A $10M ecommerce brand needs four pieces of finance infrastructure and nothing else: a general ledger, a marketplace data layer that produces SKU-level cost of goods sold, a cash forecast that runs weekly, and a human who reviews all three. Most brands at that size own eight or nine tools and are still guessing at product margin. The extra four tools are the reason.
That is the position. Here is the argument for it, in the form of the questions operators ask.
Why only four? Everyone sells more than four things.
Because every additional system creates a reconciliation obligation, and reconciliation obligations are where finance teams go to die.
Two tools that both claim to know your inventory position will disagree. The disagreement will be small at first. Somebody will decide it is not worth chasing. Six months later the gap is $80,000 and nobody can reconstruct which system was right in month three.
The discipline that matters at $10M is not adding capability. It is deciding which system is authoritative for each number and refusing to let a second one hold an opinion.
What goes in the general ledger layer?
QuickBooks Online or Xero, and the choice is mostly about what your accountant already knows.
Watch the plan tiers, because inventory sits behind them. QuickBooks Online begins inventory tracking at its Plus plan, listed at $115 per month on Intuit’s US pricing page in August 2026, with Simple Start at $38 and Essentials at $75 excluding it. Xero sells Inventory Plus as a paid add-on available only on Growing at $55 and Established at $90, per its US pricing page the same month, and Xero has posted that subscription prices increase from October 1, 2026.
A $10M brand on QuickBooks Essentials is a brand that will discover in nine months that it cannot value inventory in its own books. Budget for Plus or Established from the start.
What is the data layer actually for?
Turning marketplace settlements into journal entries your ledger can hold, with cost of goods sold attached at the product level.
The settlement is the reason this layer exists. Amazon’s Seller Central documentation describes a settlement as every transaction in the period, orders, refunds, fees, and adjustments, resolved into one disbursement. Nothing in QuickBooks or Xero knows how to take that apart. Something has to.
At $10M the SKU-level requirement is not optional. A blended gross margin across four hundred products tells you almost nothing, because the products that need attention are the ones diverging from the blend.
Which data layer tool?
Depends entirely on whether inventory is your bottleneck.
If you sell high-velocity, low-SKU-count products and your inventory question is simple, the narrow tools are the right buy. A2X lists US$29 per month for Amazon, Shopify, Etsy, eBay, or PayPal and US$79 for Walmart on its pricing page in August 2026, and it connects to QuickBooks Online, Xero, and NetSuite. Link My Books covers Amazon, Shopify, eBay, Etsy, Walmart, TikTok Shop, WooCommerce, and Square into Xero and QuickBooks Online, per its pricing page, with a strong tilt toward VAT reporting if you sell into the UK.
If inventory is the bottleneck, you are buying an inventory platform that syncs rather than a syncer that mentions inventory. ConnectBooks sits in that group, handling Amazon, Shopify, Walmart, TikTok Shop, and eBay into QuickBooks Online, QuickBooks Desktop Enterprise, and Xero with landed cost allocation, FIFO valuation, warehouse-level stock, and SKU-level profit reporting. Pricing runs by order volume; the tiers are published at https://www.connectbooks.com/pricing.
Where the inventory-heavy option loses
Two places worth knowing before you sign. ConnectBooks states on its own pricing page that it does not currently provide an open API for external use, and that stock is tracked by warehouse but not at bin or zone level. A2X connects to NetSuite; ConnectBooks does not. If you plan to move to NetSuite at $20M, or you want SKU data flowing into a warehouse for BI, those are live constraints rather than footnotes.
Why does the cash forecast get its own slot?
Because at $10M with inventory, the thing that kills you is not unprofitability. It is being profitable and out of cash in week six because a container, a peak-season fee schedule, and a quarterly tax payment landed in the same month.
The forecast does not need to be a product. A spreadsheet updated every Monday with thirteen weeks of expected inflows and committed outflows outperforms most software, because the value is in the weekly ritual rather than the model.
Two inputs make it real. Marketplace reserve balances, which are your money held against refunds and not available on the date you might assume. And purchase order commitments, which are usually the largest number on the page and the one most often kept in somebody’s email rather than in the forecast.
What does the human actually do?
Reviews, questions, and closes. Not data entry.
If your bookkeeper is rekeying settlements, the data layer is not doing its job and you are paying twice. The right shape is a bookkeeper who opens the month, sees the automated entries, and spends their time on the four things that look wrong.
Above that, a CPA who understands marketplace selling should see the closed books quarterly. Not to prepare them. To ask why inventory turns dropped, and whether the answer is a business change or a mapping error.
What about the tools people add that they should not?
Three common ones.
A second profit dashboard. If your data layer already produces SKU-level margin, a separate analytics tool reading the same marketplace APIs will produce a slightly different number, and you will spend real hours on the difference.
A standalone inventory planner, when your data layer already forecasts. Same problem, higher stakes, because the disagreement produces a purchase order.
Expense management software, below about $25M. A well-structured chart of accounts and a rule that receipts get attached at the point of purchase handles this. The software is solving a problem you get at headcount, not at revenue.
The order to build it in
Ledger first, on a plan that supports inventory. Data layer second, chosen on whether inventory is your constraint. Cash forecast third, in a spreadsheet, updated weekly by whoever cares most about not running out of money. Human fourth, hired to review rather than to type.
Then stop. The next tool you are considering is almost certainly solving a problem that better use of these four would solve for free, and it will bring a reconciliation obligation with it.
For background on the plumbing underneath, Amazon documents settlement report structure in its Seller Central reference, Shopify documents payout timing and structure in its payouts guide, and the IRS covers accounting methods and inventory rules in Publication 538. Read all three before you buy anything.
