A budget for an Amazon business is a forecast of units, prices and costs per SKU, compared every month with what actually happened. The comparison is where the value is: it shows whether a profit shortfall came from selling fewer units, selling at a lower price, or paying more per unit in fees and advertising. Each cause needs a different fix, and a single “profit is down” number does not tell you which one you have.
This is FP&A (financial planning and analysis) scaled to an Amazon seller: a driver-based forecast, a monthly variance check, and a rolling reforecast.
What goes into an Amazon forecast?
Forecast drivers, not totals. A driver is a number you can observe and influence directly. For each main SKU:
- Units sold per month, from recent sales velocity, adjusted for season and planned promotions.
- Average selling price.
- Landed COGS per unit (product cost plus freight, duties and prep).
- Amazon fees per unit: referral fee, FBA fulfillment fee, and a storage allocation.
- Advertising cost per unit, or a TACOS target (ad spend ÷ total revenue).
- Refund rate and the cost of each refund.
Then account-level fixed costs: staff, software, accounting, the owner’s salary.
Forecast net profit = Σ (Units × Net profit per unit) − Fixed costs
Group smaller SKUs together. A forecast with 15 lines you update every month is more useful than one with 400 lines you stop maintaining in March.
How do you read budget vs. actual?
Split every difference into its causes. Example for one SKU, one month:
| Budget | Actual | |
|---|---|---|
| Units | 3,000 | 2,700 |
| Average price | $25.00 | $26.00 |
| Revenue | $75,000 | $70,200 |
| Net profit per unit | $5.00 | $4.60 |
| Net profit | $15,000 | $12,420 |
Revenue variance (−$4,800):
Volume effect = (2,700 − 3,000) × $25.00 = −$7,500
Price effect = ($26.00 − $25.00) × 2,700 = +$2,700
Profit variance (−$2,580):
Volume effect = (2,700 − 3,000) × $5.00 = −$1,500
Per-unit effect = ($4.60 − $5.00) × 2,700 = −$1,080
The price went up by $1.00, yet profit per unit went down by $0.40. The per-unit bridge shows why:
| Per-unit line | Budget | Actual | Change |
|---|---|---|---|
| Price | $25.00 | $26.00 | +$1.00 |
| COGS | $8.00 | $8.00 | — |
| Referral fee (15%) | $3.75 | $3.90 | −$0.15 |
| FBA fulfillment fee | $4.50 | $4.50 | — |
| Advertising per unit | $3.00 | $4.10 | −$1.10 |
| Storage and refunds | $0.75 | $0.90 | −$0.15 |
| Net profit per unit | $5.00 | $4.60 | −$0.40 |
Advertising per unit rose by $1.10. So the month’s story is: the price increase cost some volume, and ad spend rose while units fell. The action is on advertising efficiency and on whether the price increase was worth the lost units, not on fees or COGS.
How often should you reforecast?
Monthly, as a rolling three-month forecast. After each month closes:
- Record actuals per SKU.
- Run the variance split: volume, price, per-unit cost.
- Update the drivers for the next three months with what you learned. If advertising per unit has been higher than budget for two months, it is the new baseline, not a one-off.
- Check the cash side: purchase orders due, inventory arriving, loan repayments.
Keep the annual budget unchanged as the reference point, and let the rolling forecast move. Comparing actuals with both shows how far plans have drifted and where the year is heading.
What makes Amazon forecasting different?
Fees change during the year. Amazon adjusts fulfillment fees and surcharges, and peak-season storage rates apply in Q4. Put known changes into the forecast for the months they take effect.
Inventory is bought months before it sells. The profit forecast and the cash forecast differ by the lead time of your supply chain. A strong Q4 forecast means cash goes out in Q3.
Stock limits sales. If a SKU will be out of stock for two weeks, the forecast should show it. A volume variance caused by a stockout is an inventory planning problem, not a demand problem.
Payouts don’t match the P&L. Amazon settlements arrive on a schedule that doesn’t line up with calendar months, and they net out fees, refunds and reserves. Forecast on sales-date accounting and reconcile payouts separately.
Best practices
- Forecast per SKU for the products that make up most of your profit; group the rest.
- Use net profit per unit as the core driver, built from COGS, fees, ads and refunds.
- Split every variance into volume, price and per-unit cost before deciding what to change.
- Reforecast monthly on a rolling three-month basis.
- Keep a separate cash forecast tied to purchase order dates and lead times.
- Write one sentence per month explaining the largest variance. Over a year, these notes show patterns the numbers alone don’t.
Common mistakes
- Budgeting revenue only. A revenue target says nothing about whether the plan is profitable.
- Forecasting advertising as a fixed monthly amount while volume changes. Ad cost per unit is the useful driver.
- Treating a stockout month as weak demand.
- Reforecasting to match actuals every month, which removes the reference point.
- Using Amazon payouts as monthly revenue.
FAQ
What is budget vs. actual analysis? Comparing what you planned (units, prices, costs, profit) with what happened, and splitting the difference into causes so you know what to change.
How detailed should an Amazon seller’s forecast be? Per SKU for the products that drive most of your profit, grouped for the rest, with fixed costs at account level.
What is the most important number to forecast? Net profit per unit per SKU. Revenue follows from units and price; profit depends on every cost per unit.
How do I forecast Q4? Use last year’s Q4 sales pattern as a base, adjust for this year’s trend, include peak-season fees, and check that inventory will arrive in time. Place the purchase orders in the cash forecast for the months they are paid.
Do I need FP&A software? A spreadsheet is enough for most Amazon businesses, provided the actuals feeding it are complete and per SKU.
Conclusion
A budget is only useful if you compare it with actuals and understand the difference. For an Amazon business that means per-SKU drivers, a variance split into volume, price and per-unit cost, and a rolling reforecast that absorbs what each month teaches. An hour a month is enough once the actual figures are easy to get.
sellerboard provides those actuals per product and per month: units, average price, Amazon fees, advertising, refunds and net profit, in a form you can export into your forecast spreadsheet each month.
Figures in the example are illustrative.
