The sixty-second answer
Highest interest rate first costs the least. Smallest balance first gets finished more often. If the rates are close together, clear the smallest debts to build momentum; if one rate is far above the others, attack it first. Before either, hold enough cash to cover fixed costs through a slow period - a payoff plan that forces you to re-borrow has not achieved anything.
Two methods, both defensible
There are only two orderings worth considering, and the argument between them is older than any software.
The avalanche orders debts by interest rate and pays the highest first. It is arithmetically optimal: for a given total payment, it minimises the interest you hand over.
The snowball orders by balance and pays the smallest first. It costs more in interest. It also produces a debt that is fully gone earlier, and that matters more than spreadsheets suggest, because a plan abandoned in month four saves nothing at all.
Choosing between them honestly
The useful question is not which is mathematically superior - avalanche is, by definition. It is which constraint is actually binding for you.
If your rates are clustered within a few points of each other, the avalanche's advantage is small. In that case the snowball's earlier completion is worth more than the modest interest difference, and clearing the smallest balance also removes a minimum payment from your monthly obligations, which frees cash.
If one debt carries a rate far above the others - a card balance against a term loan, say - the arithmetic stops being close. Pay that one first regardless of its size.
The step both methods skip
Both approaches assume you should direct spare cash at debt. Often you should not, at least not yet.
The prior question is whether the business can cover its fixed costs through a slow stretch. Fixed costs arrive whether or not the month was good, and a business that has emptied its buffer to accelerate a payoff has swapped a known, priced obligation for an unpriced risk.
If clearing the debt forces you to borrow again in three months - very likely on worse terms, because emergency borrowing always is - the payoff has cost money rather than saved it. Hold the buffer first. Then accelerate.
Build the plan on real numbers
Every payoff plan contains one estimate: how much is genuinely spare each month. That figure is almost always guessed, and almost always guessed high.
The number you need is what the business actually spends in an ordinary month, separated into costs that move with revenue and costs that do not. A plan built on categorised spending survives contact with a normal month; one built on an optimistic recollection breaks in the first week and takes the plan's credibility with it.
That is where the tracking connects to the plan. Penny can answer what the business really spent by category and what the fixed floor looks like, so the amount you commit to debt is one you can sustain rather than one you hoped for.
Check the terms before accelerating
Two details are worth confirming before any extra payment. First, whether the loan carries a prepayment penalty, which can erase the interest saving on a shorter remaining term. Second, whether extra payments reduce the principal or are simply applied to future instalments - these are not the same thing, and only the first shortens the debt.
Interest on money borrowed to earn business income is generally deductible [1], which lowers the effective cost of carrying the debt. It rarely changes the ordering, but it does mean the real cost of a business loan is lower than its stated rate, and that is worth knowing before you starve the business to clear it.
Start with the list
Whichever method you pick, write down every debt with its balance, rate, minimum payment and prepayment terms. Most owners have never seen all of it in one place.
That list frequently settles the question on its own - because the rate spread turns out to be wider than assumed, or because one balance is small enough to clear this month and stop thinking about.