You’ve done the hard part.
You wrote the job ad. You sifted the applications. You interviewed, you chased references, you processed the DBS, you negotiated the start date. You got someone through the door.
And then, somewhere between week three and month three, they hand their notice in.
Or they stop turning up for shifts. Or they go quiet. Or they stay on the rota but you can tell they’re already mentally somewhere else.
This edition is about the 90-day window. The period most care providers almost completely ignore. The period that determines, more than any other, whether the money, time, and energy spent on hiring someone turns into a return or another cycle of the same.
Spoiler: the hiring process ends on day one. What happens next is a different process entirely. Most care providers don’t have one.
The numbers on early attrition
The sector’s turnover rate sits at 31%, nearly double the UK average of 15% across all employment. That headline figure gets quoted a lot. What gets discussed less is when that turnover is happening.
Industry data consistently points to the same finding: a disproportionate share of care worker leavers go within the first three months. Some research puts it as high as one in three of all leavers departing within their first 90 days.
Think about what that means in cost terms. Using the JRF’s figures from earlier this year, median recruitment cost £800, training cost £884, agency cover cost £3,683 per leaver, someone who leaves in week six has cost you everything it took to hire them and most of what it cost to cover the gap while they were there. And contributed relatively little in terms of experience, stability, or continuity of care.
The maths of early attrition.
Staff member leaves in first 90 days. You’ve spent ~£1,684 recruiting and training them. You’ve spent ~£3,683 on agency cover for the gap they’ve now recreated. You’re back to where you started. Minus £5,367. And the rota still needs filling.
The sector talks about retention as a long-term problem. But for most providers, the biggest retention leak is right at the start, in the period when a new starter is deciding, consciously or not, whether they made the right choice.
What actually drives early leavers
Research into early attrition in care consistently identifies the same patterns. None of them are surprising. Most of them are fixable.
Nobody checked in.
The first few weeks in a care role are genuinely hard. New environment, new service users, new team dynamics, new processes. For someone who’s never worked in care before, and the sector recruits a significant number of people making a career change, it can feel overwhelming. A structured check-in at the end of week one and week four isn’t pastoral care for its own sake. It’s information gathering. It tells you whether someone is struggling before they decide the struggle isn’t worth it.
The rota changed.
One of the most cited reasons for early departure in the sector: the hours or patterns changed from what was discussed at interview. Care rotas are inherently unpredictable. But if someone accepted a role on the basis of certain shift patterns and those patterns shifted significantly within the first month, the implicit agreement has been broken. That erodes trust fast.
They never felt like they belonged.
This one is harder to systematise but it matters enormously. Care work is relationship-based, with service users and with colleagues. New starters who are left to find their own feet socially within a team, without any deliberate integration effort, often don’t. They stay on the periphery. And peripheral people leave.
The job wasn’t what they expected.
This one cuts both ways. Sometimes the role is genuinely misrepresented at recruitment stage, and that’s a job ad problem or an interview problem. But sometimes the issue is that nobody prepared the new starter for the realities of the work: the emotional weight of it, the pace of it, the complexity of some of the service users’ needs. An honest, practical induction that prepares people for what the job is actually like dramatically reduces the ‘this isn’t what I signed up for’ resignations.
The Employment Rights Act: what changes from April 2026
A quick update on the Employment Rights Act 2025 that’s directly relevant to early-stage employment.
From April 2026, statutory sick pay applies from day one of sickness. The previous three-day waiting period is gone. The lower earnings threshold of £123 per week is also removed, meaning more part-time and variable-hours workers qualify.
For care providers with high proportions of part-time staff, this will increase the financial impact of sickness absence. It also means absence tracking needs to be more rigorous, not to police people, but because the data matters more when every day of absence has a cost attached.
Also from April 2026: day-one rights for paternity leave and unpaid parental leave. Employees no longer need a qualifying period. This has less dramatic immediate cost implications but it requires managers to know what their obligations are from the moment someone starts, not after they’ve passed probation.
Practical implication for onboarding: your induction process needs to cover employment rights from day one, not because it’s legally required to, but because an employee who understands their rights is less likely to feel that those rights are being ignored.
Informed employees are more engaged employees. Engage them early.
What a 90-day onboarding process actually looks like
Most care providers have an induction. Most of those inductions are a day or two of mandatory training, a manual, and a shadow shift. That’s compliance, not onboarding.
Onboarding is the planned, structured experience of integrating someone into the team, the service, and the culture in a way that makes them more likely to stay. It doesn’t need to be elaborate. It needs to be intentional.
Week one: practical and relational.
- Buddy system, paired with an experienced team member for the first two weeks.
- End of week one check-in, 15 minutes, informal, ‘how’s it going’ not ‘are you compliant’.
- Clear expectations set: what does good look like in this role, in this service.
Weeks 2 to 4: embedding.
- First formal one-to-one with their line manager, not a performance review, a conversation.
- Introduction to any service-specific procedures not covered in induction.
- Check on rota patterns, are they working what was discussed.
Month 2: commitment check.
- Structured review conversation: what’s working, what isn’t, any concerns from either side.
- Development conversation: what does progression look like, any qualifications they want to pursue.
- Ask directly: ‘Is this what you expected?’
Month 3: retention signal.
- End of 90-day review, formal record, two-way feedback.
- Establish what’s next: any training, any changes, any commitments from the employer side.
- The question that matters: ‘What would make you more likely to stay long-term?’
None of this requires a dedicated HR function. It requires a system, a checklist, a calendar reminder, a record of what was discussed, and a manager who has the time and headspace to do it. That last part is often the real problem: managers who are covering shifts, chasing references, and reposting job ads can’t give new starters the attention that keeps them.
Fixing recruitment admin frees up the time that onboarding needs. They’re connected problems.
This fortnight’s exercise
Think about the last three people who left your service within the first 90 days.
For each one, answer three questions:
- Did someone have a structured check-in conversation with them in the first two weeks?
- Did they work the shift patterns they were told about at interview?
- Did anyone ask them, before they handed in their notice, if something wasn’t working?
If the answer to most of those is no, you’ve got a process gap, not a people problem. Process gaps are fixable.
Before you go
The sector spends enormous time, energy, and money on getting people in. Very little of that investment is protected by anything that happens after day one.
The 90-day window is where retention is won or lost. Not in the job ad. Not in the interview. In the weeks after someone starts, when they’re deciding, usually without telling anyone, whether they made the right choice.
Build the process. Give them a reason to stay.