Four Savings Goals Failed and the Fifth Worked. The Difference Was Specificity.

Between 2021 and 2024 I set up four savings pots and abandoned all of them. They were called, in order: Savings, Emergency Fund, Buffer, and Savings 2. The fifth, set up in late 2024, was called “Studio floor, June 2025,” and it filled on schedule and I spent it on a studio floor.

Four failures and one success is not evidence of anything in general. But the failures had a shape and I want to describe it, because the shape suggests the mechanism.

Each of the four abandoned pots was raided, and in every case the raid was justified by the pot’s own name. Money in something called “Savings” is available for anything I can describe as a good reason to need savings, and I can always describe one. The Emergency Fund got taken for a laptop, and at the time the laptop genuinely was an emergency in the sense that I could not work without one, and that is exactly the problem — a name broad enough to be motivating is broad enough to be spent.

“Studio floor, June 2025” cannot be raided for a laptop without the raid being obvious to me. There is no argument available. The money is either for the floor or it is not, and taking it means writing a different name on the pot, which is a small act I would have to perform deliberately, and I never did.

The date does similar work in a different direction. The four failures had no date, which meant no month was the wrong month to fall short on monthly savings contributions. There is always next month. With June 2025 written on it, falling short in February was visible as falling behind rather than as a normal month, and twice I made up the difference in the following month, which I had never once done in three previous years.

I have since set up two more on the same pattern and both are on track at nine and four months in. The names are equally unglamorous.

The counter-argument I would make against my own conclusion is that maybe the floor just mattered more to me than an abstraction, and specificity is doing nothing except standing in for genuinely wanting the thing. I think that is partly right. But the emergency fund also mattered — arguably more, in any sensible ranking — and it did not survive, so wanting is not sufficient on its own.

The awkward consequence of this, which I have not resolved, is that a genuine emergency fund cannot be made specific. Its whole purpose is to be available for something unnamed. So the technique that works for me does not work for the single most important thing to save for, and my current arrangement is a compromise: an emergency fund at a different bank, no card attached, three working days to reach, and I have made accessing it slow rather than making it specific. Eighteen months and it is intact, which is longer than any of the four ever lasted.

If there is something transferable here it is probably that: a savings goal needs either a name that forbids the alternatives, or enough friction that a raid takes days rather than seconds. What does not work, for me, is a pot named after the abstract virtue of having one.