The Tenant

In autumn 2024 I reported a repair. A surveyor came. An engineer came and photographed it. An electrician came, in the words of the file, "to have a look."

Then nothing happened for fifty-one days.

I know it was fifty-one because I counted them myself. One row per day in a spreadsheet, each row saying the same thing: no communication received. Across that one repair there were 128 such rows out of 252 days. For more than half the time I spent waiting, nothing whatsoever was happening.

Those 128 rows exist in my data. They do not exist in theirs. They cannot, because their system records events, and silence is not an event. Nobody did anything, so nothing was written down.

In the file, those seven weeks look identical to seven weeks in which there was nothing to report.

This is the foundation of everything that follows. Repairs data is a record of activity, not of experience. Every row is created by somebody doing something. The intervals when nobody does anything - which is where tenants actually live - generate no rows at all, and so cannot be counted, reported, benchmarked, targeted or improved.

No housing system I have ever seen measures elapsed silence. It is the largest single component of the tenant experience and it is structurally invisible.

It began in January 2016, third on a list of three, after a dripping tap and a light that wasn't working:

It ended in July 2025, when two extractor fans were fitted. Three thousand four hundred and seventy-two days.

Now ask how that looks inside their system. It does not appear as one thing lasting nine and a half years. It appears as a long series of separate jobs. Each raised. Each attended. Most, in time, closed.

Every closure is a completed repair. Every completed repair is a success. A number of them will have been completed within target, because the target runs from the moment that particular job was raised - not from the moment I first said there was mould.

Their system models jobs. I was living a problem. Those are different entities, and only one of them exists in the database.

I put this to them myself in 2024, before I had any standing to and without a word of the vocabulary above:

Until a repairs system can hold a recurring problem as a thing in its own right, with the clock running from the first report rather than the latest job, a house that defeats a family for a decade will keep presenting in the management information as a well-served address with a slightly high job count.

In its 2023/24 Tenant Satisfaction Measures report, the Regulator of Social Housing published two figures.

Asked directly, 28% of social housing tenants said they had made a complaint to their landlord in the previous twelve months.

The complaints their landlords actually recorded came to 6%.

The regulator's own explanation is that "what some tenants consider as a complaint is likely to have been recorded as a service request by a landlord."

Read that again, because it is published by the body that regulates the sector. Roughly four out of every five tenants who believe they made a complaint do not appear in the complaints data as having made one.

I know exactly how that happens, because it happened to me, and nobody involved did anything wrong. A tenant reports a thing. It is logged as a service request, because that is what it is. It doesn't get done. She reports it again. New request. It still doesn't get done. She says she wants to complain, and is told the job is being worked on. Nothing is refused. Nothing is concealed. And at the end of the year the complaints figure is a small, tidy number bearing no relationship to how many people were unhappy.

Two more from the same report. Satisfaction with complaint handling is the lowest score in the entire suite of tenant satisfaction measures - a median of 35%. And of the 250,000-plus stage one complaints landlords did record that year, 73% were answered within the required timescale.

Answered on time. The worst-rated thing in the sector. Both true at once, because the measure rewards responding and says nothing whatsoever about resolving.

My complaint about that repair was logged in late 2024. Three months later it was logged again, as a new complaint with a new reference, despite my explaining at length that it was the same one.

I understand exactly how that happened. Somebody conscientious noticed that no complaint had been raised at the time, and raised one. That is a person doing their job properly.

But look at what it does to the numbers. One grievance becomes two records. The second starts with a fresh clock and is comfortably answered within target. The first, having never formally existed, breaches nothing.

There is a related move, and it also happened to me. When I first asked for a complaint to be opened about the silence, it was not accepted - on the grounds that the job had "morphed into a more complicated job" and they were working hard to resolve it.

A complaint that is not accepted does not appear in the complaints data at all. Not upheld, not rejected. Absent.

My own complaints, when they did exist, were acknowledged promptly, investigated conscientiously and answered courteously, for years, while the thing itself stayed exactly as it was.

I asked for two things instead. Why, when my job sat untouched for seven weeks, was there nothing in the system that noticed it had been lost. And why, when a contractor failed to arrive, was there nothing that caught it and rebooked me.

Neither question was answered. I was told that complaints are reviewed regularly, and that training and feedback had been provided.

Compensation closes a record. In database terms it is a resolution code. What I was asking for was a change to the process that had generated the record - and there was no mechanism through which a tenant could ask for that.

That absence is the finding.

I am not alleging that anybody gamed anything. I don't believe they did, and I would say so to a regulator. That is the uncomfortable part. You do not need bad faith to produce flattering numbers. You need a system in which opening a new record is easier than linking to an old one, and in which refusal is a valid outcome. Both of those things are true nearly everywhere.

Eighty-four people, across sixteen years and a series of mergers and rebrands that changed the organisation's name more times than I can now reconstruct.

I am not complaining about turnover. People are entitled to move on.

But consider what it does. There is no continuity of person. Nobody has held my history in their head for longer than a year or two. Which means the record is the only thing that could possibly carry the story from one of them to the next.

And the record is organised by job.

So each new person inherits a list of closed jobs and one open one. They do not inherit nine years. They inherit today. They arrive, in complete good faith, at a problem they believe is new, and they treat it as new, because everything in front of them says that it is.

That is why I have told the same story to strangers standing in my own bathroom more times than I can count. Not because anybody was careless. The system was designed to forget, and the people were doing exactly what it told them.

And I want to be unambiguous about those people. In sixteen years not one person who came to my house was rude, or idle, or unkind. They arrived with no history, no context and no brief, and were expected to stand in a stranger's bathroom and absorb her disappointment. Several were visibly embarrassed. That embarrassment was never theirs to carry. Nothing I have written here is a criticism of any of them. Every failure I am describing is a failure of design, executed faithfully by people who deserved better tools than they were given.

Eight in the morning until five in the evening. Six separate times.

A nine-hour window is not an appointment. It is a day of my life, and it is priced at nothing by the only party in the transaction who isn't paying for it.

You cannot leave the house. You cannot commit to anything, collect anyone, or work anywhere else. If you have a job you spend a day of leave. If you have children you rearrange around it, and then rearrange again on the days when nobody comes.

Add that to the hours on the telephone that nobody logged, and to sixty-six thousand words written in the evenings - twenty-seven per cent of my emails were sent outside working hours, fifty of them after eight at night, the earliest at twenty-three minutes past midnight.

Tenant time is the largest single input into this process and it appears nowhere in any account of it. It is contributed for free, in unlimited quantities, by the person with the least of it to spare, and because it is never costed it is never conserved.

I include that not out of sentiment but because it forecloses the easiest explanation available to anyone who wants one.

I was not a difficult tenant. I was polite, in writing, for sixteen years, replying at a median of one hour, thanking people in nearly half of everything I sent.

And it did not work.

Which is the point at which the arithmetic finally became obvious to me. Reporting a repair cost me days of leave, a filing system, photographs, a spreadsheet, sixty-six thousand words, uncounted hours on the phone and a quantity of hope I could no longer afford - and it did not reliably produce a repaired thing at the end.

So I stopped. I worked around faults in my own home rather than mention them.

And on paper, I improved.

I have written all of this down because I have run out of other places to put it.

Sixteen years. Sixty-six thousand words. A spreadsheet with a hundred and twenty-eight identical rows in it. Three years on my landlord's customer committee, trying to get at the inside of the problem rather than the outside. A scrutiny framework I wrote in my own time, at nobody's request, proposing exactly the kind of deep, root-cause review that would have caught what happened to me - which I sent in, and which nobody ever acknowledged.

None of it worked. I want to be honest about that rather than tidy.

I also work now in the software that housing organisations use to schedule repairs. I'm not going to pretend that makes me an expert in it - I'm early into that part and learning in public. But it does mean I spend my days looking at the inside of the machine whose outside I stood in front of for sixteen years, and I have yet to meet many people who have done both.

That is the only thing I have that is genuinely rare. Not the years. Not the mould. Plenty of people have those, and most of them have had it far worse than I have. What I have is that I can read the database and I am in it. I know which fields exist, and I know what happened to me in the gaps between them.

Earlier I wrote that a tenant who has given up is indistinguishable, in the data, from a tenant with nothing wrong.

I was that tenant. This is me declining to be her any longer.

So I am putting it here, where it can be read, argued with and used - because sixteen years of raising it through the proper channels fixed one bathroom, and I would like the next sixteen to be worth more than that.