The pay rise conversation I got wrong

The first time someone asked me for a pay rise, I said: 'Let me see what I can do.'

It sounded reasonable. It sounded like a yes forming. What it actually was is the single worst sentence available in that conversation, and it took me losing a genuinely good employee to understand why.

What actually happened

She'd been with us a bit over two years. She'd taken on things nobody asked her to take on, she was the person clients rang first, and she'd quietly become load-bearing in a way I'd noticed and never once said out loud. She asked for a meeting, was visibly nervous, and made a well-prepared case for a rise that, looking back, was entirely fair.

I wasn't ready. Cash was tight that quarter, I hadn't looked at anyone's salary in a year, and I had no framework at all for what people should be paid. So I stalled: let me see what I can do. She thanked me. Nothing was agreed, no date was set, and the meeting ended with both of us relieved it was over.

Then I did the thing that turned an awkward meeting into a resignation — I didn't come back to her. Not out of malice. The month got busy, the answer was going to be difficult, and every week that passed made it slightly easier not to raise it. Six weeks later she handed in her notice for a job paying a bit more than she'd asked me for.

Why 'let me see what I can do' is the worst answer

Because it isn't an answer. It's a delay dressed as goodwill, and it hands all the discomfort back to the person who was brave enough to ask.

Asking for more money is genuinely hard. Whoever is sitting across from you has probably rehearsed it, likely for weeks. What they need is a clear response — yes, no, or a specific date by which you'll answer and exactly what you'll be weighing. What they get from a vague maybe is a fortnight of wondering, then another, and the slow realisation that they'll have to raise it again themselves. Most people won't. They'll do the far easier thing and let the market answer for them.

Silence after a pay conversation isn't neutral. To the person who asked, it reads as a no delivered by someone who couldn't be bothered to say it.

What I'd learned by the second time

The next time it happened, I handled it almost the opposite way, and it went fine even though the immediate answer was no.

I said I couldn't commit to anything in that meeting because I wanted to look at it properly rather than react. I gave a date — two weeks — and stuck to it. When we met again I explained where the business genuinely was, what a rise would need to look like, and what specifically would need to change for it to be affordable in six months. Then I put that in writing, because a verbal commitment from a busy owner is worth very little to the person waiting on it.

He didn't get the money that month. He got a straight answer, a plan, and the rise six months later. He's still with us. The difference wasn't the money — it was being treated like an adult who could handle the real position.

The number matters less than the reasoning

The thing I'd assumed, and got wrong, is that these conversations are about the figure. Mostly they aren't. They're about whether the person believes the process is fair and whether their work is actually seen.

People will accept 'not right now, and here's honestly why' far more readily than they'll accept an unexplained number. What they won't accept for long is the sense that pay in your business is arbitrary — that it depends on who asks, how confidently they ask, and what mood you're in that week. That perception, once it settles, does more damage than any single salary decision, because it tells everyone the only way to get paid properly is to go and get an offer somewhere else.

What I do now

Four things, none of them complicated. Salaries get reviewed on a fixed date every year for everyone, whether they ask or not — which means most of these conversations never have to be initiated by a nervous employee at all. I keep a rough sense of the market rate for each role, so I'm not negotiating from nothing. Any pay conversation gets an answer within two weeks, in writing, even when the answer is no. And if the answer is no, it comes with what would need to be true for it to become yes, and when we'll look again.

The wider point is one every small employer works out eventually, usually the expensive way. Losing a good person costs far more than the rise they asked for once you count recruitment, the months of lost output, and everything they knew that was never written down anywhere — a cost we've looked at directly in the real cost of a bad hire, and it applies just as brutally to losing a good one.

She'd have stayed for a fraction of what her leaving cost us. I just didn't have a process, so I had a vague sentence instead — and a vague sentence is what people remember.