Employee retention: why people stay, and why they leave

Key takeaways

• People rarely leave for one big reason. They leave because a handful of small things went unaddressed for long enough that leaving started to look easier than staying.
• The drivers of staying and the drivers of leaving aren't mirror images of each other. Pay can stop someone leaving without ever being the reason they stay.
• Exit interviews tell you what already happened. By the time you're running one, the decision is made and the useful window has closed.
• The signals are almost always visible earlier, in engagement data, development activity and manager relationships, if you're looking at the right things.
• Retention isn't a program you run once a year. It's what development, career pathing, fair treatment and being genuinely listened to add up to, every week.
• If you want a real number for what turnover is costing you, our turnover cost calculator will work it out from your own figures rather than a generic industry average.

A client once told me her best account manager resigned "out of nowhere." I asked when she'd last had a real conversation with him about where he saw his career going. She thought about it for a while. Eighteen months, maybe more.

There was nothing sudden about it. He'd been leaving for a year. She just hadn't been in the room.

That's the pattern I see more often than any other. Retention doesn't fail in a moment. It fails slowly, in the gap between what someone needs and what they're actually getting, until the gap gets wide enough that another offer looks like the easier choice. This guide is about closing that gap: what genuinely keeps people, what pushes them out, why the usual measurement (the exit interview) comes far too late, and what to do instead.

Why people stay

Ask most leaders why people stay at their organisation and you'll get "the culture" or "we pay well." Both matter. Neither is usually the whole answer.

In my experience, people stay when most of these are true for them:

• They can see a path. Not necessarily a promotion. A sense that their skills are growing and their role isn't a dead end.
• Their manager is decent. Fair, present, and capable of a real conversation. This one carries more weight than almost anything else on the list.
• The work means something. Not a mission statement on the wall. A felt sense that what they do matters to someone.
• They're treated fairly. Recognition, pay and opportunity that roughly track effort and contribution, and don't quietly favour certain people.
• The workload is sustainable. Busy is fine. Permanently underwater isn't, and people can tell the difference between a hard patch and a hard job.

None of these are exotic. They're also rarely all present at once, which is exactly why retention is a genuine, ongoing piece of work rather than a policy you write once.

Why people leave

Here's the part that surprises people: the reasons someone leaves aren't simply the absence of the reasons they'd stay. Pay is the clearest example. Being underpaid will push people out the door. Being well paid rarely, on its own, keeps someone who's stopped growing, stopped believing in their manager, or stopped feeling like the work matters.

The common threads I see when someone actually leaves:

• They can't see a future here. Not "there's no promotion available right now." More often it's "nobody has told me what growth even looks like."
• Their manager let them down. A pattern of being overlooked, undermined, or simply unsupported through something hard.
• They stopped being asked. Nobody checked in on how they were actually going until the resignation was already written.
• Something felt unfair. A missed opportunity that went to someone less capable. A workload that never evened out. A decision made about them, not with them.
• The workload finally won. Not one bad month. A year of bad months, with no sign it was going to change.

Notice that most of these are relational and structural, not transactional. Money is real, and it matters. It's just rarely the actual story.

Stay-drivers and leave-drivers, side by side

What tends to keep people	What tends to push people out
A visible, credible path for growth	No sense of where this role is heading
A manager who's fair, present and honest	A manager who's absent, inconsistent or unfair
Work that feels like it matters	Work that feels interchangeable or invisible
Recognition and opportunity that track contribution	Recognition and opportunity that seem arbitrary
A workload that's hard but sustainable	A workload that's been unsustainable for too long
Being asked, and being heard	Being managed, but never actually consulted

Look at the pairs. They're related, but they're not identical, and that's the practical point. You can't retain people just by removing the negatives. You have to actively build the positives too. For more on the specific signals that show up before someone resigns, see why employees leave.

Why exit interviews are too late

I understand the appeal of the exit interview. Someone's leaving anyway, so you might as well ask why. The trouble is timing. By the exit interview, the decision has been made, the next role is likely signed, and the person answering you has already emotionally left. What you get is a tidy, retrospective explanation, not the messy, real-time truth of what actually happened over the eighteen months before it.

Worse, exit interviews are self-selecting. The people most willing to give you a full, honest account are often the ones who feel safest doing so, which usually correlates with them not being the ones who felt most let down. The people with the sharpest feedback are frequently the ones who say the least on their way out the door, because they've already decided it isn't worth the effort.

An exit interview can confirm a pattern. It's a poor tool for spotting one early enough to act.

What to measure instead

If exit interviews are the post-mortem, you need something closer to a pulse check. The good news: most organisations already collect data that would tell them this, they just aren't looking at it as a retention signal. Worth tracking:

• Engagement trend for individuals, not just averages. A team score of 7.8 can hide someone who's dropped from an 8 to a 4 over two quarters.
• Development and learning activity. People who've quietly checked out tend to stop engaging with growth opportunities long before they resign.
• Internal mobility. Are people moving into new roles internally, or is every ambitious person's next move an external one?
• Manager relationship signals. One-on-one frequency, cancelled check-ins, a sudden change in how someone talks about their manager.
• Workload and capacity data. Sustained overtime or after-hours activity that's been normalised rather than addressed.

None of these alone tells you someone's about to resign. Together, tracked over time and at the individual level, they tell you a lot more than an annual survey score ever will.

What actually moves retention

If I had to boil it down, four things do most of the work.

Development that's real, not generic. A course catalogue nobody asked for isn't development. People need growth that's matched to where they actually are and where they actually want to go. That's most of what our Transformation work is built around.

A clear career path, even a sideways one. People don't need a five-year plan to a corner office. They need to believe their skills are growing and someone's paying attention to that. See career pathing that actually retains people for what this looks like in practice.

Genuine listening, on a rhythm. Not once a year. Regularly, and specifically. Stay interviews are one of the most underused tools here, precisely because they ask the exit interview's questions while there's still time to act on the answers.

Acting on what you hear. This is the step organisations most often skip. Data without action isn't retention strategy. It's a report nobody reads.

What this is actually costing you

Most leaders have a rough sense that turnover is expensive. Recruitment, onboarding time, lost productivity while a role sits empty, the knowledge that walks out the door with someone experienced. What most leaders don't have is their own number.

I'd rather you work that number out for your own organisation than take a generic industry figure and assume it applies to you, because it won't. Our turnover cost calculator does exactly that. Put in your own figures and see what turnover is actually costing you. It tends to be a more useful conversation starter with the board than any statistic I could quote you.

Frequently asked questions

Is retention really about pay, deep down?

Not usually. Pay can push someone out if it's genuinely uncompetitive, but raising pay rarely fixes retention on its own. If someone's stopped seeing a future, stopped trusting their manager, or stopped feeling like the work matters, a pay rise buys you a few extra months at best.

How often should we be checking in on retention risk?

Continuously, not annually. An annual engagement survey tells you where things stood on the day people filled it in. The organisations that retain well are watching trends month to month, at the individual level, not just the company average.

We already run engagement surveys. Isn't that enough?

It's a good start, but a survey score alone doesn't tell you who's at risk or why. You need it paired with individual-level signals, like development activity and manager relationship health, and a habit of acting on what you see rather than just reporting it.

What's the single most useful thing we could start doing this month?

Start running stay interviews with your best people, not your struggling ones. You'll learn more from one honest conversation with someone you don't want to lose than from a year of survey data.

So here's the real question. If your best person resigned tomorrow, would you be able to say honestly that you saw it coming, or would you be as surprised as my client was?