More Staff Won't Fix This

A professional services firm hits £2M revenue with 12 employees. Partners are working late. Staff are stretched. Client delivery is starting to slip. The obvious answer: hire.

They bring on four people over six months. Revenue climbs to £2.6M. Better, right?

Except the partners are still working late. Staff still feel stretched. And now there are more Slack channels, more meetings, more handoffs, and more things falling through cracks that didn't exist before.

Revenue per employee went from £167K to £163K. They got bigger but they didn't get better.

Each hire added capacity. But each hire also added coordination costs, management overhead, onboarding time, and new process bottlenecks that nobody budgeted for. The office that felt chaotic with 12 people now felt chaotic with 16 people, just in different ways. The org chart grew. The efficiency didn't.

Here's the uncomfortable part: nobody noticed. Revenue was up 30%. That's the number that went in the board deck. That's the number the partners told themselves at year-end. The fact that they needed 33% more people to produce 30% more revenue? That didn't make the slide.

The metric that would've caught it: revenue per employee. Not because it's hard to calculate. It's division. But because the answer forces a conversation most firms aren't ready to have.

Annual revenue divided by total employees. Include everyone: partners, support staff, part-timers on a pro-rata basis. The number you get is revenue per head.

Industry benchmarks vary, but here's roughly where firms tend to land:

Law firms sit between £120K and £250K per employee, with top performers above £300K. Accounting practices range from £80K to £180K. Marketing agencies: £90K to £160K. IT managed service providers: £100K to £200K. Construction tends higher at £150K to £300K because of project-based revenue concentration.

These are approximate ranges. Your specific number matters less than the direction it's moving.

Here's where the story gets uncomfortable. Watch what happens over three years at a growing firm:

Year one: £1.5M revenue, 10 staff. Revenue per head: £150K. Year two: they hire four people to handle growth. Revenue hits £2M. Revenue per head: £143K. Year three: they hire again. Revenue reaches £2.4M with 18 staff. Revenue per head: £133K.

Revenue up 60%. Headcount up 80%. Revenue per head down 11%.

That 11% decline is the scaling problem hiding behind the top-line growth. It means every new pound of revenue is costing more to produce than the last. The business is growing, sure. But it's growing in a way that makes the owner busier, not wealthier.

If your ratio is flat or declining year-over-year while revenue rises, you're not scaling. You're multiplying. Use our revenue-to-employee calculator to figure out your math.

I want to be careful here. Hiring isn't wrong. Growing your team is often necessary and good. But hiring as the default response to every capacity problem is where firms get stuck.

Every new hire adds layers that don't show up on the job description.

Management overhead. The person who manages them loses productive time. A partner who was doing 30 hours per week of billable work hires an associate. Now the partner does 20 hours of billable work and 15 hours of supervision, review, and correcting drafts. Net productive capacity didn't increase. It shifted.

Communication complexity. Meetings multiply. Slack channels grow. Email chains get longer. A 10-person firm has 45 possible communication pairs. A 15-person firm has 105. That's not abstract. That's more messages, more misunderstandings, more time spent aligning.

Process variation. New people do things slightly differently. Not wrong, just different. The onboarding documentation that doesn't exist becomes a problem when three people are doing the same task three different ways.

Onboarding drag. Three to six months before someone is fully productive. During that window, they're consuming senior time, asking questions, and learning by trial and error. Necessary, but expensive in a way that never gets invoiced.

Coordination costs. More people means more handoffs. More handoffs means more things that get dropped, delayed, or done twice.

The cruel irony: you hired because you were too busy. Now you're busier. Just differently busy. Instead of doing the work, you're managing people doing the work. The calendar that was full of client work is now full of one-to-ones, reviews, and process discussions.

A partner at an accounting firm told me something that stuck: "I used to work on clients. Now I work on staff problems. I'm not sure when the switch happened." It happens gradually, and it happens to almost everyone who hires reactively instead of strategically.

Not every task that needs doing needs a person doing it. Some should be eliminated. Some should be automated. The ones that remain should be delegated. In that order.

Three questions for every capacity problem. Ask them in this order.

Does this work need to exist at all?

Reports nobody reads. Meetings where nothing gets decided. Approval steps that add no value but add two days to the timeline. Status updates that duplicate information already in the project tool.

Eliminate before you optimise. Dead work absorbs living resources. I've seen firms where 15-20% of recurring tasks exist only because "we've always done it that way." Nobody questioned whether the monthly report that takes someone half a day to compile is actually used by anyone. Sometimes the answer is no. Sometimes it's been no for years.

Does this work require human judgment?

If yes, hire or delegate. If no, automate.

Most tasks aren't purely one or the other. They're 80% process and 20% judgment. The smart move is to automate the 80% and keep humans for the 20%.

Take conflicts checking at a law firm. The search, matching entity names, traversing corporate hierarchies, cross-referencing opposing parties, is repeatable and rule-based. The decision about what to do with a flagged conflict requires legal judgment. Currently, at most firms, one person is doing both. A partner billing at $380/hour spending three hours on what's essentially a database search before spending ten minutes on the actual judgment call.

Does this work vary or is it repeatable?

Highly variable, client-facing, emotionally complex? Human. Repeatable, rule-based, high-volume? System.

Apply this to common professional services bottlenecks and the split gets obvious fast. Client intake paperwork: automate. Engagement letter generation: automate. Follow-up reminders: automate. Conflicts checking: automate the search, keep the human judgment. Strategic advisory: keep human. New business development: human relationships, but the sourcing and outreach can be systematised.

The firms that get this right don't hire less. They hire better. Their people do the work that actually requires a person. Everything else has a system.

Companies with rising revenue per employee share a few patterns. They're not doing anything exotic. They're just asking the right question before they post a job ad.

They automate before they hire. When something becomes a bottleneck, the first question isn't "who do we bring in?" It's "does this need a person?" Often it doesn't. The Nashville law firm from this week's Blueprint is a good example. Partners spending 28 hours per month on conflicts checks. That's not a hiring problem. That's an automation problem. The agent cost $180/month. The partner time recovered was worth $10,640/month. Revenue per employee didn't improve because they added headcount. It improved because they stopped burning partner hours on a search function.

They eliminate dead work before they optimise existing work. No point in making a useless process faster. Kill it first. Then look at what's left.

They measure capacity in output, not hours. "We're all working 50-hour weeks" isn't a capacity metric. It's an exhaustion metric. The question is what those hours produce, not how many there are. A recruiter spending four hours a day on manual candidate sourcing is busy but not productive. The same recruiter spending those four hours on candidate conversations and client calls is both.

And they treat every bottleneck as a decision point: hire, automate, or eliminate. Not a default hire. Some firms we've talked to have discovered that two or three well-placed automations would've been more impactful than the last three hires they made. They don't regret the hires. But they wish they'd asked the question first.

Pull up your revenue number from this year and last year. Divide each by total headcount. Compare. Or just use our calculator —>

If the number is rising, you're scaling. Whatever you're doing, keep doing it.

If it's flat, you're growing linearly. Every new pound of revenue requires roughly proportional new expense. Sustainable, but you're not building any operational advantage.

If it's declining, at least one of these is true: you're hiring to solve problems that should be automated. You have dead work consuming live resources. Your best-paid people are spending time on your lowest-value tasks.

None of those are fatal. All of them are fixable. But fixing them starts with seeing them, and revenue per employee is the metric that makes them visible.

Want to find where the gaps are in your firm? The AI Bottleneck Audit takes 5 minutes and shows you exactly which tasks are dragging your ratio down. No pitch. Just the maths.

Want to see how other SMBs improved their ratios? Download Unstuck. Twenty-eight real stories of firms that grew revenue without growing headcount.

by SP, CEO - Connect on LinkedIn
for the AdAI Ed. Team