A caregiver hands in their notice. You already know the shape of the next few weeks: advertising, interviews, induction, and a stretch where the rota is short.

Recruitment is the part that shows up in a budget. The rest of it does not, and it is bigger. This post covers what a leaver really costs you, what the national data says reduces turnover, and the part of it you can change inside your own building.

What a leaver actually costs

  • The gap comes twice. There is the stretch before a replacement starts, and another one after, while they learn the residents, the routines and who to ask. A new starter is not a like-for-like swap on day one.
  • A resident loses a familiar face. For someone with dementia, familiarity is not a nice extra. It is part of the care, and it does not transfer with the handover notes.
  • The team absorbs the difference. The people who stay pick up the extra shifts, which is how one leaver turns into two.

What the data says moves it

Skills for Care published its latest workforce report in October 2026. Three of its findings are worth knowing.

  • Turnover more than halves when five things line up. Care worker turnover falls from 34.3% to 15.3% when an employer offers competitive pay, guaranteed hours, training, qualifications and full-time work. Not one of them on its own. All five.
  • Experienced managers keep people longer. Care worker turnover falls from 28.3% to 20.6% as registered manager experience increases.
  • Under-25s leave most. Turnover in that age group is 35.5%, the highest of any.

Read that list honestly and most of it is pay, contracts and hours. Those are decided by your fee rates and your budget, not by any software you buy. Anyone selling you a retention tool should say that first.

The part that sits inside your building

What is left is the admin. The work that is not care and not pay, but still takes up the shift.

The second one matters more than it looks. The tasks that wear people down are not always the long ones. They are the ones that arrive without warning, land on whoever is nearest, and have to be sorted out by phone.

Some of that you can remove. Notes written during the shift instead of from memory at the end of it. Records someone can find without walking to the office. Paperwork that is not entered twice. None of that is dramatic, and none of it will outweigh an uncompetitive hourly rate. It is still hours handed back to the people you are trying to keep.

One honest caveat about what software cannot do here. Granily does not tell you who is about to resign. We have no feature that watches for it, and we would be sceptical of one that claimed to. Noticing that someone has gone quiet at handover, or stopped picking up extra shifts, is a manager's job and stays that way.

Where that leaves you

You cannot buy your way out of the pay problem, and this post is not going to pretend otherwise. What the data does say is that the things which keep people are mostly ordinary: steady hours, training that happens, and a manager with enough experience to run the place calmly.

The admin load sits in that same category. It will not keep a caregiver who has been offered two pounds an hour more down the road. It might keep the one who is tired of spending the last forty minutes of every shift catching up on notes.