A mid-market business cannot win a bidding war against an employer with a larger balance sheet. It does not have to. The number that is actually recoverable is not the salary — it is the cost of the vacancy that keeps reopening.
The conversation about technology talent in mid-market businesses tends to arrive at the same place. Salaries have risen faster than the business can absorb, larger competitors pay more, and the conclusion is that the market is impossible and the business will simply have to cope.
Some of that is true. Senior software engineering salaries in London now sit in a band that a £40m-turnover business cannot match position for position, and a majority of digital leaders expect tech salary demands to become unsustainable within two years. That constraint is real and is not going away.
What follows from it is usually wrong. The response is framed as a hiring problem, and the money goes into recruitment. The larger and more controllable number sits on the other side of the equation.
The number nobody adds up
Oxford Economics research puts the average cost of replacing an employee at around £30,600. That figure scales with seniority: roughly 16% of salary for roles under £30,000, 20% between £30,000 and £50,000, 25–35% for senior positions, and dramatically higher at executive level.
£30,614
the average cost of replacing an employee earning over £25,000, on Oxford Economics research — recruitment, onboarding, and the productivity lost either side
Set that against UK turnover averaging around 34% — more than one in three workers changing jobs in a year — and the arithmetic gets uncomfortable quickly. A forty-person business losing a third of its people annually is carrying a replacement cost in the region of £400,000 a year, and almost none of it appears as a line item anyone owns.
It is distributed instead: agency fees in one budget, the contractor covering the gap in another, the delayed project in a third, and the least visible portion of all — the six months during which a capable new joiner is still learning where things are and why they are that way.
Turnover does not appear on the P&L. It appears as agency fees, contractor cover, slipped projects and a permanent shortfall in institutional memory, spread across four budgets and owned by nobody.
Why paying more does not fix it
The obvious response — raise salaries — fails for mid-market businesses in a specific way.
Pay is the strongest single driver of moves, cited by around 53% of people considering one. So salary matters and cannot be ignored. But a business that competes purely on salary against organisations with deeper pockets is committing to a contest it will lose repeatedly, and each loss is expensive.
There is a related trap that is worse. Hiring below market to control cost produces people who leave within twelve to eighteen months when a better offer arrives — which means paying the replacement cost repeatedly for the same seat. Underpaying is not cheaper. It converts a salary difference into a recurring £30,000 charge plus the disruption around it.
Career progression is the second driver, cited by around 39%. That one is not a bidding war, and it is where a mid-market business has an advantage it usually fails to press.
What mid-market businesses can actually offer
The advantage is scope. In a large organisation a capable engineer owns a component. In a forty-person business the same person can own a system, talk to the customers who use it, see the commercial consequence of their decisions, and change something that matters within a quarter.
That is genuinely attractive to a particular kind of person, and it costs nothing. What it requires is that the business actually deliver it, which is where most fall down. The scope is offered at interview and then withdrawn in practice — the person who was going to own the platform spends their time on tickets, because the ticket queue is urgent and the ownership was never scheduled.
Three things make the difference between offering scope and providing it.
Someone must protect the time. Ownership requires uninterrupted stretches. A person interrupted daily cannot own anything; they can only respond. That is a management decision about interruption routing, not a cultural aspiration.
Progression has to be legible. Not a formal framework — most mid-market businesses do not need one — but an honest answer to “what is different about my job in two years”. Absence of an answer is read, correctly, as absence of a path.
Key-person dependency has to be treated as a risk, not a compliment. The person who is the only one who understands the billing system is not an asset; they are a single point of failure who cannot take leave, cannot be promoted without leaving a hole, and knows it. Removing that dependency is usually experienced by the individual as relief rather than diminishment.
Four numbers to put in front of the board
- How many people left in the last twelve months, and what did replacing them cost — agency fees, contractor cover, and the months to full productivity?
- Of those, how many were within eighteen months of joining? That is the below-market-offer tax, and it is repeat business.
- Which systems have exactly one person who understands them, and what is the plan for each?
- How much of our senior engineering time went on work that did not require senior engineering — and what did that cost against what we pay for it?
The honest limit
None of this means a mid-market business can ignore pay. Where a salary is materially below market, no amount of interesting work compensates for long, and pretending otherwise produces exactly the eighteen-month churn described above. The first move is usually to correct the genuinely uncompetitive cases, which is cheaper than it looks precisely because the replacement cost is already being paid.
What it does mean is that the framing is wrong when the whole conversation is about hiring. A business that halves its turnover has changed its capacity more reliably than one that hires two more people, and has done it with money it was already spending.
The starting point is the first question in the box. Most management teams have never calculated what leaving cost them last year. It is a morning’s work from data the business already holds, and it tends to reframe the conversation faster than any argument about culture.