If you run a homecare agency, you know the feeling. A letter from CQC, or just the thought of one, and the week changes shape.
We run an agency too, so we went and read what CQC actually writes: the register entry of every active homecare service in England, 7,577 inspection reports, the newest assessments, and every rating going back more than ten years.
One thing came out of it that changes what an owner should spend their worry on: in the reports, low ratings track how the agency is run, not how it cares.
Where your agency sits
Across England, 12.5% of rated services are rated Requires improvement or Inadequate. The Midlands, the East and the South East sit above that; the North, London and the South West below it.
Your council
You know your patch better than any register does. Type your council and see whether the numbers match what you see.
The office gets marked down, not the carers
Every inspection report rates five key questions: safe, effective, caring, responsive and well-led. Look at the 1,017 latest reports with a low overall rating and one pattern is clear.
It runs the other way too. Among Good services with exactly one weak key question, the weak one is well-led in 465 cases and safe in 203. It is caring in two.
The breaches say the same. None of the Outstanding reports record one, 6% of the Good do, 83% of the Requires improvement and all of the Inadequate. When there is a breach, it is nearly always good governance. Then safe care and treatment, staffing, and the checks on new staff.
Numbers only go so far, so we read the reports themselves. The praise repeats: kind staff in about six Good reports in seven, a regular team that turns up on time in about half. The faults are not about the carers either:
| What was faulted | Share of low-rated reports |
|---|---|
| Audits and oversight that did not work | 90% |
| Thin or missing risk assessments | 64% |
| Care plans thin or out of date | 49% |
| Medicines | 49% |
| Recruitment checks | 37% |
| Training and supervision | 33% |
| Late, short or missed calls | 25% |
The same faults, in small and large services alike.
What praise sounds like. "People had a regular team of care staff who arrived on time and knew them well."
What a low rating sounds like. "Systems for oversight and governance were absent, poorly developed, incomplete, or ineffective."
Then there is the worry many small owners have: that CQC is harder on a small agency. The reports say how many people a service supported on the day; the median is 27, and four in ten support 20 or fewer. Size does not predict a low rating.
Our take
CQC does not measure care. It measures the capacity to prove care. In a one-service agency, and nine in ten providers run one service, that capacity is the owner's evenings. Every hour on the audit file is an hour off the rota, while the rating reads as if it were about quality.
A thin file next to a good team is a capacity problem, not a care problem. Capacity can be bought in months. A team cannot. That is how we read an agency, and why we would rather buy a Good team with a weak file than the reverse.
For owners, the order of work is the order inspectors fault things: audits, risk assessments, care plans. We think the daily watching will increasingly be done by software built for one agency, and that is where we are putting our own effort.
Your rating is years old, and so is everyone else's
4,558 homecare services, 34%, have never had a CQC rating. Many are new, but among services registered three to five years ago, 54% still have none. Where a rating exists, it is usually old.
Part of the reason is how few reports CQC has published since 2020. For 96% of services with a published report, that report is now more than three years old.
The new framework has not caught up yet. 2,303 active services have a current assessment under it, 1,153 of them published this year. For 948 of those services, 41%, it is the first rating they have ever had.
Under the new framework 19% of rated services are low, against 14% in the last reports. CQC now chooses which services to assess, sometimes after a concern, so this is partly about who gets looked at.
We also read the new assessment summaries, the short text CQC publishes with each rating. What inspectors praise and fault has not changed: governance in 94% of the low-rated ones, kind staff in 96% of the Good and Outstanding. The method has:
- some assessments were done fully off site, with no visit;
- some joined new scores to key question ratings from the old inspection;
- a Good or Outstanding rating can now sit next to a live breach.
Our take
The register has stopped being a quality signal and become a date stamp. Seven in ten dated ratings are over three years old, a third of services have none, and many of the new ones go to services that never had one. So the fresh ratings skew low and the old ones say nothing about today.
Commissioners, families and buyers will lean on other evidence whether CQC likes it or not. The agencies that publish their own numbers will set the terms; the ones that wait for an inspection will be judged by a photograph.
So keep one file current: call logs, audits, training, complaints and what you did about them. We think most buyers look at the rating first. We ask for that file, and this register is us doing the same thing in public.
After Requires improvement, two in three get back to Good
We put every rating we hold in date order for each service. Among services with two or more ratings, half have had at least one low rating along the way, and of those, 64% are Good or better today.
After Inadequate, 56% rose to Requires improvement within a median of seven months, and a third went straight to Good. After Good, 73% stayed Good at the next rating, a median 3.2 years later, and 22% fell to Requires improvement.
The rates differed by period. Between 2020 and 2023, about one Good service in three fell below Good at the next inspection. If that was you, you were not alone.
One more number belongs here, because anyone can see it on the register. A service run by a company or charity is expected to have a registered manager. About 1 in 12 of those services has none listed, and the low-rated group is 1.6 times as likely to be without one.
Our take
The sector's hidden risk is not quality. It is succession. One service per provider, a median age of nearly six years, one in twelve with no manager listed, and a low well-led rating behind 99 low ratings in 100.
The pattern we see is an owner-manager who steps back, a gap on the register, and a low rating that shows up with it.
The recovery numbers point the same way: two in three climb back. What we take from that is that the team, not the file, is the asset. A buyer who keeps the manager and the team is not being generous; it is the only way to keep what was bought, and it is how we bought A&L.
For an owner, the one question worth answering now is who takes over tomorrow if you are not there. Write that name down. It is the first thing we look at.