Your Best Team Is One Departure From Being Average

UK vacancies fell to 707,000 in the three months to July 2026, down 6,000 on the previous quarter and continuing a decline that has run for most of three years (ONS, Labour market overview, UK: August 2026, August 2026). Payrolled employees were down 78,000 on the year to June 2026. Vacancies fell in 10 of 18 industry sectors in the second quarter, with the largest falls in professional, scientific and technical activities (ONS, Vacancies and jobs in the UK: July 2026, July 2026).

From a distance that looks like a market where employers hold the cards and retention takes care of itself. From inside a live hiring process it looks nothing of the sort. Inflows into payrolled employment are running 16% below pre pandemic levels, and the ratio of unemployed people to vacancies reached 2.6 to 1, the highest in a decade outside the pandemic (Institute for Employment Studies, Labour Market Statistics, February 2026, February 2026). Plenty of people are available. Movement between good jobs has slowed to a crawl, and the senior person a desk needs is not sitting in that queue.

Which is what turns key person risk into a commercial problem rather than an HR one this year. Concentration of revenue in one or two names has always been expensive when it breaks. The replacement clock is what has changed.

1. The Org Chart Is Not a Map of Where the Money Comes From

Most desks know, informally and to the name, which two people the franchise runs through. It is rarely written down anywhere formal. Headcount plans record seats and titles, budgets record cost, and nothing in either captures the fact that one director holds the relationships behind a third of the revenue line.

The Institute for Employment Studies has argued for decades that retention risk needs two axes rather than one, and the framework has held up well. How likely somebody is to leave is a separate question from what happens if they do. High likelihood alongside high consequence is the danger zone that warrants intervention. Low likelihood alongside high consequence is the category firms ignore, and the one that needs preventative monitoring, because a loyal irreplaceable person remains a single point of failure (Institute for Employment Studies, Managing Staff Retention, Stephen Bevan, 2000).

Try this test: pick the two highest consequence people on the desk, then name who covers each of their top three client relationships next Monday. If the names have to be invented on the spot, the second axis has never been plotted.

2. Replacement Has Got Slower, Not Easier

The intuitive read of a soft labour market says a departure can be backfilled quickly and cheaply. That intuition rests on the volume of available candidates rather than on the flow of people prepared to move.

The flow is where it falls down. Inflows to payrolled employment sit 16% below pre pandemic levels and payrolled employment fell in 10 of the last 14 months (Institute for Employment Studies, Labour Market Statistics, February 2026, February 2026). By May 2026 vacancies had fallen to their lowest level since early 2021 and real pay growth had slowed to a standstill, with private sector real pay growth at its lowest since 2020 (Institute for Employment Studies, May Labour Market Statistics, May 2026). Where pay is flat and openings are scarce, the people good enough to matter stay where they are, and whoever does move is moving for something other than a small rise.

Pricing a key departure at the recruitment fee plus three months of cover understates it badly. The real number is the count of quarters the line runs below plan before a replacement is producing.

3. Pay Is Rarely the Trigger and Rarely the Fix

When a senior producer resigns the reflex is to reach for compensation, because it's the fastest lever and the only one that can be pulled in an afternoon. Evidence on why people leave has been consistent for a very long time, and it doesn't support the reflex.

Only around 10% of departing employees cite dissatisfaction with pay as their primary reason for leaving. The reasons given far more often are work that fails to use their skills, poor management, limited advancement prospects and workload pressure (Institute for Employment Studies, Managing Staff Retention, 2000). Later IES survey work with 1,400 current employees at a professional services firm found that job satisfaction alone explained 43% of the variance in intention to leave, and that five factors, training, work life balance, enjoyment of the work, career development and rewards, together explained 62% of the variance in job satisfaction (Institute for Employment Studies, The 'Great Resignation': is the grass always greener?, June 2023).

Compensation remains the only retention lever most desks can operate at short notice, which is a statement about how the desk is run rather than about what the departing director wanted.

4. Leavers Do Not Usually Regret It

There is a comforting story told about people who leave, in which they find the grass browner than advertised and wish they had stayed. It saves anyone having to ask what the leaver got right.

IES surveyed 1,600 alumni of a large professional services firm and found the opposite. Four in five said the firm had been a good employer and more than 90% said working there had enhanced their marketability, so this was not a population nursing a grievance. Yet fewer than one in five felt their subsequent employer was worse in any significant way, over 40% had received two or three promotions since leaving, and the vast majority expressed no regrets (Institute for Employment Studies, The 'Great Resignation': is the grass always greener?, June 2023). Around 40% would consider returning, which is worth knowing for anyone running an alumni network.

Good people leave and then do well. Plan on that basis rather than on the assumption that the market will teach them something.

5. Your Pay Calendar Announces the Window

Bench depth usually gets discussed as an abstraction. It turns concrete the moment you look at when departures are possible, and in this industry that answer is published.

Average weekly total pay in financial and insurance activities was £3,247 in March 2026, against £1,510 in May 2026 and £1,398 in November 2025 (ONS, AWE: Financial and Insurance Activities, Total Pay Including Arrears, August 2026). The spike is bonus. Compensation in this sector concentrates into a narrow annual window that every competitor can read off a public dataset, which means key person risk is not spread evenly through the year. It crystallises across a few weeks, in every firm at once, at the point when replacement supply is tightest.

A risk that arrives on a known date belongs on the same page as any other seasonal exposure the desk carries.

A Note of Balance

There is a version of this argument that turns into a licence to over hire, and it should be resisted. Bench depth costs money in a year when 3.5% regular pay growth is running ahead of a shrinking payroll and vacancies have fallen for most of three years (ONS, Labour market overview, UK: August 2026, August 2026). Carrying a spare director against a departure that may never come is a real cost set against a hypothetical loss.

Concentration is not automatically a defect either. Small senior teams built around one or two exceptional originators outperform broader, flatter ones in plenty of markets, and diluting that to reduce risk can destroy whatever made the desk profitable in the first place. The aim is to know how much concentration you are carrying and to have decided in advance whether the number is acceptable, rather than to eliminate it.

Conclusion

Key person risk is an unhedged exposure sitting inside a revenue line, and in a market where hiring inflows are 16% below pre pandemic norms it takes longer to close than it used to. Firms that handle it well know, before anything happens, which two departures would move the P&L and what a realistic replacement path for each looks like.

That path cannot be built at the point of crisis. Once the resignation is in, you are shopping among people who have already decided to move, which is a small fraction of the market and rarely the best fraction of it. The alternative is having mapped the function beforehand: knowing who the credible successors are across London and Paris, which of them are originating rather than inheriting, and who would take a call. That is research done from the mandate outwards, and it earns nothing unless it exists before the phone rings.

Sources: ONS, Labour market overview, UK: August 2026, 18 August 2026; ONS, Vacancies and jobs in the UK: July 2026, 21 July 2026; ONS, AWE: Financial and Insurance Activities Level, Total Pay Including Arrears (series K58I), 18 August 2026; Institute for Employment Studies, Labour Market Statistics, February 2026, 17 February 2026; Institute for Employment Studies, May Labour Market Statistics: comment from the Institute for Employment Studies, 19 May 2026; Institute for Employment Studies, Managing Staff Retention, Stephen Bevan, 2000; Institute for Employment Studies, The 'Great Resignation': is the grass always greener?, Stephen Bevan, June 2023.

‍