SKU-level profit and loss reveals more, and it is not close. Channel-level P&L tells you that Amazon made money last month. SKU-level P&L tells you that Amazon made money last month despite four products that lost money on every unit sold, and it names them. If you sell more than about thirty products, the channel view will hide a losing SKU inside a profitable channel for as long as you let it.
Channel-level reporting still has a job. It is the right resolution for a specific set of decisions, and building SKU-level analysis when you only need the channel view costs you setup time and data hygiene you may not have to spare. The useful question is which decisions you are making this quarter.
What each view actually measures
Channel-level P&L aggregates revenue, fees, and cost of goods sold by sales channel. Amazon in one column, Shopify in another, Walmart in a third. It answers questions about channel economics: is Walmart worth the operational overhead, is our Shopify margin strong enough to justify pushing traffic there, did TikTok Shop pay for itself.
SKU-level P&L allocates the same revenue and costs down to the individual item. Every unit carries its landed cost, its referral fee, its fulfillment fee, its share of storage and returns. It answers a different question: which products should we keep.
The gap between them is arithmetic. A channel that nets twelve percent margin across four hundred SKUs is almost never made of four hundred products each netting twelve percent. It is made of a few products netting thirty, a long middle netting eight, and a tail netting below zero. Averaging hides the tail.
Why marketplace fee structures make this worse
Marketplace fees are not flat, which means two products with identical sale prices can carry very different fee loads. Amazon publishes its referral fee schedule publicly at sell.amazon.com/pricing, and the structure there is category-dependent by design. Home and Kitchen carries a fifteen percent referral fee. Electronics Accessories runs fifteen percent on the portion of the price up to one hundred dollars and eight percent above that. Clothing and Accessories is tiered at five, ten, and seventeen percent depending on price band. Most categories carry a thirty cent minimum referral fee per item.
Stack fulfillment, storage, and return processing on top of that and the per-unit cost spread across a catalog gets wide. A channel-level report averages all of it into one number. It cannot tell you that your thirteen dollar accessory is being eaten alive by the minimum referral fee and a fulfillment cost that does not scale down with price.
A decision framework
Use channel-level P&L when you are deciding:
- Whether to open or close a sales channel
- How to allocate ad budget across platforms
- Whether channel-specific overhead, such as a Walmart-dedicated ops hire, pays for itself
- What to show a lender or a buyer who wants revenue concentration by platform
Use SKU-level P&L when you are deciding:
- Which products to discontinue, reprice, or reorder
- Whether a supplier price increase can be absorbed or has to be passed through
- Where to spend inventory capital when you cannot fund every reorder
- Whether a bundle or multipack is actually more profitable than the single unit
Most sellers under roughly fifty SKUs on a single channel do fine with channel-level reporting and a periodic manual margin check. Past that, or once a second and third marketplace is in play, the manual check stops happening and the tail grows quietly.
The tooling split, honestly
The tools in this category are built around different jobs, and picking by feature list rather than by job is how sellers end up paying for two.
A2X sits at the settlement reconciliation end. It takes a marketplace payout and turns it into a summarized journal entry that reconciles to the deposit in your bank account, which is exactly what an outside accountant wants to see. It can calculate cost of goods sold per SKU per payout. One thing to know before you budget: A2X publishes its Amazon pricing openly, starting at twenty nine dollars a month for up to two hundred orders, but the COGS feature is excluded at that entry tier and starts on the fifty nine dollar plan. If COGS is the reason you are buying, price the Starter plan, not the Mini.
Sellerboard approaches it from the analytics side, describing itself as a profit analytics tool for Amazon FBA sellers, with published pricing starting at nineteen dollars a month billed monthly and product-level profit and loss on every tier. For a single-channel Amazon seller who wants per-product profit visibility and nothing else, Sellerboard is cheaper and faster to stand up than anything built around a general ledger, and it beats the accounting-first tools on time to first useful report. That is a real advantage and worth saying plainly.
The accounting-first platforms, including ConnectBooks, are built for a different situation: multiple marketplaces feeding QuickBooks Online, QuickBooks Desktop Enterprise, or Xero, where SKU-level profit reporting, automated COGS, and settlement reconciliation need to agree with each other and with the books. ConnectBooks publishes its per-product profit view at https://www.connectbooks.com/profits. The tradeoff is that this class of tool asks more of you at setup, because it will not produce a trustworthy SKU margin without accurate landed costs loaded first.
The input problem nobody solves for you
SKU-level P&L is only as good as your cost data, and that is where most implementations fail. If your cost of goods is the supplier invoice price and nothing else, your SKU margins are wrong by however much freight, duty, and inbound handling actually cost. The IRS guidance on figuring cost of goods sold for small businesses, in Publication 334, Tax Guide for Small Business, walks through what belongs in the calculation, including the treatment of inventory and the lines on Schedule C where it lands. It is worth reading once even if your accountant handles the filing, because the definition you use for tax should be the definition your reporting uses.
The second input problem is timing. A product purchased at one landed cost in March and restocked at a higher cost in July has two different margins depending on which units sold. Costing method matters here, and it is a conversation to have with your accountant rather than a setting to flip.
What to do this quarter
Pull your last full month of channel-level P&L. Then pick the twenty SKUs with the highest unit volume and calculate true per-unit profit by hand: sale price, less referral fee, less fulfillment, less landed cost, less an honest return allowance. If more than two of those twenty come out negative, the channel view is hiding real money and it is worth building the SKU-level reporting properly. If none do, your catalog is tight enough that channel-level is serving you, and you can revisit after the next round of product launches.
Either way, do the twenty by hand once. It is a tedious afternoon that tends to change what people believe about their own catalog.