Peak season budget planner
Every November budget has a point where the next dollar stops paying for itself. Enter your plan, your margin, and today's ROAS, and see week by week where that point is, including BFCM week.
Weekly mode raises daily spend by the ramp each row. BFCM week has its own higher stop level when demand lift is above zero.
| Week | Spend | Revenue | Avg ROAS | Marginal ROAS | Stop at |
|---|---|---|---|---|---|
| Week 1Nov 1–7 · 7 days | $10,000 | $35,000 | 3.50 | 2.45 | $13,844 |
| Week 2Nov 8–14 · 7 days | $11,500 | $38,597 | 3.36 | 2.35 | $13,844 |
| Week 3Nov 15–21 · 7 days | $13,225 | $42,564 | 3.22 | 2.25 | $13,844 |
| Lead-inNov 22–25 · 4 days | $8,691 | $26,822 | 3.09 | 2.16 | $7,911 |
| BFCM weekNov 26–30 · 5 days | $12,493 | $48,066 | 3.85 | 2.69 | $23,711 |
| November | $55,909 | $191,050 | 3.42 | $73,154 |
Marginal ROAS in green clears breakeven; red rows are past the stop point. "Stop at" is the row spend where marginal ROAS equals 2.22.
The math is the easy part.
The calculator gives you the number. We send what moves the inputs — from accounts we run every day.
The stop rule
Keep spending while the next dollar returns at least breakeven ROAS, and stop when it does not. Breakeven ROAS is 1 divided by your contribution margin: at a 45% margin it is 2.22. The rule is about the next dollar, not the average. By the time average ROAS reaches breakeven, you have already spent a long stretch of money that lost contribution.
Past the stop point, revenue still goes up. That is what makes November dangerous: the top line keeps growing while contribution after ads shrinks. The planner shows both, so the decision is about what you keep rather than what you book.
How the curve is built
Revenue follows a diminishing-returns curve, revenue = a × spend^b, with b between 0 and 1. The planner calibrates a so the curve passes through today: at your current daily spend it returns your baseline ROAS. The curve is applied per day, so a four-day week is not penalized for being short. BFCM week multiplies the curve by your demand lift.
daily revenue = lift × a × (daily spend)^b
a = baseline ROAS × (current daily spend)^(1 − b)
average ROAS = revenue ÷ spend
marginal ROAS = b × average ROAS
breakeven ROAS = 1 ÷ contribution margin
stop spend/day = (lift × a × b × margin)^(1 ÷ (1 − b))
past the stop = marginal ROAS < breakeven ROASMarginal ROAS is simply b times average ROAS on this curve. With b at 0.7 and a 3.5 average, the last dollar returned 2.45. That one line is the whole case against judging November on average ROAS.
The weeks
November 2026 is split into five rows that cover every day of the month: Nov 1 to 7, Nov 8 to 14, Nov 15 to 21, a four-day lead-in from Nov 22 to 25, and BFCM week from Thanksgiving on Thursday Nov 26 through Cyber Monday on Nov 30. Demand lift applies only to BFCM week. If your lead-in usually softens as shoppers wait for the deals, the plan will slightly overstate that row.
What the model leaves out
One curve stands in for every channel, so it cannot tell you which channel saturates first. It holds margin constant, so enter the margin after your planned BFCM discount, not your everyday one. And it counts first-order revenue in the week it happens. Customers acquired in November who return in the new year are worth more than this table shows, which is a reason to push slightly past the stop point on purpose only if you know your repeat rate.
What is marginal ROAS and why does it matter more than ROAS in November?
Average ROAS is revenue divided by all the spend. Marginal ROAS is the revenue the last dollar bought. As spend rises, returns diminish, so marginal ROAS falls faster than the average. A week can report a healthy 3.0 average while its last few thousand dollars returned 1.8. If breakeven is 2.2, those dollars lost money even though the dashboard looked fine.
How is the stop point calculated?
Breakeven ROAS is 1 divided by contribution margin. The stop point is the spend level where marginal ROAS falls to breakeven. Below it, one more dollar returns more contribution than it costs. Above it, one more dollar costs more than it returns. The planner shows that spend level for each week and flags the first week the plan crosses it.
What elasticity should I use?
If you have raised spend before and watched revenue, use your own number: b equals the natural log of the revenue ratio divided by the natural log of the spend ratio. If spend went up 50% and revenue went up 30%, b is ln(1.3) ÷ ln(1.5), about 0.65. Without a test, 0.7 is a starting assumption, not a benchmark. Try 0.5 and 0.85 to see how sensitive your plan is.
Why does BFCM week get a different stop point?
Because the same spend buys more revenue when demand peaks. The planner multiplies the curve by your BFCM demand lift from Thanksgiving through Cyber Monday, which raises both ROAS and the spend level where marginal ROAS reaches breakeven. Use your own lift from last year. If you enter zero, BFCM week is treated like any other.
Should I use a total budget or a weekly ramp?
Use the total mode when finance has already set a November number and you want to know how to split it and whether it is too much. It splits spend so every week has the same marginal ROAS, which is the most revenue a fixed budget can buy on this curve. Use the weekly mode when you plan by stepping spend up each week and want to see the week where the ramp overshoots.
Want these numbers computed from your actual account?
The free audit reads your real data and shows where the economics leak — no generic benchmarks, no sales theater.