Revenue is the easiest number in ecommerce to verify and the least informative one to own. What you are buying is the margin after landed cost, and the traffic that keeps producing it.
The report spine is the same as for SaaS — the sections that change are the two that have to. Churn becomes repeat purchase; customer concentration becomes supplier and SKU concentration. Everything else, including the conversion from asking price to a real profit multiple, works identically.
01
Repeat purchase rate and time between orders are the ecommerce answer to churn. A store where 80% of revenue is first-time buyers is an advertising business wearing a brand, and it is priced very differently.
02
Cost of goods is the easy part. Landed cost, shipping, returns, payment fees, app subscriptions and the ad spend it takes to move the next unit are where the margin actually goes. Gross margin on a listing is rarely the margin you inherit.
03
One supplier with no written agreement is a single point of failure you are paying a multiple for. Transferability of that relationship — and whether the terms survive a change of owner — belongs in the price, not in a footnote.
04
A hero SKU carrying most of the revenue is concentration in the same sense a whale customer is. It matters what happens when it is copied, when the trend passes, or when the supplier raises the price.
05
A store living on paid acquisition has costs that scale with growth and stop the moment the card does. Organic search, an email list and a real domain history are assets; a well-run ad account is a habit that has to be re-earned monthly.
06
Inventory is part of the purchase price and part of the risk. Ageing stock, the transfer mechanics, and whether the count reconciles to the sales history are ordinary questions that go unasked surprisingly often.
Take the price, strip the real cost base — goods landed, shipping, returns, apps, payment fees, the ad spend that is genuinely required rather than optional — and normalise the owner's hours to what an operator costs. The multiple that comes out is comparable. The one on the listing is not.
Where the public data cannot settle it, the report says so and lists the questions that would. A seller declining to answer one of them is itself information.