How to reduce a housing repairs backlog

The short answer

1. Build one list of every open repair, dated from the tenant's first report.
2. Triage by risk and vulnerability, with the Awaab's Law timescales on top.
3. Measure the days from the tenant's report to a home that is right.
4. Count first-time fix on the whole repair.
5. Diagnose at first contact and send the right part.
6. Agree appointments with the tenant, so fewer visits end in no access.
7. Count how often tenants chase, and remove the causes.
8. Hold contractors to the whole repair.
9. Add temporary capacity last, for a fixed period.

Regulator of Social Housing, Tenant Satisfaction Measures 2024/25. Analysis: The Chase Report, Part 1.

1. Build one list of every open repair

Start by putting every open job in one place, whichever system or contractor holds it, and take out the duplicates. Against each one, record the date the tenant first reported the fault and what the job is waiting for: a part, access, a second trade, an approval or a contractor.

Date the repair, not the job. A job that was closed and raised again is one repair with one start date, and the tenant has been counting from the first.

Then look at what has already left the list. In 2024/25, of 18.7 million responsive repairs raised in England, more than 2.5 million were closed without being completed. The median landlord closes 11% of the repairs it raises this way. Some of those are duplicates and cancellations. Others are faults the tenant still has, so sample them before you trust the size of the backlog.

2. Triage by risk and vulnerability, not by age alone

Age is the wrong first sort. Put hazards first: damp and mould, anything that is an emergency, and homes where the household is vulnerable.

In England this is law for part of the list. Under Awaab's Law a landlord must investigate an emergency hazard and make it safe within 24 hours, and must investigate a significant damp and mould hazard within 10 working days of becoming aware of it. From 30 November 2026 the same timescales cover seven more groups of hazards. The guide to Awaab's Law Phase 2 sets them out.

A backlog job that falls under these rules is already on a statutory clock. That clock started when the landlord became aware of the hazard, not when the job reached the top of the list.

3. Measure from the tenant's report to a home that is right

Most boards watch the share of repairs completed within target. Across 352 large landlords in England in 2024/25, that figure explains 12% of how residents feel about the time their repair took. Landlords reporting 95% or more inside target range from 54% to 96% of tenants satisfied with the time taken. The length of the target a landlord sets itself has no relationship with satisfaction at all (r = 0.00).

Repairs targets measure the landlord's promise, not the tenant's wait.

Scotland publishes a number that does track satisfaction. Across 161 Scottish landlords in 2025/26, those averaging five working days or fewer to complete a non-emergency repair had a median repairs satisfaction of 92.5%. Those averaging more than twelve days had 84.4%.

So run the backlog on the days from the tenant's report to a home that is right, and look at the spread of those days as well as the average. The average hides the longest waits.

4. Count first-time fix on the whole repair

First-time fix is the measure most often recommended for a backlog, and as usually reported it is flattered. Once a repair can be split into a first visit that counts as fixed and a follow-on that counts as a new job, almost every landlord passes. The median Scottish landlord reported 97.7% right first time in 2025/26, and the figure had almost no relationship with satisfaction (r = 0.07).

Count it on the whole repair instead: one fault, one count, however many jobs it took. A separate page explains why the reported figure misleads and how to count it properly.

5. Diagnose at first contact and send the right part

A repair that needs a second visit sits in the backlog twice. Every translation between the tenant's words and the job that gets raised is a place where the wrong trade, the wrong part or the wrong priority gets booked. So let the report reach someone who can diagnose it, with photographs and the repair history of the home in front of them.

Then look in the van. When one council's contractors measured it, operatives were spending an average of 1.8 hours a day each collecting parts, before van stock was set against the demand for parts. By the council's own count, 15% of repairs took four or more visits. Both figures come from one landlord's published case study (2011). They are not a sector benchmark.

6. Agree appointments with the tenant

A visit that ends in no access uses a slot and fixes nothing. Before treating it as the tenant's fault, check how the appointment was made. Was a time agreed with the tenant, or were they given the next free slot? Did anyone tell them the operative was coming?

When Tees Valley Housing counted its repairs contact for the government's housing pilot in 2005, the largest single cause of its failure demand was tenants asking when someone would be coming.

7. Count how often tenants chase

Some of the work arriving each week is new repairs. The rest is tenants chasing an appointment, reporting the same fault again or saying it still is not fixed. That second kind is failure demand, and a clearance team does not reduce it.

When two landlords counted theirs before changing anything, it came to 45% of repairs contact at Tees Valley Housing (2005) and 60% at one council (2011). Those are the landlords' own counts. No public source counts chasing across landlords, so there is no benchmark and you need your own number.

Take one day of repairs calls and messages and sort them into two piles: a new repair, or a repair already reported. The share in the second pile is your chase rate. List the reasons in that pile and start with the biggest.

The cost shows up in complaints. Landlords in the bottom quarter for repairs satisfaction record 1.8 times as many stage 1 complaints per 1,000 homes as those in the top quarter (a median of 69 against 38), and 2.7 times as many at stage 2.

8. Hold contractors to the whole repair

If a contractor is paid and judged job by job, a follow-on is a second job completed on time. Ask contractors for the same numbers you ask of your in-house team: days from report to finished, visits per repair and repeat reports.

Check who the tenant has to ring when a second trade is involved. If the tenant is relaying messages between you and the contractor, the repair has no owner.

9. Add temporary capacity last, and for a fixed period

Overtime, agency operatives or a second contractor will clear aged jobs, and for hazards you should not wait. Give the extra capacity an end date and keep it apart from the day-to-day service.

Then watch two numbers every week: repairs arriving and repairs finished. If arrivals are still ahead once the old jobs are gone, the list refills. A backlog is rarely a capacity problem alone. Much of it is often the same repairs coming back, which is why adding people tends to slow its growth rather than clear it.

What the public data shows, and what it does not

The question behind this page is whether a landlord's measures show what tenants live through, or only make the service look fine. On repairs, the public data gives an answer.

Landlords measure repairs against the clock they set for themselves, not the clock the tenant lives by. Completed within target counts whether the landlord kept its own promise. It does not count how long the home was broken, how many visits it took or how often the tenant had to ring.

My proposition is that four numbers from the tenant's side track satisfaction and trust more closely than the target does: the days from the tenant's report to a home that is right, first-time fix counted on the whole repair, the spread of those days, and the number of times the tenant had to chase.

The regulator's own headline report for 2024/25 notes the gap between landlords' within-target results and tenant satisfaction (paragraph 5.7) without quantifying it. I am not aware of a published analysis across landlords that tests one against the other.

Only repairs has been tested. The same targets run voids and lettings. There is a scorecard for each of the three services: repairs, voids and allocations.

Sources: Regulator of Social Housing, Tenant Satisfaction Measures 2024/25 (large landlords, low cost rental accommodation); Scottish Housing Regulator, Charter data 2025/26; regulatory judgements graded C3 or C4, retrieved on 2 October 2026 and classified by keyword. Analysis: The Chase Report, Part 1. Correlations are Spearman rank correlations; the 12% is R² from a least-squares fit across the 350 landlords with both measures. These are associations across landlords in a single year. They say nothing about cause at any one landlord.