Back to Blog

What does it really cost when an experienced employee leaves?

More than your turnover report says. The rule most finance teams use prices a leaver at about half of salary, which covers the empty seat but not the judgement that walked out with them. Price the decisions a successor gets wrong while they learn the role, and a late career exit can cost several times what the formula books. That's why most retention budgets are aimed at the wrong end of the career.

Every retention budget I've been shown was built from a headcount number. Not one was built from a value number. You can prove that inside your own data in about forty minutes.

The number on your dashboard counts bodies

Voluntary turnover, as a percentage. It goes to the board, it gets benchmarked against the sector, and somebody owns bringing it down.

It counts bodies. The analyst who resigns in March after fourteen months and the operations manager who has been quietly holding a region together since 2009 land on the same line of the same report, each one worth exactly one.

So the money follows the volume. Most of your attrition sits under 35, because it always does, which means most of your retention spend sits there too. Onboarding redesign. Engagement pulses. The early careers pathway somebody presented at the last People committee. All defensible, all aimed at the biggest pile of leavers.

The biggest pile isn't the most expensive pile. Those are different questions, and your dashboard only answers one of them.

Work the number

Take a company of 1,200 people running voluntary turnover at 14 percent. That's 168 people out the door in a year.

Split them the way most organisations actually split: 90 under 35, 50 between 35 and 49, 28 over 50. Use your own numbers when you do this properly. The shape won't be far off.

Price them the way your finance team does. Replacement costs roughly half of salary: agency fees, the empty seat, the ramp.

90 early career leavers at $65,000 comes out around $2.9 million. 28 late career leavers at $130,000 comes out around $1.8 million.

The dashboard isn't lying to you. On that arithmetic early career attrition costs sixty percent more than late career attrition, and the budget belongs exactly where it is.

What the half of salary rule can't see

Half of salary prices the seat. It doesn't price the judgement that was sitting in it.

The 26 year old's replacement is doing the same job to the same standard inside six months, because that job is documented. The 22 year operator's replacement is doing about eighty percent of the job. The missing twenty percent was never a process. Which supplier quietly fails every December. Why that clause is in that contract and what it cost to learn. Which of three plausible explanations for a variance is the real one, without running the analysis.

So put a number on the gap. Say eighteen months before a successor is making calls of the same quality, and one decision a quarter across those eighteen months that the predecessor would have made differently. Six decisions per exit.

Ask an operator what a wrong call costs in their function. In a plant it's a scrapped run. In a multi unit business it's a bad site on a ten year lease.

You don't need six figures a decision for this to flip. You need about $50,000. Twenty eight exits, six decisions each, fifty thousand a decision: $8.4 million. Against the $1.8 million your model booked, and against the $2.9 million you're actively spending to prevent.

"They were leaving anyway"

That's the objection I get next. They're retiring.

In 1996 the labour force participation rate for Canadians aged 55 to 64 was 47.2 percent. In 2025 it was 67.5 percent. Over the same stretch the 25 to 54 rate moved 5.1 points (OECD data via FRED). The older band did twenty points of work and almost nothing about how companies are structured moved with it.

More than half of workers over 50 leave before they chose to. That's a retention failure wearing a demographic alibi, and it's the one category of exit nobody runs a root cause on.

I spend a good part of my week with people on the other side of this. Pushed out, still perfectly capable, now working for themselves. That's where I first noticed how many exits get filed as retirements by companies that never asked.

The one pull to run this week

Out of systems you already own, no survey required. For every voluntary exit over 50 in the last 36 months: how long the role sat open, and how long until the successor was signed off as fully effective by their own manager. Then the same two columns for your under 35 exits.

Months to full effectiveness is the column almost nobody has. If it runs long at one end and short at the other, and it will, you've just priced the thing the dashboard was hiding. Then you can have the real argument about where the retention money goes.

Most organisations find they don't need more money. They need the same money, pointed thirty years later than it's pointed now.

If you'd like this argument made to your leadership team in a room, it's the core of my keynote, The Expiration Date Delusion. You'll find it on the speaking page.

Questions people ask about the cost of losing experienced employees

How much does it cost to replace an employee?

The rule most finance teams use is about half of annual salary, covering recruiting fees, the empty seat and the ramp. That's a fair price for the seat. It leaves out the decisions a less experienced successor gets wrong while they learn the role.

Is it more expensive to lose older or younger employees?

On the half of salary formula, early career attrition usually costs more in total, because there's so much more of it. Once you price the judgement that leaves with a late career employee, the answer usually flips. In the worked example above, 28 late career exits come to about $8.4 million against $2.9 million for 90 early career exits.

What is time to full effectiveness?

It's how long it takes a replacement to be signed off by their own manager as fully effective in the role. Very few organisations track it. It's the column that shows where losing experience really costs money.

Aren't most older employees just retiring?

Many aren't. Participation among Canadians aged 55 to 64 rose from 47.2 percent in 1996 to 67.5 percent in 2025 (OECD data via FRED), and more than half of workers over 50 leave before they chose to. An exit filed as a retirement is often one nobody asked about.

Where should a retention budget go?

Start with your own data: compare months to full effectiveness for the replacements of under 35 and over 50 leavers. Most organisations don't need more retention money. They need some of the same money pointed at the people whose leaving costs the most.