Last updated: September 2026 | Sabiha, Talent Acquisition Director
A stay interview is a short, structured conversation with an employee who is still in post, designed to find out what keeps them and what might push them out. UK SMEs run them because they are cheaper, faster and more useful than the alternative, which is finding out the same information on the day someone resigns.
A founder I worked with last year had a spreadsheet of every person who had left in eighteen months. Nine people. She had exit interview notes for seven of them. When we read the notes together, the same two themes came up again and again. No route upward. A manager who never had time.
She had all the information. She had just collected it too late, seven separate times.
That is the argument for the stay interview in one paragraph. Everything below is how you actually run one in a business of fifty to two hundred people, without an HR department of six, and where AI genuinely helps rather than adding another tool nobody opens.
| Scope note: This article covers retention practice and workforce strategy. Where it touches employee data or AI tools, it describes regulatory direction of travel rather than legal instruction. For decisions about processing employee data, consult a qualified UK data protection or employment law specialist. |
A stay interview asks a current employee what would make them leave. An exit interview asks a departing employee why they already decided to. One gives you time to act. The other gives you a record.
The mechanics are simple. A twenty to thirty minute conversation, usually between an employee and their line manager or someone in HR, held on a schedule rather than in response to a crisis. It is not a performance review, and it should not feel like one. Nobody is being scored.
The distinction matters more in a smaller business, where losing one senior person can strip out a whole capability.
| Stay interview | Exit interview | |
| Timing | While the person is still employed and still persuadable | After the resignation has been handed in |
| Question | What would make you consider leaving? | Why are you leaving? |
| Best use | Preventing a resignation you can still prevent | Spotting patterns across departures over time |
| Honesty risk | People soften answers to protect the relationship | People soften answers to protect the reference |
| Cost of acting | Usually a conversation, a schedule change, a development route | Usually a full recruitment cycle |
Exit interviews still earn their place, particularly once AI is used to read across them at pattern level. That is covered in more depth in AI exit interviews and what UK employers miss. Stay interviews sit in front of that process, not instead of it.
Because a business of 51 to 200 people feels a single departure structurally, and because most SMEs have no other mechanism for hearing the truth early.
The CIPD Resourcing and Talent Planning Report 2024 found that 69% of UK organisations reported competition for well-qualified talent had increased over the previous year, and 56% said talent had become harder to retain. Replacing people is not getting easier or cheaper.
There is a more uncomfortable finding underneath that one. CIPD research has previously found that only 17% of organisations calculated the cost of their labour turnover at all, and just 12% collected data specifically to evaluate and improve retention initiatives. Most businesses are losing people without a number attached and without a feedback loop.
Larger employers paper over this with engagement surveys. In a fifty person business an anonymous survey is not anonymous, and everyone in the room knows it. A direct conversation is often the only honest channel available.
The broader picture on what actually holds people in place is set out in employee retention strategies for UK SMEs.
Ask about the day, the future and the exit. Six to eight open questions is enough for a thirty minute conversation, and the follow-up matters more than the question list.
These are the ones that consistently produce usable answers:
That last one is where the useful discomfort lives. Managers rarely ask it because the answer is usually yes.
Keep notes in the employee own words where you can. Paraphrasing at the point of writing is how the signal gets sanded off.
Two to four times a year for most roles, and separately from the appraisal cycle. Annual is too slow, because most of the reasons people leave build over months rather than years.
Three moments are worth adding on top of the standing schedule. Around the six month mark for a new starter, when the honeymoon has worn off and the decision to stay is genuinely live. After any change of line manager. And during restructures, acquisitions or system changes, when the ground is moving under people.
The first of those overlaps heavily with early tenure risk, which is dealt with in the first 90 days and what managers get wrong.
If you are starting from nothing, do not attempt the whole workforce. Start with the ten to fifteen people whose departure would hurt most, and expand once managers have run a few and stopped dreading them.
AI is useful for the pattern layer, not the conversation. It reads across many conversations at once and surfaces themes a manager holding fifteen separate memories cannot see.
Three uses hold up in a small business:
What AI should not do is score people on flight risk and hand managers a ranked list. Predicting who is about to resign, from data employees gave you in confidence, changes the nature of what you asked them for. It also drifts towards the kind of automated profiling the ICO has been actively examining in employment contexts. Aggregate for insight. Do not build a leaderboard of who is about to quit.
The honest version of this is that AI makes the analysis possible at SME scale. It does not make the conversation happen.
Not the questions. The silence afterwards.
The failure pattern is always the same. A round of conversations, a lot of useful feedback, and then nothing visible changes. The second round then produces polite, empty answers, because people have learned what the exercise is worth. You have spent the goodwill and bought nothing.
Two more things break them. Running the conversation with the manager who is the problem, which is common and rarely planned for. Offer an alternative person as standard rather than as an exception. And treating the meeting as a retention negotiation, where the only lever anyone reaches for is money.
Set the expectation openly. Some things will change, some will not, and you will say which is which. Managers can be trained on that in an afternoon.
Track what you did, not just what you heard. Coverage, actions closed, and turnover in the covered group against the rest of the business.
That fourth measure is the one that tells you the truth. If the same themes come back word for word, the conversations are working and the response is not.
Where the answer to a stay interview is a lack of route upward, the practical response is usually structural. That is covered in internal mobility for UK SMEs.
Treat the content as confidential to the people who need it to act, and say so at the start. Themes can be shared upward in aggregate. Individual comments should not travel without the person knowing. Being vague about this is how trust breaks on the first round.
The line manager by default, because they own most of what can actually be fixed. Always offer an alternative, because the manager relationship is one of the more common reasons people leave.
Twenty to thirty minutes. Longer starts to feel like an assessment and people become guarded. Book it separately from the appraisal so it does not inherit that atmosphere.
Yes, and they tend to do it better once they start, because the founder can act on what they hear without a committee. Start with a small group, keep a simple shared log, and hold the ninety day review of commitments.
They answer a different question. A survey tells you the temperature across the business. A stay interview tells you why one specific person is considering their options, which is the thing you can act on this month.
Pick fifteen people. Send a short note explaining what the conversation is and what it is not. Run them over three weeks. Write down every commitment made, with a name and a date against it. Review that list in ninety days.
The conversation is not the intervention. What you do in the eight weeks afterwards is.
| About the author Sabiha is a Talent Acquisition Director, speaker and author with more than 16 years of hiring experience across the UK, Dubai, South Africa and Malaysia. She helps UK organisations move past reactive hiring and build workforce strategies that hold up for the long run, using AI alongside human judgement. Shortlisted as Best Career Coach UK by the CDI, she has helped businesses improve the way they hire and retain talent. |
Global Talent. Ethical AI. Strategic Hiring. Sustainable Retention.
