The 24-Month Debt Payoff Plan: Set the Payment, Freeze It, Survive the Months That Break It — With 2026 APRs

You have the list. The order question has its answer elsewhere on this site, counted in full. What nobody has handed you is the plan: the payment, the table, the months where the plan gets tested, and the checkpoints that tell you it is still alive. This article builds one. One $16,800 list at dated 2026 rates, one $835 payment, twenty-four months, every row recomputed from the numbers printed here — and the logic to rebuild the whole thing on your own list in an evening.

The debt is not a moral failure and this is not a scolding. It is ten numbers — three balances, three rates, three base payments, one budget — and a calendar. Let’s lay them out.

Why twenty-four months

A payoff plan needs a horizon the way a trip needs a destination, and two years is the horizon that survives contact with a real household. Twelve months asks for luck twice; thirty-six months asks for patience nobody has tested yet. Twenty-four is long enough that the interest math bites — you cannot borrow your way out of a bad payment for two years at 2026 card rates — and short enough that you can put the finish date on a wall and watch it approach.

It is also the one horizon you can build backwards from a formula. The payment that clears a debt in exactly 24 months, at a fixed rate, is not a guess; it is arithmetic, and it goes in one line:

payment to clear principal P in 24 months at rate APR:
  i   = APR / 12                      the monthly rate
  PMT = P x i / (1 - (1 + i)^-24)     the same number every month

That formula is the whole planning tool. Everything below is it, run on a real list, plus the parts the formula cannot tell you — which order, which months break, what a new bill does.

The list

One household, three debts, $16,800, October 2026. Each rate below is the APR — the annual percentage rate, what that debt costs a year, the rate — and every one of them is a published number the site has already dated: the Federal Reserve’s G.19 Consumer Credit release of September 8, 2026, the second-quarter averages that the avalanche-vs-snowball article, the payoff calculator and the minimum-payment article all run on. This run could not re-pull the Federal Reserve’s own pages — they did not open, and I will not cite what I could not read this week — so the rates are carried from those dated pages, linked, not reprinted. The rate is the number that decides, so it carries its date.

The scenario is a scenario — the balances are mine, chosen so every rate on it is a dated federal average and the list is nobody else’s list: not the $25,000 four-line schedule the comparison runs, not the single $5,000 card the minimum-payment article runs to death. Your list has different rows. The machine below does not care.

Every base line on this list earns one sentence, because none of them was typed off a statement as-is. A card’s minimum is a formula — on the site’s model, the greater of $25 or 3 percent of the new balance — and the plan freezes it. Card A’s first minimum, computed in month one, is $99.68 of interest plus the percentage part landing at $164.99; the plan sends $165 every month, whether the formula agrees or not. That freezing is the whole difference between this and the minimum-payment article next door: that one runs a card on the live formula and counts 200 months and $6,829.98 on $5,000 — because the minimum shrinks as the balance shrinks. A plan does not shrink. The $165 stays. Read that article for why; this one is what you do after.

The loan’s line is not a minimum at all. It is the contract payment for $7,800 at 11.86 percent amortized over exactly 24 months — the formula above with P = 7,800 and i = 0.0098833: $366.66 a month, rounded to $367 in the plan. It will die at month 24 no matter what the plan does to it. That matters later.

The payment: where an $835 line comes from

Three rates, three base lines — $165, $110, $367. That is $642 to keep the accounts conforming; the plan sends $835, and the $193 difference is the attack. The first honest question is not which order. It is whether the household can send $835 for twenty-four straight months, because that number, not the order, is the plan.

Put it against dated money. The Bureau of Labor Statistics’ median monthly earnings for full-time workers aged 25 to 34 — $5,027 gross in Q2 2026, dated on this site in the first-paycheck sequence and the reserve calculator — makes $835 16.6 percent of gross. The same article’s take-home under those brackets — $4,220.86, which I re-derived this week from the Rev. Proc. 2025-32 tables pulled off irs.gov: $60,324 less the $16,100 standard deduction, through the 10 and 12 percent brackets — makes the line 19.8 percent of what actually lands. And here is the coincidence the budget people will spot first: the 50/30/20 stress test runs the printed rule on a $60,000 salary — $5,000 a month, its own words, “a hair under” that median — and prices its 20 percent column at $839.83 a month. This plan fits inside the rulebook’s saving column with $4.83 to spare.

Which is exactly why the stress test matters more than the fit. That article’s finding is that at dated 2026 rents the needs column runs 69 percent in a calm month and 90 percent in a bad one — the 30 percent and the 20 percent do not both fit, and the card is the gap’s favorite way in. So the plan’s first step is not the table. It is the line: find $835 in the household, name what it replaces, and defend it. The order question is worth a few hundred dollars on a two-year list, counted below; the payment question is worth thousands. Get the line from the stress test, then come back.

One more rule before the table, and it belongs to the payoff lane, not the budget lane: no line in a 24-month plan may run uphill. A payment under the month’s interest does not pay the debt — it feeds it. Card A at $165 covers its month-one interest of $99.68 with $65 to spare; Card B at $110 covers $62.82; the loan at $367 covers $77.09. Three lines, all moving from month one. The calculator page prints the cliff this rule guards: a flat $25 minimum loses to the interest above about $1,354 of balance at 22.15 percent. A plan made of uphill lines is a spreadsheet, not a plan.

The order: why the choice you agonize over is the second choice

Here is the part the accepted debt literature on this site lets me say plainly, because it ran the numbers both ways twice already. On this list, at this budget, run to the last payment:

• The avalanche — Card A first (22.15), then Card B (20.94), the loan riding its own contract line: 24 months, $3,132.94 of interest.
• The snowball — Card B first ($3,600, the smallest), then Card A, same loan line: 25 months, $3,265.69 of interest.
• No order at all — every line on its own 24-month formula payment ($280.54, $184.88, $366.66 — $832.08 a month total): $3,170.18 of interest, and twenty-six cents still owed at month 24 — the rounding cents no formula payment quite finishes.

The order, on a twenty-four-month horizon, is worth $132.75 and one month. The payment is worth $1,683.17 and ten months — because the same list on nothing but the three frozen minimums, $641.53 a month with every freed dollar rolled back in, runs 34 months and $4,816.11, and its first victim is the loan, at month 25. That is the finding, and it is the same finding the calculator’s budget table prints at the other end of the range — where a $666 budget with zero extra makes the two orders $2.71 apart, “a rounding error, because there is barely any money for the order to point anywhere.” Run that list longer and the order compounds into the four figures the avalanche-vs-snowball article counts: $1,078.50 on a schedule that runs past three years. Two-year plans are short enough that the order barely gets time to earn its keep.

So the plan sets the payment first and the payoff order second — the order is still free money once the payment is decided: on this list the avalanche costs $37.24 less than no-order-at-all and a month less than the balance order. What the balance order buys for its $132.75 is its first win four months sooner — Card B dead in month 14 instead of Card A in month 18 — and the comparison article carries the research on why that earlier win moves people. Four months of win for a hundred and thirty-two seventy-five, on a list this short: both orders are defensible here, which is the point. The avalanche is the math, the snowball is the follow-through, and on a plan this length the math says take either one that makes you send the $835 for twenty-four months. The one you finish is the one that counts.

One federal footnote while we are on orders, because it surprises people at the kitchen table: if two of these balances live on the same card, you do not choose the order for the money above the minimum — the rule does. Regulation Z, 12 CFR 1026.53, read from the official 2025 CFR edition on govinfo this week: when you pay more than the minimum on a card, the issuer must allocate the excess “first to the balance with the highest annual percentage rate and any remaining portion to the other balances in descending order.” Unless you ask it not to — the rule lets you — a card with two balances runs the avalanche by default. The order you choose is the order across cards.

The 24-month plan, month by month

The mechanics, stated once so every row is checkable. Payment day the 15th, first payment October 2026. Interest charges on each open balance at its APR ÷ 12, rounded to the cent. The three base lines go out of the $835 first. What is left of the $835 goes at the top of the rate order. When a line dies, its payment does not go anywhere — it joins the attack, and the outflow stays $835 until the final month needs less. That budget-whole convention is the same one the payoff calculator runs, so the table below and that box speak the same arithmetic; put your own list into the box and check mine.

Twenty-four payments, last one September 15, 2028, $3,132.94 of interest, $19,932.94 out on $16,800 borrowed. Every cell is the formula block above run in cents; the balances round to the cent once, on the interest, the same single rounding the calculator states as its convention.

Read two things in it before we go on.

The interest curve is the motivation section. Month one feeds the bank $239.59; month 24 feeds it $9.85. By month 6 the plan has paid $1,308.03 of interest — 42 percent of everything the plan will ever pay — while retiring only 22.0 percent of the principal. By month 12: 73 percent of the interest, 46.0 percent of the principal. The plan pays the bank hardest exactly when the balances have barely moved, and that is what the extra money is for: every dollar of extra in months 1–6 is buying down the base the whole next year’s interest gets computed on.

Card B outpays Card A. The interest table by debt: Card A, the 22.15 percent card, carries $987.56 across the run. Card B, the cheaper 20.94 percent card, carries $1,146.44 — more. Because the avalanche aims everything at A, B gets attacked only from month 19 — eighteen months of compounding with nothing but its own base line against it. The rate is the number that decides the order; the total interest is decided by how long each balance gets to compound, and the plan decides that. It is not a flaw in the order — it is what any order does to the line it waits on, and it is why the extra, not the sequence, is where the money is.

The checkpoints: month 6, month 12, month 24

A plan without checkpoints is a hope with a spreadsheet. On this plan, the three rows that tell you whether it is alive:

Put the two balance numbers on the calendar itself, not in the spreadsheet: March 2027, owe $13,098. September 2027, owe $9,067. When a month goes sideways, you check the date against the checkpoint, and the checkpoint tells you how bad “sideways” actually was — which is the difference between a detour and a quit.

Build it on your list: the template logic

The plan above is disposable; the machine that made it is the article — which is also the straight answer to how do I make a debt payoff plan, and to anyone who typed “debt snowball spreadsheet” into a search box to get here: the template is five steps and one formula, rebuildable in an evening. Five steps, each with the number it produces:

1. List them. Every debt: balance, APR, contract line. Dated rates — the APR on your statement, and for averages the G.19 release with its date, like every schedule on this site. Rates move mid-plan; when yours moves, re-run, same as the dated-schedule rule states.
2. Freeze the minimums. Compute each card’s month-1 minimum once and pin it: on the site’s model, the greater of $25 or 3 percent of the first new balance — Card A’s $164.99, Card B’s $109.88. Loans keep their contract line. The plan sends these whether the formula shrinks or not; the next-door article is what the unfrozen version costs.
3. Price the horizon. Run the 24-month formula — PMT = P x i / (1 - (1+i)^-24) — on each line and add: on this list, $280.54 + $184.88 + $366.66 = $832.08. That is the honest price of two years. Round the plan up from it (here, $835), never down.
4. Test the line against dated household money — gross and take-home, not a feeling — and fund it in the stress test before anything else. Also test each line uphill: payment under the month’s interest is not a plan line.
5. Aim, then run it once and read the checkpoints. Extra at the highest APR (the avalanche; the smallest-balance order if that is the one you will finish — both, counted), freed payments rolled into the attack, budget held whole. Then the month-6 and month-12 targets fall out of the table and go on the calendar.

If step 5’s engine is not a spreadsheet you want to build: the payoff calculator on this site is this machine, in the browser, with your four lines on the input side and both orders on the output side — run your list through it and the plan’s table is the printout. What the box does not do is the parts below, which is what a plan is actually made of.

When the plan gets tested

Three things happen inside any 24 months. Each one gets an answer in this plan’s own arithmetic, not in motivation.

A month breaks. You cannot send the $835. Send the three base lines and skip the extra, and the plan lands at 25 months, $71.70 more interest. Send nothing at all that month and it lands at 26 months, $248.06 more. Two zero months and it is 27 months, $423.91 more. The plan absorbs a broken month for the price of one dinner out per month of slippage — if the rule was written down first: a bad month sends the base lines and rolls the extra; a bad month is a checkpoint to re-read, not a verdict. The plan is the $835 and the order and the targets. A month is a month.

New debt happens. A $600 repair goes on Card A in month 10 — the plan holds the $835 line and the same order, and the finish moves to month 25, $183.10 more interest. That is the whole damage, counted. Now watch what “being careful” does: pausing the extra for three months to “build the cushion back” instead lands the plan at 26 months and $345.92 more interest — nearly double the price of just carrying the bill inside the plan. The attack is the cheap part to keep. The cheap absorber is money, not restraint: a mini-reserve, or a $50 add-on for twelve months. That is why the emergency-fund calculator is part of this plan and not a different life — its worked floor is $3,784 of essential outgo, and even a deliberate few-hundred-dollar mini-cushion sitting under the card turns most $600 months back into no-charge months. A payoff plan with zero cushion in it is a plan with a timer on it.

A payment frees. When Card A dies at month 18, its $165 has to go somewhere, and this is where plans quietly die: the freed line drifts to the checking account and the household starts its next debt at month 19. The rule is in the mechanics above and it is not optional — the outflow stays $835 until the last month needs less; every freed payment joins the attack. That is the snowball’s own mechanism, and on this plan it is what turns month 23’s $718.09 into a two-line, same-month finish. It is also the convention worth understanding, because schedules disagree without meaning to: the calculator page reconciles the two schedules this site prints on one list — same debts, same rates, and when the freed money keeps running instead of leaking out of the household, both orders land sooner and cheaper. Money that used to feed a debt has exactly one job in this plan.

And the one that does not get a paragraph: if the same list keeps getting used — new charges on a card the plan is paying down — none of the above applies, because there is no plan. The list above assumed no new charges, the way every schedule on this site states its assumption. Close the spending before you start the paying; that is the only behavioral rule here that is not a number.

Does the plan survive 24 months: the checklist

Before the first payment, six questions, each answerable with a date or a number:

1. Is the line funded before it is promised? $835 named against dated income (here: 19.8% of the site’s dated $4,220.86 take-home), with what it replaces written down. The stress test prices what happens when a line is promised to a rulebook that already does not fit.
2. Does every line beat its own month-one interest? Here: $165 beats $99.68, $110 beats $62.82, $367 beats $77.09. One uphill line and the plan is decoration.
3. Is there cushion under the cards? Even a mini-reserve — the calculator’s floor is $3,784, its worked fill-line is $503 a month — turns a $600 month from a $345.92 decision into a $0 one: the cushion pays, the attack never pauses.
4. Are the two checkpoints on a calendar? March 2027: owe $13,098. September 2027: owe $9,067. Dates, not rows.
5. Is the broken-month rule written down? Base lines ship in a bad month; extra rolls; re-read the checkpoint; the horizon moves one month and the plan does not end.
6. Is the payoff order picked, on purpose? The avalanche on paper; the balance order if the first win is what keeps you sending it — both ways are counted, and on a two-year list they are $132.75 apart. The plan you finish is the plan that counts; the order you finish is the order that’s yours.

If all six answer yes, the plan survives 24 months. The math was never in danger; the months were.

The decision

One list, one line, twenty-four months: $835 a month, Card A first, freed payments pressed into service, $3,132.94 of interest, done September 2028 — and rebuilt in five steps on whatever list you actually have. The order was worth $132.75; the payment was worth $1,683.17; the months that break are priced at $71.70 to $423.91 and survive inside the plan if you wrote the rules first. On this list, the execution is the strategy — which is the one thing the avalanche-vs-snowball debate quietly assumes and never says.

Run your own four lines through the payoff calculator, take the payment fight to the stress test, put a cushion under the whole thing with the reserve calculator, and settle the order with the full comparison — the four pages are one machine, and this is the page where it gets turned on.

List them, pick the order, start.