When you have several debts and a limited amount of extra money each month, the order you pay them in changes both what the debt costs you and how likely you are to finish. Those two things do not always point the same direction, which is the whole reason this question is interesting.
There is a step zero that outranks everything below it: make every minimum payment, every month, without exception. A late payment can damage your credit, trigger a penalty APR that is often several points above your normal rate, and on some accounts accelerate the balance. Optimising the order of your extra payments while missing a minimum is like re-routing traffic through a bridge that is on fire.
First, what the interest rate actually costs
APR is the annual cost of borrowing, as a percentage of the balance you carry. A card at 24.99% APR on a $1,000 balance costs roughly $250 a year — and the real figure is a little higher, because most cards compound daily.
Daily compounding works like this. The daily rate is the APR divided by 365, so 24.99% is about 0.0685% per day. On $1,000 that is $0.69 added every day. Because that interest then earns interest, the effective annual rate lands above 28%.
Put in dollars rather than percentages, the urgency is obvious:
| Debt type | Typical APR | Annual cost on $5,000 |
|---|---|---|
| Payday / cash advance | 200%+ | $10,000+ |
| Store card | 26–30% | $1,300–1,500 |
| Credit card | 20–25% | $1,000–1,250 |
| Personal loan | 10–15% | $500–750 |
| Auto loan | 5–8% | $250–400 |
| Federal student loan | 4–8% | $200–400 |
| Mortgage | 3–7% | $150–350 |
The top of that table is a genuine emergency. The bottom is a line item. Every dollar you put against a 25% balance earns you a guaranteed 25% return, which is better than you will reliably get anywhere else — and unlike an investment return, it is certain.
Avalanche: cheapest by arithmetic
Pay the minimums on everything, then put every spare dollar on the highest APR. When it clears, roll its payment into the next-highest, and continue. Avalanche always produces the lowest total interest. There is no scenario where another order is cheaper, because you are always starving the most expensive balance first.
The catch is that the highest-rate debt is frequently also a large one. If it takes fourteen months before a single account disappears, you spend over a year making payments with nothing visibly finishing. For some people that is fine. For others it is exactly where the plan dies.
Snowball: the one people finish
Same minimums, but the extra goes at the smallest balance. When it clears, its payment rolls into the next smallest, so the amount you are throwing grows each time an account closes.
Snowball costs more in interest, by definition. Its advantage is that accounts start disappearing quickly, and research from Harvard Business School found people using it were more likely to clear all their debt — despite paying more — because visible progress kept them going. A strategy you abandon at month eight is worse than a slightly costlier one you finish.
That rollover is the mechanism in both methods, and it is the part people leave out when they estimate this by hand. Each cleared account permanently increases what you can throw at the next one, so progress accelerates rather than staying linear.
Run both on your actual debts: the debt payoff calculator simulates each method month by month, rolling cleared minimums forward, and tells you what the cheaper one saves and which debt the faster one clears first.
How different are they, really?
This is the question almost nobody answers before choosing, and it is the one that should decide it. The gap depends entirely on your particular debts.
When your smallest balance is also your highest rate, the two methods pick the same target and the answer is identical — no trade-off exists, and the debate is moot. When your smallest balance carries your lowest rate, the gap is at its widest. In a realistic four-debt scenario, the difference often lands somewhere between a few hundred and a couple of thousand dollars across the whole payoff.
So the honest framing is not "which method is better". It is: here is what finishing sooner costs me, and is that a price worth paying for momentum I know I need? If avalanche saves $120, take the snowball and the early win without a second thought. If it saves $2,400, that is a real sum and worth some discomfort.
The decision, in order
- Every minimum, every month. Non-negotiable, before any optimising.
- Anything above about 15% APR is the priority. Cards, store cards, payday loans. At these rates the compounding outruns most things you could do with the money instead.
- Check whether the methods even disagree. If your smallest balance is also your highest rate, stop — both answers are the same.
- Price the trade-off. Run both. If the difference is small, choose the one you will stick with. If it is large, lean avalanche.
- Below roughly 6–7%, it becomes preference. Mortgages and federal student loans are cheap enough that paying ahead is a values question, not a math question. Some people want the certainty of being done; others would rather hold the cash. Both are defensible.
Two things that beat the ordering entirely
If the minimum payment on an account is smaller than the monthly interest, the balance grows even when you pay on time. No payoff ordering fixes that — the only fix is more money against it, or a lower rate. Check this first if a balance has been stubbornly flat for a year.
And a rate reduction outruns any sequencing. A balance transfer at 0% for eighteen months, a credit-union consolidation loan, or a hardship rate from the issuer can each do more than perfect ordering ever will. They are worth an afternoon of phone calls before you spend two years optimising around a rate you might not have to pay.
The ordering matters. It matters considerably less than the amount you send and whether you keep sending it.