The Plan You'll Finish

Imagine five debts and a fixed amount each month to throw at them. A financial adviser, or a spreadsheet, will tell you the correct thing to do, and they’ll be right: attack the one with the highest interest rate first, put every spare pound there, ignore the balances, and only move on once it’s gone. It’s called the avalanche method, and it provably minimises the interest you pay. There is no arguing with the arithmetic.

Now picture actually doing it. Your highest-rate debt happens to be a large one, so for the first eighteen months you pour everything into it and watch almost nothing happen. The other four debts sit there, untouched, mocking you from the top of every statement. You get no sense of progress, because on the avalanche plan there isn’t any to see yet, just a slightly smaller version of the same big number, month after month. And somewhere around month fourteen, tired and unrewarded, you quit. You stop the plan, drift back into old habits, and the interest you were so carefully minimising resumes on all five debts at once. Congratulations: you followed the optimal strategy, right up until you didn’t, and now you’re worse off than if you’d never optimised anything.

That is the whole subject, and it exposes a truth about money that the spreadsheets systematically hide: personal finance is not, mostly, a mathematics problem. It’s a behaviour problem. And the single most important number in any financial plan is not its theoretical efficiency, it’s the probability that you actually finish it, which is exactly the number the maths leaves out.

Consider the alternative the adviser frowned at, the snowball method: pay off your smallest debt first, whatever its interest rate, then take the payment that debt was eating and roll it onto the next-smallest, and the next. On paper it’s worse, sometimes by real money. In practice it gets far more people out of debt, and we’re not guessing about this. When two researchers at Northwestern went through the records of thousands of real people working their way out of debt, they found something the arithmetic can’t explain: what predicted whether someone actually became debt-free was the fraction of their separate accounts they’d managed to close, not the number of dollars they’d paid down. Wiping out whole debts, one at a time, was what kept people going. The size of those debts barely mattered. It was the count of finished things, not the sum of the money, that carried them to the end.

The reason isn’t mysterious once you stop treating people like calculators. Every completed debt is a small, total, unarguable victory, an “I did that,” and those are the moments that build the belief that the whole thing is possible. Psychologists call it self-efficacy, and it’s fed by visible wins, not by the abstract knowledge that you’re on the cheapest theoretical path. There’s a related quirk, the goal-gradient effect, where effort surges as a finish line comes into view, and clearing a small debt outright drags a finish line right up close where you can see it. And there’s the sheer weight of the number of debts, which the spreadsheet ignores entirely. Each one is its own open loop in your head, its own statement, its own due date, its own low hum of dread. Closing one, even a tiny one, doesn’t just reduce what you owe; it removes a whole worry from the pile, and that relief is a real resource even though it earns nothing and appears nowhere in the interest calculation.

Put those together and the avalanche’s flaw comes into focus. It concentrates all its reward at the far end, and asks you to run the longest, most barren stretch, no visible progress, no closed accounts, for months, precisely at the point where most people give up. It saves its money at exactly the moment it’s most likely to lose its owner. The snowball, by contrast, front-loads the encouragement, spending a little efficiency to buy the one thing that actually gets people to the finish: the feeling that it’s working.

And here’s a detail that makes the trade less painful than it sounds: the money avalanche saves is usually smaller than people imagine. The interest gaps between ordinary consumer debts tend to be modest, a nineteen-percent card, a twenty-two, a twenty-four, and total interest depends not just on the rate but on how long you take. If the snowball gets you finished faster because you never quit, that shorter run can quietly eat most of the theoretical penalty. In a lot of real cases the “cost” of doing it the wrong way is a rounding error, and the “benefit” of doing it the right way is a plan in a drawer.

I don’t want to oversell the snowball, though, because there’s a case where the maths genuinely wins and you should listen to it. If one of your debts is carrying a savage rate, a payday loan at several hundred percent, a store card in the thirties, that’s not a debt to be politely queued behind smaller ones for morale’s sake. That’s a fire. It compounds fast enough that the interest gap really is serious money, and clearing it first isn’t optimising, it’s putting out the fire before it spreads. And there’s a neat compromise that captures most of both worlds: kill your single smallest debt first, purely for the launch of momentum and one fewer creditor, then switch to the avalanche for everything that’s left. You give up a sliver of interest to buy a morale boost, and keep the efficient path for the long tail.

So the honest way to choose is to choose for the person you actually are, not the person the spreadsheet assumes. If you’ve started financial plans before and abandoned them, or your debts are lots of small ones at similar rates, take the snowball and don’t feel bad about it, the version of you that finishes is worth more than the version that’s technically correct for four months. If you’re genuinely disciplined and one debt is clearly the expensive one, take the avalanche and pocket the difference. The mistake is not picking the wrong method; the mistake is picking the method that suits an idealised human who isn’t you.

Which is the part worth carrying well beyond debt, because this is not really about debt at all. It’s about the gap between the optimal plan and the plan that gets done, and that gap governs almost everything people try to improve. The perfect training programme that’s miserable gets abandoned; the merely good one that you enjoy gets completed, and completed beats perfect every time, because the binding constraint on results is almost never the quality of the plan, it’s the consistency of the execution. We are not optimisation engines. We’re story-following creatures, and we keep going through a narrative, a beginning, a struggle, visible progress, an ending, far more reliably than through a cost-minimisation curve. A plan that gives you that arc will out-perform a cleverer plan that doesn’t, whether you’re clearing debt, building savings, learning a language, or trying to write a book.

So the next time someone hands you the mathematically optimal way to do something hard and slow, ask the question the spreadsheet can’t: is this the plan I’ll still be running in a year? Because a good plan you finish beats a perfect plan you quit, every single time, and the whole trick of getting anything done is to stop designing for the machine you aren’t and start designing for the human you are.