Let’s talk about the thing everyone in the sector quietly knows is bleeding them dry but somehow keeps doing anyway.
Agency staff.
Not because agencies are evil. They’re not. They fill gaps. They show up on a Sunday morning when your rota falls apart at 11pm on a Saturday. They serve a purpose.
But somewhere along the way, for a lot of care providers, agency stopped being the emergency plan and became the actual plan. And that shift, from backup to default, is costing the sector an extraordinary amount of money. Money that could be paying your permanent staff better, funding training, or just keeping the lights on with a bit less dread.
This edition is about the agency habit. How it forms, what it actually costs, and, without being preachy about it, what the alternative looks like.
The numbers, and they’re worse than you think
New research from the Joseph Rowntree Foundation published this year finally put proper numbers on something care providers have felt for a long time but rarely quantified.
The median agency cost per departing care worker: £3,683. Not per month. Per person who leaves.
Add the median recruitment cost of £800 and training cost of £884 for the person who replaces them, and you’re looking at over £5,300 every time someone walks out the door.
For a provider replacing 10 people a year, which is modest, given the sector’s 31% turnover rate, that’s £53,000. Not in agency markup alone. In the total cost of that churn cycling through your service.
Quick maths.
10 leavers x £5,300 average replacement cost = £53,000 per year. 20 leavers x £5,300 = £106,000 per year.
Most providers have no idea this number exists on their books because it’s spread across invoices, manager time, and payroll in ways that never appear in one place.
The reason the agency bill stays high isn’t complicated. When someone leaves, the rota has a gap. The gap needs filling today. The agency answers. The invoice arrives. Repeat.
What doesn’t happen, because there’s no time and no system for it, is anyone asking why that person left in the first place, or whether the process that hired them gave them any real reason to stay.
The domestic pool just got smaller. Again.
If your agency habit was already uncomfortable to read about, here’s the part that makes it worse.
The pool of people you’re competing to hire from just got significantly smaller.
The Health and Care Worker visa ban came into full effect in July 2025, ending international recruitment for care worker roles. The sector had been leaning on overseas hires heavily. By 2024/25, nearly one in four workers in adult social care came from outside the UK and EEA, up from one in twelve in 2019. That source is now effectively gone for new entrants.
At the same time, according to Skills for Care data, there were 30,000 fewer British workers in the sector in 2024/25 compared to the year before.
So: more demand for care, fewer domestic workers, no international pipeline, and 131,000 vacancies still sitting there according to Care England.
The government’s answer, a Fair Pay Agreement that might deliver pay increases in 2028, implemented by a Negotiating Body being established in October 2026, is, to put it charitably, not a short-term solution.
What this means in practice:
- Every care provider in your area is now fishing in the same smaller pond.
- The ones who win are not the ones spending the most on job boards.
- They’re the ones whose hiring process is fast enough, warm enough, and organised enough to convert candidates before someone else does.
The Employment Rights Act: the zero-hours piece you need to know about
The Employment Rights Act 2025 received Royal Assent in December and implementation is rolling out through 2026 and 2027. There’s a lot in it, but for care providers specifically, one area demands attention now.
Zero-hours contracts.
Around 21% of posts in adult social care were on zero-hours contracts in 2023/24, compared to 3.5% in the wider economy. The Act doesn’t ban zero-hours working outright, but it changes the rules significantly.
From January 2026, zero-hours workers cannot be restricted from taking work elsewhere. Any practice, written or informal, that penalises someone for picking up shifts at another provider creates legal risk. That includes cutting their hours after finding out they’ve accepted work somewhere else.
Later in 2026 and into 2027, workers on variable hours will gain the right to request a guaranteed-hours contract based on their average hours over a 12-week reference period. Employers will also need to give reasonable notice of shifts and pay compensation for cancellations at short notice.
For care providers, this has two implications worth thinking through now rather than in a hurry later.
- Your bank staff and zero-hours workers may start requesting guaranteed contracts. If a significant portion of your flexible workforce does this, your rota planning assumptions need revisiting.
- If your flexible staffing model relies on the expectation of exclusivity, that your bank workers only work for you, that expectation is now legally unenforceable.
None of this is a crisis. It’s a process question. What does your current zero-hours and bank staff management look like? If the honest answer is ‘a spreadsheet and a WhatsApp group’, the Employment Rights Act is a reasonable prompt to tighten that up.
Breaking the agency habit in three steps
Nobody breaks a default pattern overnight. But there are three practical moves that shift the dynamic, and none of them require a large budget or a transformation project.
Step one: find out where your agency spend is actually coming from.
Most providers don’t have a clean answer to this. Is it the same roles repeatedly? The same service? The same shift patterns? The answer shapes everything that comes next. Pull your last three months of agency invoices and map which roles, which days, and which services are driving the majority of the spend. You’ll almost certainly find it’s not evenly distributed.
Step two: look at what happened before the gap appeared.
For the roles you’re regularly covering with agency, look back at why those positions became vacant. Was it turnover? Sickness? Rota gaps from not having enough contracted hours? Each has a different fix, and treating them all the same is why the spend keeps recurring.
Step three: build a bank before you need it.
The providers who use agency least tend to have a warm pool of people who’ve worked for them before, done their DBS, know the service, and are willing to pick up shifts. Building that bank takes time and a basic system to manage it, but it costs a fraction of what agency does, and the people in it already know how you work.
This fortnight’s exercise
Pull your last three months of agency invoices. Total them. Then divide by the number of individual shifts covered.
Now calculate what that shift would have cost with a bank worker on your standard pay rate.
The difference between those two numbers is your agency premium. For most care providers it sits somewhere between 40% and 120% above standard cost.
Write that number down somewhere visible. Not to feel bad about it, to give yourself a concrete reason to change it.
Before you go
The agency habit is understandable. It exists because it works, in the narrow sense of filling a shift on short notice. But it keeps working because the alternative, a functioning direct recruitment process, never quite gets built.
The sector can’t wait for the government’s 2028 pay agreements to make domestic recruitment easier. The providers who will be in the best shape when those changes arrive are the ones who started building proper processes now.