Repairs attempted 13 April to 4 May 2026 · Published: 28 October 2026 · Last updated: 28 October 2026 · Reading time: ~8 minutes · The parts receipts and the photographs in this post are my own.
In the winter I replaced three things in my flat and then worked out, three months later, that two of them were not decisions at all. They were fault repairs with a shopping trip attached. The obvious question, and one I did not ask in that post, is: why didn't I just repair them?
So I tried. Not the same three. The chair in that post had nothing wrong with it, which was the whole point of that post, and the lamp had already been replaced. The kettle was the one I could genuinely have repaired, and I did not, because by the time I had finished writing that post I had already bought a new one and the question had stopped being interesting. That is its own finding, and not a flattering one: the window for repairing something closes the moment you stop being annoyed by it.
These are three different things that were genuinely broken and that I had been walking past for months. Two worked. One did not, and the one that did not cost me more than replacing it would have.
The three attempts
I picked things that were actually broken, not things I was vaguely unhappy with. That distinction is the whole reason this experiment exists, and I will come back to it.
| Item | Fault | Parts | Time | Result |
|---|---|---|---|---|
| Hallway lamp | Loose switch, frayed flex | £4 | 25 min | Worked. Looks and behaves like new. |
| Kitchen chair | Wobble at the base, one bolt housing cracked | £9 | 120 min | Failed. Part discontinued, universal one did not fit. |
| Toaster | One slot stopped heating | £12 | 90 min | Worked. Casing does not quite seat and it rattles now. |
| Total | £25 | ~4 hrs | Two repairs, one replacement, £110 all in | |
The £110 on the bottom row is the £25 in parts plus the £85 I spent on a new chair when the repair failed. If I had replaced all three outright it would have been around £160, so I saved roughly £50 for four hours of work. That is £12.50 an hour, which is almost exactly the rate I worked out in my food-waste experiment, and I am starting to think that is just what my time is worth in these situations.
The one that worked: the lamp
This was easy in a way that made the whole experiment feel promising. The switch was loose, the flex was frayed near the plug, and both were visible faults I could name before I bought anything. A new switch and a length of flex came to £4. The repair took 25 minutes, most of it unscrewing the base and being careful not to cross the wires.
It works. It looks exactly like it did before, which is the right result for a lamp. Total cost £4, total time 25 minutes, and I would do it again without hesitating.
The one that failed: the kitchen chair
The chair had a wobble at the base. One of the bolt housings had cracked, and the bolt was no longer gripping. I looked up the model and found that the manufacturer had discontinued it two years earlier, and that the specific housing was not sold separately.
So I bought a universal replacement housing for £9, because £9 is worth trying. It did not fit. Not "did not fit well." It did not fit at all, because the bolt thread was a different pitch. I spent about forty minutes on the first attempt, another twenty looking up whether anyone sold an adapter, and then a further hour taking it apart, putting it back together with the cracked housing, and deciding that I was not going to solve it.
Two hours and £9 later, I bought a new chair for £85. Total for the chair: £94 and two hours, for a chair that cost £85. I was £9 worse off than if I had simply replaced it on day one, and I had also spent two hours, which I am not counting in the £9.
The one that half-worked: the toaster
One slot stopped heating. This was the most repairable of the three in principle. The element is a consumable, it is usually sold as a part, and the toaster was a model I could look up. It took a week to find the right element and it cost £12.
Fitting it took 90 minutes, and most of that was not the element. It was the casing. The casing has six screws and two plastic clips, and once you have the clips off, getting them back on without breaking them is a genuinely fiddly job. I got it back together. It works. It also rattles now, and the seam where the casing joins does not sit flush like it used to.
So it is a functioning toaster that looks slightly wrong. That is not a failure, but it is not the clean result the lamp gave me. If I had known about the casing before starting, I might have decided differently.
What the three have in common, and what they don't
The lamp and the toaster both worked, and they worked for the same reason: the fault was singular and the part was available. A loose switch is one fault. A dead element is one fault. Neither required me to understand the whole object, only to swap one piece of it.
The chair failed for the opposite reason, and it is not the reason I would have guessed before starting. The fault was also singular: one cracked housing. The problem was that the part did not exist. The chair was not unrepairable because it was complicated. It was unrepairable because it had been discontinued, and no amount of skill would have changed that.
Which means the thing that determines whether a repair is possible is not the fault. It is whether the manufacturer still sells the part. That is a supply question, not a repair question, and it is the first thing to check, before you take anything apart.
The thing I got wrong
I assumed the obstacle to repair was difficulty. It is not. The obstacle is uncertainty, and it costs you money in a way that difficulty does not.
Before you repair something, you know three things and you do not know one. You know what the fault looks like. You know roughly what it will cost in parts. You know roughly how long it will take. You do not know whether it will work. Replacement has none of that. You buy the new thing, it works, and the uncertainty is zero.
That asymmetry is why people replace things that could be repaired, and it is rational. If I spend £9 and two hours on a chair and it fails, I have lost £9 and two hours and I still have to buy the chair. If I buy the chair, I am done in twenty minutes. The repair is only worth attempting when the expected value is clearly positive, and expected value is a terrible thing to calculate while standing in a shop holding a broken part.
So the rule I came out with is this. Before you repair something, look up the part. Not the fault, the part. If the part exists, is cheap, and can arrive in a few days, repair is probably worth attempting. If the part is discontinued, or only sold as part of an assembly that costs most of a new one, stop. You are not repairing anything. You are gambling, and the house has already taken the odds.
What I would do differently
I would have looked up the chair part before spending the £9. That sounds obvious, and it is, and I did not do it because I wanted the repair to work and looking it up would have told me it would not. The £9 was a way of postponing a decision I had already made.
And I would have asked, before starting the toaster, whether the casing was easy to get back together. I did not, and the rattle is what I got. If I had known, I would still have done the repair, but I would have been prepared for the result being "functions, looks wrong" rather than "like new."
Should you do this?
Yes, but only on the things where the part exists. That is the whole test and it takes five minutes.
Walk round your flat, pick the broken thing you have been meaning to deal with, and search for the specific part. If a seller comes up with the part in stock at a sensible price, attempt the repair. My success rate on that condition was two out of two, and neither cost me more than 90 minutes. If the only results are discontinued notices, forum threads from 2019, and a universal part described as "fits most models," skip it. You are about to spend money on a coin flip.
The honest summary: three attempts, £25 in parts, four hours, two successes and one failure. The failure cost me £9 and two hours more than just replacing the chair would have. The successes cost me £16 and about two hours for things that would have been £75 to replace. That is a real saving and it is not a large one, and the rate works out at about £12.50 an hour, which is the same rate I got for reducing food waste and is probably just what my time is worth when I am doing chores.
Frequently asked questions
Is it safe to repair electrical things yourself?
For a lamp or a toaster, with the plug out, I am comfortable. Both are simple circuits and neither stores charge once unplugged. I turned the toaster upside down over the sink before opening it, which produced about a teaspoon of crumbs and is the only preparation I do beyond unplugging it. I would not say the same about a microwave or anything with a large capacitor in it, and I did not touch either. If you are not sure which category a thing is in, that uncertainty is the answer.
Did the toaster rattle bother you?
For about a week. Then it became the noise the toaster makes. That is the same thing that happened with the lamp. The fault disappeared and then the object disappeared with it. The difference is that the toaster's disappearance is slightly louder than the lamp's.
Would you repair rather than replace next time?
Yes, if the part exists. No, if it does not. Which sounds like a trivial answer and is actually the whole finding of this post. The decision does not depend on how good you are at repairing, or how much you want to save, or how much you care about waste. It depends on whether the manufacturer still sells the part, and that is a fact you can check in five minutes.
What about the environmental argument?
I did not repair these for environmental reasons and I am not going to pretend otherwise. If I had, I would have been more upset about the chair, because the failure was not my fault and the object was structurally fine apart from one small housing. That is probably the strongest argument for repair: that a sound object gets thrown away over a part nobody makes any more. And it is not the one I used.
About this post
Written by Maha Alkumaish. I write about the routines and systems that survive a real working week, tested on myself before they get written about. The parts receipts, timings and photographs in this post are my own.
All photographs in this post are my own. I was not paid or sponsored by any brand mentioned, and there are no affiliate links in this post. More about me and how these experiments are run.
Related reading on this blog:
- I replaced three things I'd owned for years, the experiment that asked this one: two of the three were fault repairs, so why not repair them instead?
- I logged everything I threw away for six weeks, where the same £12-an-hour figure came up, for a different reason
- More Home & Space experiments
0 Comments