Amazon FBA due diligence checklist (2026)

Why verify from source data?

Signed LOIs fall apart over diligence more often than over financing: earnings that don’t hold, revenue concentrated in one listing, assets the seller does not own. By the time a deal is called off, it has often spent two to three months under exclusivity. That is the lower middle market as Axial’s Dead Deal Report reads it. No public data exists for e-commerce deals under $3 million.

An Amazon business makes the first of those problems easy to miss. When the seller’s P&L is a spreadsheet built from Seller Central reports, the bank sees something else: settlements on Amazon’s calendar, net of fees, refunds and often advertising. Diligence that starts from the seller’s spreadsheet checks the spreadsheet. Diligence that starts from Amazon’s payout records and the bank checks the business.

If you finance with an SBA 7(a) loan, the stakes are higher. Under SOP 50 10 8.1, historical earnings must cover debt service at 1.25x, and projections cannot make up the difference.

What should you ask the seller for?

Ask for sources, not summaries. Each one proves something the others cannot.

In what order should you do the work?

Diligence is cheapest before you commit and most expensive after. Put the checks that can end the deal first, and leave the ones that only move the price for later.

1. Before the LOI: the earnings and the ownership. Rebuild the P&L from payout records, match it to the bank, and run the coverage test on the verified figure. Ask the five ownership questions. A fatal finding here costs you nothing but the time to find it.
2. Before exclusivity ends: the rest of the six pillars. Unit economics, trend, concentration, supply, account health. These findings are mostly priceable or fixable, and they shape the final price and the transition plan.
3. With the lender: the formal file. Tax returns against IRS transcripts, the business valuation, and, at a purchase price of $3 million or more on an SBA loan, the quality of earnings report. The lender can pass the cost of these reports on to the borrower, and what the buyer spends on them counts toward the equity injection. The comparison with a quality of earnings report explains which one you need, and when.
4. Before closing: the transfer. The seller account, Brand Registry, the trademark assignment, the supplier relationship and the domain, each moved to you or your entity, with the documents to prove it.

What are the warning signs in a seller’s P&L?

None of these proves a problem on its own. Each one is a reason to go back to the payout records and the bank.

• Revenue that matches Seller Central but not the bank. Sales reported at order value, before Amazon’s fees, refunds and deductions, overstate what the business earns.
• Refunds that are missing or suspiciously low. Every Amazon business has them. A P&L without them is incomplete.
• Advertising booked as an add-back or left out. Ad spend is an operating cost of an Amazon brand, not a discretionary owner expense.
• Cost of goods from a price list, not from invoices. The landed cost includes freight, duties and the price the supplier actually charged.
• Add-backs without documents. Personal expenses, one-off costs and owner pay each need evidence.
• A trailing 12 months that ends on the best quarter. Compare it with the last two fiscal years before you price on it.

What should you check? Six pillars

The Scalerify Score weighs a brand across six pillars, out of 100. The weights below are the ones the Score uses; they are a reasonable order of attention for your own diligence too.

1. Financial health (25)

Start here, because every other number depends on it. Rebuild the P&L from Amazon’s payout records for the last two fiscal years and the trailing 12 months: gross sales, Amazon’s referral and fulfillment fees, refunds, reimbursements, advertising and other deductions. Then match each payout to a deposit in the business bank account. A payout that never reached the bank, or reached a different account, is the first question for the seller.

Set the declared figures against the rebuilt ones, line by line. Small differences are normal; a pattern of differences in one direction is not. Then list every add-back the seller claims, with its evidence. An add-back without a document behind it is a claim, and an SBA lender will not count it either: under SOP 50 10 8.1 the lender must justify each adjustment in its credit memorandum.

Finish with the coverage test on the verified figure, at 1.25x, with the loan you actually expect. The DSCR calculator shows the largest loan and price the verified earnings support.

2. Unit economics (20)

Total margin can hide products that lose money. For each of the top SKUs, work out the landed cost from supplier invoices and freight bills, then subtract Amazon’s fees, storage, returns and the advertising attributable to it. What is left is the contribution each unit makes.

Watch advertising closely. It can be a large share of sales: in our sample report ad spend ran at 46% of sales. Look at it month by month, and at margin with and without promotions and coupons. A brand that only sells at a discount is priced on its discounted margin.

3. Trend (15)

Look at monthly sales and margin over at least 24 months, so seasonality shows up twice. Separate the seasonal shape from real growth or decline. Explain every spike: a promotion, a competitor running out of stock, a price change, a one-time wholesale order. A trailing 12-month figure that includes one exceptional quarter overstates what you are buying.

Review count and rating on the top listings tell the same story from the customer side. A rating that drifts down while sales hold up is an early warning.

4. Concentration and platform (15)

Measure how much of revenue comes from the top listing and the top three, from each marketplace, and how much of purchasing goes to one supplier. Concentration is not a defect on its own. It is a risk to price. One listing suppressed, one supplier lost, one keyword lost to a competitor: each should be a scenario you have already run.

5. Supply and IP (15)

Check that the trademark is registered to the entity you are buying, and who administers Brand Registry. Ask who owns the formula, the mold and the tooling, and ask to see it in writing. Read the supplier terms: lead times, minimum orders, exclusivity, and whether a second source exists. Ask about supplier disputes, open or recent.

This is where the sample report’s finding sits. The manufacturer owned the formula: the buyer would have acquired a listing and a label, not a product.

6. Account health and CX (10)

Review the account health status and its history: warnings, suspensions, intellectual property complaints, policy violations. Read the return rate and the return reasons on the top listings, seller feedback and product ratings. It weighs least in the Score, but an account problem can still end a deal outright: a suspended account is a fatal finding.

Which findings end a deal? Fatal, priceable, fixable

Not every finding matters the same way. Grade each one before you decide what it means for the deal.

Who owns what? Five questions to ask before you sign

These are the questions that end deals after closing. Ask them before, and ask for the document that answers each one.

Doing it yourself, or having it verified?

Everything above can be done by a buyer with time, access and a spreadsheet. The financial health pillar is the slowest part: rebuilding a P&L from settlement data and matching it to the bank, payout by payout.

A Scalerify Verified report does that part from source data and covers the rest of the list: the P&L rebuilt from Amazon’s payout records and matched to the bank to the cent, a Score from 0 to 100 across the six pillars, a risk register with every finding graded, and the 1.25x coverage finding. It is prepared by an Amazon operator. See what a buyer gets, what it costs and how fast, or read the sample report first.

A Verified report is not an audit, a quality of earnings report, a valuation, investment advice or a commitment to lend.