featured article

Building a 2027 Ecommerce Budget From SKU-Level Actuals

Konstantine Christofakis

November 15, 2026

A 2027 ecommerce budget built from SKU-level actuals starts with each product's 2026 unit economics: units sold, net price, landed cost, fees, ad spend, refunds, and storage per unit. You then change only the assumptions you have evidence for, such as a new supplier quote, a planned price, a launch, or a discontinuation, and multiply out. Channel totals, the P&L, and the inventory purchase plan all roll up from those rows. A top-down budget that adds 15 percent to last year's revenue cannot show that the growth might cost margin. A bottom-up one shows it in the first SKU.

Why top-down fails for product sellers

A services business can budget revenue and headcount and be roughly right. A product business has a different cost structure on every SKU. A heavy item in a high fulfillment fee band, a light item with a 30 percent return rate, and a bundle with three components do not scale the same way. Applying one growth rate to all of them hides the fact that the SKUs growing fastest may be the ones earning least. The budget also has to produce a purchase plan, which is only possible by SKU, because the supplier ships units, not revenue.

Step 1: Pull 2026 actuals per SKU

For each active SKU and channel, twelve months:

  • Units sold, net of returns
  • Average net selling price after discounts
  • FIFO landed cost per unit
  • Marketplace and fulfillment fees per unit, as charged on settlements
  • Ad spend per unit
  • Refunds per unit
  • Storage per unit

Use what was charged, not the fee you assumed when you priced the product. The year-end SKU ranking from SKU rationalization produces most of this table already, and it also tells you which rows to delete before you budget them.

Step 2: Set volume one SKU at a time

Start each SKU at 2026 units and adjust for something you know. A product that stocked out for six weeks in 2026 gets those weeks added back at its normal velocity. A product with a new competitor gets a haircut. Discontinued SKUs get their sell-through units and nothing more. New launches go in as separate rows with their own assumptions, clearly labeled, so a miss on a launch does not hide inside the core catalog.

Step 3: Change cost assumptions only with evidence

Landed cost. Use the current supplier quote plus freight and duty for the next purchase order, not the 2026 FIFO average. If the quote went up, the budget should say so in January, not when the new layer starts selling in April.

Fees. Amazon publishes its next-year US fee update near year-end. Its update for 2026 was posted on Amazon's selling partners site on October 15, 2025, effective January 15, 2026. Until the 2027 update is out, budget a placeholder increase and label it. Two dated details from Seller Central matter for the first months of the year: holiday peak fulfillment fees run from October 15, 2026 to January 14, 2027, so the first two weeks of January 2027 still carry peak rates, and the 3.5 percent fuel and logistics surcharge on FBA fulfillment fees has applied since April 17, 2026.

Advertising. Budget per unit by SKU, not as a share of total revenue. A share-of-revenue rule hands the biggest ad budget to the product that already sells without help.

The worked example: one SKU, two years

Illustrative numbers. Take a seller's top SKU on Amazon.

2026 actuals

  • 12,000 units at a net price of $29.50
  • Landed COGS: $9.20 per unit
  • Fees: $9.10 per unit
  • Ads: $2.95 per unit
  • Refunds: $1.18 per unit
  • Storage: $0.40 per unit

Unit contribution: $29.50 minus $9.20 minus $9.10 minus $2.95 minus $1.18 minus $0.40 equals $6.67. Times 12,000 units: $80,040.00.

2027 budget, first pass

  • Units up 8 percent: 12,960
  • Price held at $29.50
  • Landed COGS from the new supplier quote, freight included: $9.65
  • Fees with a 2 percent placeholder increase: $9.28
  • Ads, refunds, and storage held per unit

Unit contribution: $29.50 minus $9.65 minus $9.28 minus $2.95 minus $1.18 minus $0.40 equals $6.04. Times 12,960 units: $78,278.40.

The budget sells 960 more units and earns $1,761.60 less. That result is the reason to budget by SKU. A top-down plan would have shown this product contributing 8 percent more.

Second pass: what price holds 2026 dollars?

The seller needs $80,040.00 divided by 12,960 units, which is $6.18 per unit, $0.14 more than the first pass. Each extra dollar of price adds about $0.85 of contribution after a 15 percent referral fee, the rate for most categories on Amazon's sell.amazon.com pricing page. So the price has to rise about $0.16 to $29.66, assuming refunds and volume hold. The seller tests $29.99 instead. At that price, unit contribution rises by $0.49 times 0.85, which is $0.42, to $6.46, and 12,960 units would earn $83,721.60.

That last figure assumes volume does not fall at the higher price. It might. How much it falls depends on the product's price elasticity, and the honest budget carries the $29.99 row with a lower unit count until a test says otherwise.

Step 4: Keep fixed costs below the line

Salaries, rent, software, and professional fees go in as a separate block under total contribution. Phase them by month as you expect to pay them. They do not belong in SKU rows, where an allocation would make small products look unprofitable for reasons that have nothing to do with the product.

Step 5: Turn the budget into a cash plan

The P&L budget says what you will earn. The purchase plan says when the cash leaves. For each SKU, budgeted units by month, less inventory on hand and inbound, plus a safety buffer, gives units to order, and the supplier's lead time tells you when to place the order. ConnectBooks' restock report works from sales history and velocity with lead times and inbound stock factored in, and purchase orders can be created in the platform, so the budget's purchase plan can be checked against the inventory position each month. Settlement timing goes on top: marketplace cash arrives on each platform's disbursement cycle, not on the day of sale. The full method for the weekly view is in how to build a 13-week cash flow forecast.

Step 6: Phase by month

Use 2026's monthly shape per SKU, adjusted for known calendar effects. Cyber Monday falls on November 30 in 2026 and November 29 in 2027, so both years keep it in November, but promotions you ran in a different month this year belong where you will run them next year.

Where AI fits, and where it does not

Crunch, the AI analyst inside ConnectBooks, is useful for the questions behind the assumptions. It can tell you which products had better margins in some months than others and why, or what changed on a SKU whose contribution fell, working from the seller's reconciled data. It does not build the budget, and it does not forecast 2027 demand. It cannot know your supplier quote, your launch calendar, or a competitor about to enter your category. If landed cost is missing on a row, any margin it reports for that row is wrong. The assumptions, and the decision to commit cash against them, stay with you and your accountant.

Take Control of Your E-Commerce Business with ConnectBooks

Running an e-commerce business comes with plenty of challenges, but ConnectBooks is here to make your life easier. With real-time insights, seamless integrations, and detailed tracking of your profitability and inventory, you can stay ahead of the game. Whether you’re selling on Amazon, Shopify, Walmart, TikTok or eBay, ConnectBooks helps you manage your finances with 100% accuracy and confidence, so you can focus on growing your business.

Ready to level up? Start making smarter, data-driven decisions every step of the way. Try ConnectBooks Free Today or Schedule a Demo