pod holds a mixed set of clients end to end — keeps one relationship owner per client and makes onboarding and continuity simpler, at the cost of each pod needing broader technical range. Firms with a narrow client base and complex technical work usually do better organised by service line; firms built on relationship-led, varied small-business clients usually do better organised by portfolio. There is no wrong answer, but there is a wrong amount of thought given to the choice, which is usually none.
Worked example: what changes when a flat firm moves to pods
The figures below describe an illustrative 22-person firm, not a client of ours, moving from a flat structure to three pods of six to seven people, each led by a promoted senior. It is meant to show the shape of the effect, not a number any specific firm should expect to match.
| Measure | Flat structure | Three pods with team leads |
|---|---|---|
| People per reviewer | 22 report to the owner | 6–7 report to each of three team leads |
| Owner’s review queue | 34 jobs waiting, oldest 9 days | 9 jobs waiting (exceptions only), oldest 3 days |
| Owner interruptions per day | Around 11 | Around 4 |
The cost side is smaller than owners expect. Say each promoted team lead moves from a £42,000 senior salary to a £46,000 team-lead salary — a £4,000 uplift. Employer’s National Insurance at 15% above the £5,000 secondary threshold adds £600 on that uplift, and the 3% employer minimum pension contribution on qualifying earnings adds a further £120, since both the old and new salary sit inside the £6,240–£50,270 qualifying band for 2026/27. That is £4,720 of extra annual cost per team lead, or £14,160 for three.
Against that, if the owner genuinely redirects the freed time — roughly seven fewer hours a week of review and interruption, 44 working weeks a year, at an illustrative £150 an hour for owner time — that is £46,200 of owner time no longer spent processing other people’s decisions. The arithmetic only works if that time goes somewhere specific: advisory delivery, business development, or planning. Recovered hours that simply refill with more of the same interruptions, redirected from three people instead of twenty-two, are not a saving at all, just a smaller version of the same problem. That is the same trap described in how to delegate effectively as a firm owner, and it applies to structure exactly as it does to individual decisions.
Getting the spans and the grouping right
Two design choices decide whether a new structure holds or quietly reverts to flat within six months.
- Span of control follows review load, not affection. The practical ceiling for one team lead reviewing everyone’s work in an accountancy pod is around five to seven people, because review time is what actually limits the number, not how many names fit tidily on an org chart. Push past seven without changing what has to be reviewed at team-lead level — raising the sign-off threshold for routine work, using checklists so juniors self-check first — and you have simply built a second bottleneck one level down, which solves nothing.
- Something has to come off the team lead’s plate. The single most common way a new pod structure fails is a team lead who keeps a full chargeable workload and is handed review, allocation and one-to-ones on top of it, with nothing removed. Decide the percentage of chargeable time coming off before the change is announced, not after someone is visibly drowning.
- Draw the lines before you announce them, not while people are asking questions. Who is in which pod, who each pod reports to, and what moves to the team lead versus what still comes to the owner should exist on one page before the first conversation happens. A structure that is worked out client by client, in real time, teaches the team that the lines are negotiable, and negotiable lines get walked around.
It is worth deciding, at the same time, what the new shape does to key-person risk. A pod structure that concentrates one client group entirely behind one team lead can quietly recreate the founder-dependency problem one level down, just with three founders instead of one. Building in a genuine second reviewer per pod, even lightly, is the difference — see how to reduce key-person risk in your firm for how to check for this before it becomes visible the hard way.
Making the change without unsettling clients or the team
Restructuring an accountancy firm is not like restructuring most other businesses, because every internal reporting line also touches a live client relationship, an ongoing job and a deadline that does not pause for the reorganisation. Three things keep the change from becoming disruptive.
- Move at a quiet point in the year, not the busiest one. Announcing a new structure in the run-up to the 31 January self assessment deadline, or immediately before a company’s year-end crunch, guarantees the change is judged by how it performed under maximum stress rather than on its own merits. A quieter month gives new team leads time to find their feet before anything is tested hard.
- Tell clients about continuity, not organisation charts. Clients do not need to hear about pods and spans of control. They need to know who their point of contact is and that nothing about service is changing for the worse — ideally introduced by the outgoing contact alongside the new one, not by letter.
- Put the new structure into the existing operating rhythm immediately. A structure only becomes real once it shows up in how the firm actually runs week to week — who attends which meeting, whose numbers get reviewed where, who reports what in the monthly operating review. A structure that exists on a chart but not in the weekly rhythm reverts to the old habits within a term, because habits are stronger than diagrams.
Run the new shape for a full quarter before adjusting it further. The first few weeks will surface genuine mistakes — a pod that is unbalanced, a team lead who needs more support than expected — and the temptation is to fix each one as it appears. Resist tinkering weekly; collect what is not working, then make one deliberate set of adjustments at the three-month mark. A structure that changes every fortnight teaches people not to bother learning it, which defeats the purpose of having one. Sequencing a change like this properly, alongside everything else competing for a growing firm’s attention, is exactly the kind of work the Optivo method is built to stage, and the Optivo Monthly COO exists to keep on track once it is live.
Common questions
At what headcount should an accountancy firm move away from a flat structure?
There is no fixed number, but most firms feel the strain somewhere between eight and fifteen people, when the owner’s review queue stops emptying overnight and interruptions start arriving throughout the day rather than at natural gaps. Watch the symptom rather than the headcount: a review queue with a genuine backlog, jobs waiting on a decision only the owner can make, or the owner routinely working evenings to clear what could not be reviewed in the day. A firm of twelve with excellent delegation habits can stay flat comfortably; a firm of nine where everything already funnels through one person is already overdue for pods. Structure should follow where the bottleneck actually is, not a headcount milestone on a spreadsheet.
What's the difference between organising pods by service line and by client type?
Organising by service line — a compliance pod, an advisory pod, a payroll pod — concentrates technical expertise and makes review straightforward, because everyone in the pod does similar work, but most clients then touch two or three pods and no single person owns the whole relationship. Organising by client type or portfolio keeps one team, led by one team lead, responsible for a client end to end, which makes onboarding, continuity and accountability simpler, at the cost of each pod needing broader technical range across its members. Firms with narrow, technically complex client bases usually do better by service line; firms built on varied, relationship-led small-business clients usually do better by portfolio. Pick deliberately rather than by default, and expect to revisit the choice as the firm grows further.
How many people should report to one team lead?
The practical ceiling in an accountancy pod is around five to seven people, and the constraint is review capacity rather than management skill or how many names look tidy on a chart. A team lead reviewing everyone’s output in a pod of six or seven is already close to full; push the group to ten or twelve without changing what has to reach them for review, and the pod simply relocates the original bottleneck one level down instead of removing it. If a wider span is genuinely needed, reduce what requires team-lead review first — raise the sign-off threshold for routine, low-risk work, and use documented checklists so juniors self-check before anything reaches the team lead at all.
Should a firm change its structure gradually or all at once?
Decide the whole shape at once — who is in which pod, who leads each one, and what decisions move to team-lead level — then announce and implement it as a single, deliberate change rather than a series of individual promotions made as the need arises. A structure that accumulates one ad-hoc decision at a time tends to end up organised around who happened to be available when each gap appeared, rather than around what the firm actually needs, and it is far harder to unpick later than to design properly once. That said, once the new shape is live, resist adjusting it every few weeks in response to early friction. Run it for a full quarter, note what genuinely is not working, and make one considered set of changes at that point rather than continuous small revisions.
How is restructuring the team different from building a management layer?
They are closely related but answer different questions, and most growing firms eventually need both. A management layer is about authority — which decisions stop one level below the owner, and who has been given the right to make them, covered in how to build a management layer in your firm. Team structure is about shape — how many people sit between a decision and the owner, and how they are grouped, whether by service line, by client portfolio, or not grouped into pods at all. A firm can grant a manager real authority while still having everyone report through one flat line, and it can reorganise into tidy pods while every meaningful decision still travels all the way to the owner. Getting the shape right makes the authority easier to exercise; it does not replace the need to grant it.
How do you decide who leads each pod?
Look for someone the team already defers to informally on technical questions and who is willing to have an uncomfortable conversation, rather than automatically choosing the most senior or longest-serving person in the group. Technical strength and the willingness to review other people’s work, allocate it fairly and hold a standard under pressure are genuinely different capabilities, and assuming the best technician will also be the best team lead is one of the more expensive mistakes a growing firm makes. Whoever is chosen needs a defined reduction in their own chargeable workload before the role starts, agreed and communicated in advance, and a fixed weekly slot with the owner in the first quarter to work through what is landing on their desk. Without both, the promotion is a new title on an unchanged job.
Where does your firm actually stand?
Score it across the ten things a COO would look at first. Three minutes, an honest read on where you’re strong and where you’re exposed, and a 90-day plan for the areas holding you back.
Or book a confidential call if you’d rather just talk it through.