Executive summary
- 86.8% of care home beds in England were occupied in mid-August 2026, according to the government's Capacity Tracker. Nobody publishes the matching figure that would tell an owner whether that's enough: the occupancy at which a home breaks even.
- The line differs by home because the payer mix does. Councils paid an average of £956 a week for residential care in 2025/26. Self-funders are 45% of residents but 55% of the market's value, so a home's line depends on who is in the beds, not just how many.
- In the worked example below, a six-home group is 88.0% full and 54 residents above its combined break-even. Two of its homes hold five of those between them. Each is two or three residents away from a loss.
- April 2026's fee rises came mostly from self-funders. In the example they are 42% of residents and paid 62% of the increase. At the most council-funded home, the uplift covered April's extra costs with £30 a week to spare.
Healthcare at DataHexis
"Patient flow, capacity planning, clinical KPIs, and CQC compliance reporting." This piece is about capacity planning where it decides whether a business survives: the independent care home groups that provide almost all of the UK's residential care for older people.
— from the Industries section of the DataHexis websiteAsk the owner of a care group how full the homes are and you'll get a number to one decimal place, often from memory. Ask how many residents each home needs to cover its costs this month, and the answer is usually a range, a rule of thumb, or a promise to ask the accountant.
That gap matters more than it used to. Pay is well over half of a care home's income, the National Living Wage rises every April, and council fees rarely move at the same speed. A home can stay exactly as full as last year and end up with less to show for it. The occupancy figure won't warn you. The line underneath it will.
88% full of what?
Occupancy sounds like the one care home measure everybody agrees on. It isn't, because it has two denominators. Registered beds are what the CQC registration allows. Available beds are what can actually be sold this week, after rooms closed for refurbishment, doubles used as singles and beds held back for staffing.
The published figures come from different samples and don't always say which denominator they use; Grant Thornton's 2026 report labels its supply chart "occupancy as % registered beds". Inside a group the same thing happens. One manager reports against registered beds because that's on the certificate, another against available beds because that's what the rota is built for, and the group total quietly mixes the two.
In the worked example below, the group is 88.0% full against available beds and 84.6% against registered ones. Both are true. Only one of them can be compared with a break-even line, because costs are set by the beds a home is staffed and run for.
Occupancy tells you how full a home is. Beds above the line tell you whether that's enough.
The line nobody prints
Break-even occupancy = the home's fixed weekly costs ÷ (average weekly fee − the cost that rises with each resident), shown as a share of available beds. Fixed costs are mostly the staffing rota, plus running costs and rent. The per-resident cost is food, continence products, laundry and other consumables.
Beds above the line = residents today minus break-even residents. It's the cushion a home has before the next death, hospital stay or delayed admission takes it into a loss, counted in people rather than percentage points.
Sector reports publish occupancy, fees, staff costs and margins in detail. None of the benchmark reports we checked publishes a break-even occupancy. The 80–85% often quoted in conversation is a rule of thumb, and it can't be right for every home, because the inputs aren't the same for every home.
The biggest difference is who pays. The government's provisional figures for 2025/26 put the average council fee at £956 a week for older people's residential care, and £1,089 with nursing (excluding the NHS-funded nursing contribution). LaingBuisson estimates that self-funders now make up 45% of residents but 55% of the market's value, and that the gap between private and council fees has narrowed from around 40% in the mid-2010s to about 25%.
Size counts as well. Grant Thornton's 2026 report puts EBITDARM at 22.6% for homes of 1–39 beds, against around 32.6% for homes over 60. A small home carries a manager, a cook and a night team whether it has 30 residents or 36, so its line sits higher and each empty bed costs it more.
Two homes at 85% can be ten beds apart. One of them is having a good year.
April redraws it
Every April the inputs move. The National Living Wage rose 4.1% to £12.71 an hour on 1 April 2026, after a 6.7% rise the year before. Employers' National Insurance went from 13.8% to 15% in April 2025, with the threshold cut from £9,100 to £5,000 a year. The NHS-funded nursing care rate rose 5.4% to £267.68 a week for 2026/27. Council fee uplifts are set locally, and some are confirmed weeks after the pay rise has gone through.
The two staff-cost benchmarks disagree because they measure different businesses: one mainly large corporate groups, the other the wider independent sector. A six-home group will usually sit somewhere between them, and the spread between its own homes is often wider than the spread between the benchmarks.
What April does to the line depends on the payer mix. If a home's fees rise faster than its costs, the line falls. If the pay bill rises 5% and the council uplift is 3%, a home funded mostly by its council sees the line rise, even while the group as a whole improves. That's why a group-level April bridge can look comfortable while one home's cushion gets thinner.
A group can have a good April while one of its homes has a bad one. Only a home-level line shows which.
Agency spend arrives late
Agency staff were the cost story of 2022 and 2023. Both main benchmark reports now put agency at around 3% of staff costs, down from around 7% a year earlier. Skills for Care puts the sector's vacancy rate at 6.2% for 2025/26, the lowest since 2015/16, but international recruitment fell to 30,000, its lowest in four years. Grant Thornton expects agency use to rise again in the medium term, now that care workers have come off the shortage occupation list.
The reporting problem is timing. An agency shift is booked by a home manager at short notice, worked that night, and invoiced weeks later. If agency spend is only visible in the management accounts, the board sees each month's problem about a month after the rota could have fixed it. The shift record knows on the day; the invoice is the last place to find out.
An agency shift costs the most on the day it's booked, and gets reported a month later.
A worked example: the group occupancy view
Here's what the reporting looks like for an independent group of six homes in the Midlands: two nursing homes and four residential homes, 312 registered beds, 300 available and 264 residents. It's a composite of typical group structures rather than a single client. The figures are illustrative, but they're calibrated against the benchmarks cited throughout, and every number reconciles to the others.
The conventions are stated rather than assumed. Occupancy is against available beds unless it says otherwise. Fees are gross weekly fees, including NHS-funded nursing care and any top-ups. "Home EBITDA" is after rent and before central costs. The April 2026 changes are assumptions for the example: the pay bill up 5.0%, other costs up 3.0%, self-funder fees up 6.0%, council fees up 3.0% and NHS-funded fees up 4.5%. At today's rates the group takes £17.69m a year in fees, at an average weekly fee of £1,285.
| Line | % of fees | Note |
|---|---|---|
| Staff costs | 57.2% | Pay, on-costs and agency, by home; from 53.6% at the largest nursing home to 65.5% at the smallest |
| Running costs | 12.5% | Utilities, maintenance, insurance, rates, home administration |
| Per-resident costs | 8.1% | Food, continence and medical consumables, laundry |
| Rent | 2.9% | Two of the six homes are leased |
| Home EBITDA | 19.4% | £3.42m; before central costs of 5.5%, leaving 13.9% for the group |
Five sources feed the model:
The operations director opens this view every Monday; the managing director takes the same model to the bank and the board each month. Both read the same figures, so the bank and the board never see two versions of a home.
Group occupancy view: six homes
Week commencing 12 October 2026 · refreshed nightlyOccupancy keeps its place in the row, but beds above the line sits next to it: 54 sounds comfortable until you see where they are. Each status carries a word as well as a colour, so the row still reads when printed in greyscale.
Forecast beds above the line, next seven weeks
Built from what's already known: confirmed admissions, assessments booked, residents in hospital and notice given. Orchard View dips below its line in the week of 2 November unless an assessment converts. The lightest cell on the grid is the one to act on, three weeks before it happens.
Occupancy against headroom: where each home sits
Riverside, Church Lane, Mill House. Protect the fee levels and the staff who hold them.
The Beeches. Plenty of headroom; every admission now is mostly margin.
No home today. Selling beds won't fix this box: only fees, payer mix or the rota will.
Orchard View, Westgate. Two or three residents from a loss. Enquiries and fees reviewed weekly.
The empty box is the one to watch. A council-funded home that fills up after a block contract can land there, and its occupancy figure will look like good news.
The board view: five charts for the monthly pack
Home managers live in the weekly view. The board, and increasingly the lender, needs to see which homes carry the group, what April did to each of them, whether agency is under control and where the fee income actually comes from. Same model, fewer rows.
Break-even and actual occupancy, by home
Riverside breaks even at 64.9% and runs at 92.6%. Orchard View needs 79.0% and runs at 84.1%. The group's 88.0% describes neither of them. Put the line on the same axis as occupancy and the board stops asking how full the homes are, and starts asking how safe they are.
Agency share of the pay bill, by month
From 7.6% to 4.2% in a year. The fall started when each home's unfilled shifts became visible on the day, not when the invoices arrived. At this group's pay bill, each percentage point is about £100k a year paid through agencies.
What April 2026 did to home EBITDA
The group comes out £0.18m better off. Self-funders are 42% of residents and paid £0.48m of the £0.77m fee rise, almost exactly the cost of the pay rise. At Westgate, where two-thirds of residents are council-funded, the uplift covered April's extra costs with £30 a week to spare, and its line moved up.
Average weekly fee, by who pays
The same nursing bed earns £1,732 from a self-funder and £1,363 from the council, including the NHS nursing contribution. A home's line is set as much by this chart as by its rota.
Where each £1 of fees goes
57p goes on staff, which is why the line moves when pay does. Most of it is fixed by the rota rather than by the number of residents, so an empty bed saves only about 8p in the pound.
Under the hood (for the data team)
The charts are the easy half. The work is getting a care system, a billing system, a rota and a set of accounts to agree on what counts as a bed, a resident week and a home's costs.
Six tables and the bed-nightDated bed counts, fee periods and why the denominator has a history
Everything lands in one small warehouse before Power BI sees it. Whichever care and rota systems each home uses, they end up in the same six tables:
| Table | Grain | Why it matters |
|---|---|---|
| Fact Bed Night | Room × night | Occupied, vacant, held for hospital stay, closed; the only grain that gets occupancy right mid-week |
| Fact Fee | Resident × payer × week | The rate actually charged, split between council, NHS, self-funder and top-up |
| Fact Shift | Shift × home × role | Planned, worked and agency-filled, with hours and cost rate |
| Fact Cost | Home × account × month | Running costs and rent from the accounts, tagged fixed or per-resident |
| Dim Home | Home, with bed counts by date | Registered and available beds as at each night, so a closed room changes the denominator from the day it closed |
| Dim Resident | Pseudonymised resident, payer history | Funding changes mid-stay (self-funder to council, FNC added), so payer is dated |
The join that breaks first is the denominator. Available beds change when a room closes for a refit or a double becomes a single, and if the bed count isn't dated, last month's occupancy is quietly recalculated against this month's beds.
The second is the fee. Council uplifts are often agreed after April and backdated. The model has to hold the rate that was billed and the rate that was finally agreed, or the April bridge changes every time a council letter arrives.
The measuresSample DAX for occupancy, break-even and beds above the line
Available Bed Nights =
CALCULATE ( COUNTROWS ( 'Bed Night' ), 'Bed Night'[Status] <> "Closed" )
Occupied Bed Nights =
CALCULATE ( COUNTROWS ( 'Bed Night' ),
'Bed Night'[Status] IN { "Occupied", "Hospital hold" } )
Occupancy % = DIVIDE ( [Occupied Bed Nights], [Available Bed Nights] )
Average Weekly Fee =
DIVIDE ( SUM ( 'Fee'[Weekly Amount] ), DISTINCTCOUNT ( 'Fee'[Resident Week Key] ) )
Break-even Residents =
DIVIDE ( [Fixed Cost per Week], [Average Weekly Fee] - [Cost per Resident Week] )
Beds Above the Line =
DIVIDE ( [Occupied Bed Nights], [Nights in Period] ) - [Break-even Residents]
The detail that decides whether a home manager trusts the number is the hospital hold. A resident in hospital is usually still paying, and the bed can't be sold, so it counts as occupied for fees and break-even. Count it as vacant and the home looks emptier and riskier than it is; count a closed room as available and it looks fuller than it should.
Break-even uses the average fee of the residents the home actually has. If the next admission is council-funded at a lower rate, the line for the following week rises, so the forecast should use the expected payer of each confirmed admission, not the home's average.
Already using your care system's dashboards, or a benchmarking service?Where this adds something, and where it doesn't
The main digital care record and rota systems now ship good dashboards: occupancy, care delivered, dependency against staffing, incidents. Accountants and benchmarking services compare a home's fees, payroll and margins with its peers. If all your homes are on one system, your accounts are by home and month, and someone already works out break-even after every fee change, you have most of what this article describes, and we'd tell you so.
What none of them sees is the data held by the others. The care system knows who is in the beds but not what they pay. Billing knows the fee but not the rota. Benchmarks arrive once a year, a long way after the April that changed them. Beds above the line needs all three, by home and by week. That join is what we build.
What we need, and what we don't
A registered manager will ask about resident data before anything else, and should. This reporting needs operational data: room status by night, admission and discharge dates, payer and weekly fee, shift hours and costs, and costs by home. It doesn't need care plans, care notes, medication records, diagnoses or anything else from a resident's health record, and resident identities can be pseudonymised before anything leaves your systems.
On the contract side, DataHexis works as a UK GDPR data processor, with a signed data processing agreement, named sub-processors and storage in the UK. Nothing we build writes back: care records and rotas stay exactly as your systems hold them. Because the source systems hold special category health data even where the extract doesn't, the sensitive order is your due diligence first and data second.
Single-home owners
Many care homes in the UK are run by an owner with one home and a manager they've worked with for years. One home is one spreadsheet's worth of data, not a warehouse's. It needs one sum, done properly: the weekly cost of running the home with the rota it actually uses, divided by what each resident pays after food and consumables. Often a spreadsheet fed from payroll and the fee schedule does it, and we'd recommend that before anything bigger.
If anything, a single home needs the line more than a group does. Grant Thornton's figures show homes under 40 beds earning about ten points less margin than homes over 60, because the fixed staff don't shrink with the building. In the worked example, Westgate breaks even at 26 residents and has 29. Redo the sum every April, and every time a council fee letter or a new placement changes the mix. Once a year after the accounts is too late to act on it.
Frequently asked questions
What is a good occupancy rate for a care home in the UK?
For context, the government's Capacity Tracker showed 86.8% of care home beds in England occupied in the week ending 14 August 2026, and Knight Frank's corporate sample averaged 88.7% for 2024/25. But "good" depends on the home's break-even occupancy, which varies with fees, payer mix, size and staffing. Check which beds a figure is measured against: registered and available beds can give results several points apart.
How do you calculate break-even occupancy for a care home?
Add up the weekly costs that don't change with the number of residents (mainly the staffing rota, plus running costs and rent). Divide by the contribution from each resident: the average weekly fee minus the cost per resident of food, consumables and laundry. The answer is the number of residents needed to break even; divide by available beds for a percentage. In the worked example, Orchard View has £34,240 a week of fixed costs and £986 of contribution per resident, so it needs 34.7 residents, or 79% of its 44 beds.
What percentage of care home income goes on staff?
It depends on the sample. Knight Frank's review of the corporate sector found staff costs at 55.3% of income in 2024/25, down from 56.7%. Grant Thornton's report on the wider independent sector put them at 62% for residential and 70% for nursing homes. Smaller homes and those funded mainly by councils usually sit at the higher end, which is why the same pay rise moves their break-even line further.
How much do councils pay for care home places?
Provisional government figures for 2025/26 put the average council fee for older people at £956 a week for residential care (up 5.3%) and £1,089 a week with nursing, excluding NHS-funded nursing care (up 4.9%). Rates vary widely between councils. LaingBuisson estimates the gap between private and council fees has narrowed to about 25%, from around 40% in the mid-2010s.
What is the NHS-funded nursing care rate for 2026/27?
The standard weekly rate of NHS-funded nursing care in England rose 5.4% to £267.68 from 1 April 2026, up from £254.06. The higher rate rose to £368.24. It's paid for residents assessed as needing care from a registered nurse, on top of whatever the council or the resident pays for the rest of their care.
Sources
- DHSC, Adult social care provider statistics, England: quarterly update to August 2026, published 3 September 2026 (86.8% of care home beds occupied, week ending 14 August 2026; nursing 87.4%, residential 86.5%). gov.uk
- Knight Frank, UK Care Homes Trading Performance Review 2025, financial year April 2024 to March 2025 (80% of the corporate market, over 100,000 beds; occupancy 88.7%; staff costs 55.3% of income; average weekly fee £1,298; EBITDARM 30.1%; agency 3.3% and 2.9% of staff costs). knightfrank.co.uk
- Grant Thornton, Care homes for the elderly, 2026 (staff costs 62% residential, 70% nursing; EBITDARM 22.6% for 1–39 beds, around 32.6% for over 60 beds; agency outlook). grantthornton.co.uk
- CQC, State of Care 2024/25: adult social care access, 24 October 2025 (occupancy 78% in 2021/22, 84% in 2024/25). cqc.org.uk
- DHSC, Market Sustainability and Improvement Fund: provider fee reporting 2025 to 2026, 21 August 2025, provisional (£956 a week without nursing, up 5.3%; £1,089 with nursing excluding FNC, up 4.9%). gov.uk
- LaingBuisson, Care home market reaches £27bn as inflation drives growth and fee gap narrows, 23 April 2026 (self-funders 45% of volume, 55% of value; private–council fee gap about 25%; 96% independent). laingbuisson.com
- GOV.UK, National Living Wage increases to £12.71 per hour, 1 April 2026 (up 4.1%). gov.uk
- HM Treasury and HMRC, Changes to Class 1 National Insurance contributions, 13 November 2024 (employer rate 13.8% to 15%; secondary threshold £9,100 to £5,000; Employment Allowance £5,000 to £10,500, from 6 April 2025). gov.uk
- DHSC, Better community care thanks to nursing funding boost, 9 March 2026 (FNC standard rate £267.68 from 1 April 2026, up 5.4%). gov.uk
- THIIS, Care England welcomes 5.4 per cent increase in the NHS-funded nursing care rate for 2026/27, 11 March 2026 (higher rate £368.24). thiis.co.uk
- Skills for Care, The size and structure of the adult social care sector and workforce in England, 2026, reported by National Health Executive, 24 June 2026 (vacancy rate 6.2%; 30,000 international recruits). nationalhealthexecutive.com
Related reading
How many beds above the line is each of your homes?
In a free 60-minute discovery call, we'll look at where your bed, fee, rota and cost data lives, and tell you what a home-by-home break-even view would take. Or start with our fixed-price Occupancy Health Check, from £750, credited in full against any build.
Book a free discovery call → Pseudonymised resident data · Delivered by a partner · First dashboard usually within about two weeks