Employee Retention: How to Build a Retention Strategy That Actually Holds

Too often, retention only becomes a conversation when someone resigns. By then, most of the moments that shaped their decision have already happened. The counter-offer may buy time, but it rarely solves the problem. Real retention starts much earlier, in the everyday experience people have of your business, their manager, their role, and their future.

The commercial case is clear. When a valued person leaves, the cost is not just the recruitment fee. It is the lost productivity while the role is open, the pressure on the people covering the work, the manager time spent hiring, and the months it takes for someone new to get fully up to speed. Most businesses feel that cost, but very few name it for what it is: a retention issue that could often have been spotted sooner.

There is also the human and cultural cost, which never quite makes it onto the spreadsheet. When a well-liked colleague leaves, especially if the exit feels sudden or poorly handled, others start to ask their own questions. Was there something they knew? Should I be looking too? One avoidable resignation can create a ripple effect months later. That is why retention is not just about keeping people; it is about understanding what makes them stay, and acting before the decision has already been made.

What is employee retention, and how is it different from low employee turnover?

Employee retention is your ability to keep the people you genuinely want to keep for as long as the business needs them. Employee turnover is the rate at which people leave overall, whether that is wanted or unwanted, voluntary or involuntary. The two are often used interchangeably, but they answer different questions. A business can have what looks like healthy turnover and still be carrying a serious retention risk.

That distinction is the one that needs to be fixed first. A 12% turnover rate looks fine on a board pack. If eight of those twelve points are your strongest performers and your two most experienced managers, you do not have a healthy business with normal churn. You have a retention failure hidden inside an average.

The CIPD's guidance on employee retention makes the same separation, and it is worth reading alongside anything you build internally. What matters commercially is not how many people left. It is who, from where, and whether you could have known sooner.

Why does staff retention matter more in a founder-led business?

In a business of thirty, one person leaving takes with them a percentage of your organisational knowledge that no handover document captures. In a business of three thousand, that person is a line in a report.

There is a second issue that is particularly true in founder-led companies. Your best people are often doing jobs that have grown around them, rather than around a clearly defined role. When they leave, you are not simply recruiting a replacement. You are rebuilding the role, the rhythm of the team, and the knowledge that sat with that person. That takes time, and often far longer than expected.

And then there is the harder truth. In smaller businesses, people often leave because of the experience they are having with individuals, and those individuals may be close to the founder, or part of the original team. That can make the issue uncomfortable to name, but avoiding it only makes the retention problem more expensive.

The cost you can actually put a number on

Before you build any retention strategy, work out what one avoidable leaver costs you. Include the recruitment fee, the productivity gap while the role is open, the pressure on the people covering the work, and the time it takes before the replacement is fully effective.

Most founders who do this find the real cost lands somewhere between six and twelve months of the departing person’s salary once everything is counted. I would not rely too heavily on a national average, because the honest numbers vary significantly by sector, seniority, and role complexity. The important thing is to work out yours.

Then set it against what a retention intervention costs. A salary benchmarking exercise, a round of stay interviews, or a set of management workshops will almost always come in under the cost of one avoidable exit. That is the entire commercial argument, and it holds in every business we have run it in.

How do you measure employee turnover properly?

You cannot manage retention from a feeling. You also cannot manage it from a single headline percentage, which is where many businesses stop. The useful insight is usually in the detail underneath the number.

The turnover rate, and why it flatters you

The standard calculation, as set out in CIPD guidance, is the number of leavers in a period divided by the average number of people employed in that period, multiplied by 100.

It is a useful baseline and a poor diagnostic. It treats a resignation from someone you were about to promote as identical to an exit you were relieved about. It also hides tenure, so a business that keeps losing people at eighteen months looks the same as one losing people at six years.

The stability index, which is the number worth watching

CIPD's stability index is the number of employees with one year's service or more, divided by the number of employees a year ago, multiplied by 100.

This one tells you whether you are keeping the people who have learned how your business works. It is the number that moves first when something is going wrong, and it is the one most small businesses have never calculated.

The five cuts worth running on your own data

Turnover by manager. Turnover by tenure band, especially the first twelve months. Turnover by performance rating, if you have one.

Then turnover by pay quartile, and voluntary against involuntary, separated properly.

Run those five and the answer is usually sitting in one of them. In founder-led businesses, it is often the first. That can be difficult, because the manager in question may be someone who has been central to the business from the beginning, but that is exactly why the conversation matters.

What our own data told us

Our internal engagement survey ran at 86% overall, with the whole team taking part.

Those are strong scores, but the value of an engagement survey is not only in the percentage. It is in what people write when you ask what would make the business better, and that is often the part that gets lost in the summary slide. If you are not running an employee engagement survey yet, start there, and read the free text before you read the percentages.

Five things that make people stay

The question comes up consistently. In practice, five themes account for most retention: good management, trust and autonomy, clear progression and pay, flexibility, and a sustainable workload.

1. The manager, before anything else

Gallup's research on manager impact found that around 70% of the variance in team engagement is attributable to the manager. Every retention intervention you run sits downstream of that figure.

Where this tends to go wrong in growing businesses is not usually because people are deliberately managing badly. It is accidental management. Someone brilliant at the work gets four people to look after, nobody makes clear that this is now a different job, and no one trains them for it. Then the business is surprised when those people become quieter, less confident, or start looking elsewhere.

The fix is unglamorous, and it works. Give managers a structure for one-to-ones, a process for probation and performance that they did not have to invent, and training in the two conversations they dread. That is most of what our Learning + Development workshops exist to do, and it is consistently the highest return retention work we are asked for.

2. Trust and autonomy: the Timpson example

Timpson is the example that bears repeating because it is British, it is not a technology company, and the retention effect is a consequence of the operating model rather than a benefits package.

The business runs what it calls upside-down management. Colleagues in shops have real authority over pricing, service decisions, and how they run their day, while the head office function exists to support rather than instruct. Alongside that, Timpson has recruited from prisons for years and has been open about how much of its workforce comes through that route.

What it demonstrates is that trust has to be structural, not just stated. It shows up in what people are allowed to decide without asking, which is very different from a values poster. If you want to know whether your own culture is real, look at the last five decisions someone junior had to escalate and ask whether they genuinely needed to.

3. Progression and pay clarity

The most common thing that surfaces in exit conversations is not that someone was underpaid. It is that they could not see how pay and progression worked, so they assumed the worst, and the only way to test the assumption was to interview elsewhere.

This is fixable without spending money. Write down what each level does, what it pays as a band, and what moving up requires. Career development pathways and honest salary benchmarking between them resolve more retention risk than most benefits budgets.

The businesses that handle this well tend to publish the framework before anyone has to ask for it. The ones that struggle are often trying to create clarity in a hurry, after a resignation, for an audience of one.

Case study: Tillman v Egon Zehnder, and why you cannot contract your way out of it

Mary-Caroline Tillman's employment contract with Egon Zehnder contained a six-month non-compete preventing her from being "concerned or interested in" a competing business. When she left to join a competitor, the enforceability of that clause reached the Supreme Court.

In Tillman v Egon Zehnder Ltd [2019] UKSC 32, the court accepted that the words "interested in" were too wide, because on their face they would have prevented her holding even a minor shareholding. The clause survived only because those specific words could be severed and the rest of the restraint upheld.

It is worth raising for a reason that has nothing to do with drafting. Litigation of this kind is what retention failure looks like at the end. By the time you are relying on a restrictive covenant, you have already lost the person, the relationship, and usually the client knowledge, and you are spending money to limit the damage.

Covenants are a sensible last line, and they should be reviewed, because the drafting has to be proportionate to survive. They are not a retention strategy. If yours have not been looked at since the template was written, that belongs in an HR Health Check rather than in a crisis.

4. Flexibility and consistent management

Since 6 April 2024, under the Employment Relations (Flexible Working) Act 2023, the right to request flexible working has been a day-one right. Employees can make two requests in any twelve-month period; employers must consult before refusing, and a decision is due within two months. Acas guidance on flexible working requests sets out the process.

The Carer's Leave Act 2023 also took effect on 6 April 2024, giving a day-one entitlement to one week of unpaid leave a year for employees caring for a dependant.

Both matter for retention beyond compliance. People rarely leave because of the wording of a flexibility policy. They leave when they discover the policy is applied differently depending on who manages them. That is not really a policy issue; it is a management consistency issue showing up through the policy.

5. Workload: the difference between busy and burnt out

Burnout is not the same as being under pressure, and the distinction is worth holding because businesses respond to the wrong one. The Maslach Burnout Inventory, the most widely used research instrument in the field, measures three separate dimensions: emotional exhaustion, depersonalisation or cynicism about the work, and a reduced sense of personal accomplishment.

The second and third are the ones that predict a resignation, and they are the ones a founder is least likely to notice. Someone who is exhausted still cares. Someone who has stopped caring about work they used to be proud of has usually already started looking.

What to watch for is not only hours. It is a change in how someone talks about the work, particularly someone who used to care enough to challenge, debate, or push for better, and has now stopped.

How to build a retention strategy in ninety days

You do not need a complicated programme. You need five things done in the right order. The order matters because the first two steps tell you where to put your effort, rather than spreading attention thinly across things that may not move the dial.

Step one: get the numbers on one page

Turnover rate, stability index, and the five cuts above. Twelve months of data minimum, twenty-four if you have it.

Do not interpret anything yet. Most businesses find the pattern is more concentrated than they expected, which is good news, because concentrated problems are cheaper to fix than cultural ones.

Step two: ask the people who have not left

Stay interviews, not exit interviews. Exit interviews tell you what someone is willing to say once the decision is irreversible and they want a reference. Stay interviews tell you what is fixable while it still is.

Four questions will usually do it. What would make you think about leaving? What is the best part of your week? What has changed in the last six months? And what would you change if this were your business?

Ask them somewhere that is not a meeting room, and do not take the answers to a leadership meeting with names attached.

Step three: fix the manager layer before the benefits

If the data points at specific managers, and it usually does, that is where the money goes. Coaching for the individual, structure for the process, and honesty about whether the person wants to manage at all.

Some of the best outcomes come from letting a brilliant senior specialist stop managing people and stay senior. That requires a career framework with two tracks, which is not a huge piece of work, but it can save people you might otherwise lose twice over: first from the management role, and then from the business altogether.

Step four: make pay and progression visible

Bands, levels, and criteria written down and shared. If your bands are not defensible, benchmark them before you publish, because publishing an indefensible structure is worse than having no structure at all.

Step five: choose three priorities and assign ownership

Choose no more than three retention priorities for the next six months. Give each one a named owner, a deadline, and a clear measure of success.

Then review the same turnover and stability data from step one after six months. A focused plan that is completed will do more than a long list of initiatives that never make it into practice.

How JourneyHR can help

JourneyHR has worked with founder-led UK businesses since 2010, and retention is often the reason they call us for the first time. Usually it is after the second or third resignation, which is later than ideal, but still early enough to fix.

Our Retained HR Team model gives you HR Director input on the strategic decisions and consultants and assistants to deliver everything else, for less than the cost of a standalone in-house hire. Every retained relationship starts with an HR Health Check and an employee engagement survey, so we begin with your actual data rather than an assumption about it.

If you want to start smaller, we run stay interviews, salary benchmarking, career framework design, and management training as standalone projects. If you would rather have senior capacity without a full-time hire, our fractional HR Directors work on a part-time retained basis.

Get in touch: book a call orcontact us

Final thought

The businesses that keep their people are rarely doing anything dramatic. They get a handful of small things right, consistently, including in the months when it would be easier to let them slip.

The change that makes the difference is almost never a new benefit. It is asking the right question earlier, while someone is still deciding whether to stay, rather than after they have already decided to leave.

FAQs

What is a good employee turnover rate for a small UK business?

There is no single benchmark that works across sectors, which is why comparing your figure to a national average is less useful than tracking your own trend. Look at your stability index alongside your turnover rate, and separate voluntary from involuntary leavers. A rising rate among people with one to three years' service is the pattern worth acting on, whatever the headline number says.

How do you calculate an employee turnover rate?

Divide the number of leavers in a period by the average number of people employed during that period, then multiply by 100. CIPD also recommends the stability index, which is the number of employees with a year or more of service divided by the number employed a year ago, multiplied by 100. Run both. The first shows volume; the second shows whether experience is staying.

Do stay interviews actually work, or do people just say what you want to hear?

Stay interviews work when three conditions are met: someone other than the person's direct manager runs them, answers are reported as themes rather than attributed quotes, and at least one visible change follows within a quarter. Without that third condition, they stop working immediately, because people conclude that nothing comes of speaking up and revert to saying nothing.

Is pay the main reason people leave a small business?

Pay is the reason people give more often than it is the reason they go. In most exit conversations, the underlying issue is that progression and pay criteria were not visible, so the only way to test whether they were being treated fairly was to interview elsewhere. Publishing levels, bands, and criteria resolves more retention risk than an unplanned salary increase does.

How long does it take to see a retention strategy work?

Expect leading indicators within one to two quarters and a change in your stability index over twelve months. Engagement scores, one-to-one completion rates, and internal promotion numbers move first. Turnover moves last, because the people who were already thinking about leaving when you started will mostly still leave. Judge the work on the cohort that came after it.