James (not his real name) had done good work before we met.
He’d inherited a turnover problem running at 40%. He suspected salary was part of the story, looked at the data, confirmed it and acted. By the time we started working together, things were improving.
But something still didn’t sit right. The fix worked. The friction remained.
Staff were acutely sensitive
They were complaining about even the smallest differences between colleagues.
Who was coming in only a minute or two later and ‘getting away with it’, who was getting the supposedly more interesting or easier work. Like younger siblings fighting over which shape of TeeVee Snacks they got, angling for the larger ones, when they’re virtually identical. They’re an Australian chocolate biscuit, if you’re not familiar.
That was a signal.
Best practice: Decide before you see
Before I became an executive coach, I spent years as a market researcher. Best practice is to decide what you want to know before you even design the research.
You think about the questions you want answered. You decide what to measure, what to ignore and how you’ll analyse it. You anticipate which combinations of the data might reveal something useful.
You start with your framework and hypotheses — and remain open to being wrong. The data either confirms your expectations or points to somewhere more interesting.
Re-examining the data
So we slowed down and re-examined James’ turnover data.
When were people leaving? Not mid-winter, as he’d assumed. Just before summer.
Were certain areas losing more people than others? He hadn’t checked. One manager with poor people skills can skew numbers to look systemic.
Scheduling patterns emerged. Staff had been underpaid relative to market; but the data shows they also didn’t have enough consecutive days off. Early starts were wearing people down in ways a pay rise couldn’t fix.
And the students? They were different again with different drivers, expectations, and retention levers.
If he’d stopped at salary, the numbers would have stabilised but the underlying strain would remain.
Numbers don’t tell the full story
The numbers were accurate. The meaning wasn’t obvious.
One staff member was on lighter duties. Colleagues noticed. But when James looked at the context he found the person was on a final warning, unknown to everyone else. All they saw was lighter duties. A perceived injustice.
Resentment can form from incomplete information. A complaint isn’t always what it seems. A pattern in the numbers doesn’t always point to what you think it does.
Without context, you can’t tell.
Better data helps. Slicing it carefully helps. Exit interviews help. But none of it is sufficient without interpretation.
Interpretation requires judgment
What’s actually driving this behaviour? What can people see, and what can’t they? What assumption might be wrong?
Even disciplined leaders can slip into solution mode too early. Especially when the first answer seems obvious.
If James had kept layering bonuses and allowances without asking harder questions, he would have spent money and still lost people.
Misdiagnosis is expensive.
It drains budget, energy and credibility.
The discipline is not in acting quickly. It’s in asking better questions.
That kind of discipline is tricky from inside the problem. That’s where executive coaching can help.
The question for you
What are you solving right now that might not be the real problem?
