I spent £400+ trying to calm my anxious dog. Here are all 5 things, ranked worst to best.

This was the first thing the pet shop recommended and the first thing I'd skip. Two plug-ins, one in the living room, one in the hall, replaced every month at £17 a go.

Four months in I ran a test I should have run on week one: I unplugged both and didn't tell anyone. Nobody noticed. Zeno didn't notice. I was £68 down on a thing whose only measurable effect was on my electricity bill.

I'm not saying they never work for any dog. I'm saying that if you can't tell whether it's switched on, you're not solving the problem — you're just spending money to feel like you're doing something. And when you're worried about your dog, that feeling is very easy to buy.

The logic seemed sound — dogs are den animals, so give her a den. Everyone online said the same thing. I bought the crate, covered three sides, put his bed inside, fed him in there for a week to build the association.

He hated it. Not mild reluctance — he'd get in, then stand at the open door, panting, and would not lie down. On night nine he scraped at the tray until 2am and I gave up.

What I understand now, and nobody told me then, is that enclosure and confinement are not the same thing. A rescue dog who has been shut in somewhere before doesn't read a crate as shelter. He reads it as the door closing. The instinct I was trying to trigger was real — I'd just picked completely the wrong object to trigger it with. That distinction cost me £89 and nine bad nights.

The cheapest thing on this list and the only one I'd have felt daft paying for. Radio 4 on low from about eight in the evening, or a white noise track on a speaker.

It didn't fix anything. What it did was flatten the spikes — the bin lorry, a car door, someone walking past the window stopped producing a full alert. He still didn't settle, but she stopped un-settling every eleven minutes.

Worth doing because it costs nothing. Not worth expecting anything more from it than that.

I went through three brands. The first did nothing I could detect. The second and third genuinely took something off the top — he was less wired in the evenings, and fireworks night was survivable for the first time.

But here's the thing I only worked out later, and it's the reason this is number two and not number one: a supplement lowers the volume, it doesn't give her anywhere to go. He was calmer and still circling. Calmer and still repositioning four times a night. The underlying problem — that he had no fixed, enclosed, safe spot in the house — was completely untouched.

My honest read after fourteen months: supplements are a multiplier, not a foundation. On top of the right sleeping setup they're good. On their own, at £26 a month indefinitely, they're an expensive way to manage a symptom.