E-commerce Financial Due Diligence: The Metrics That Matter

E-commerce sellers pitch themselves in a language of their own: GMV, take rate, CAC, LTV, blended ROAS, contribution margin two. Founders quote these with confidence, decks are built around them, and every one of them can be defined in a way that flatters the business. The FDD analyst's job is to translate that vocabulary back into accounting reality - to find the revenue the company actually keeps, the profit that survives returns and fulfilment, and the customer economics that hold up once you separate the paid-for growth from the organic base. This is a metrics-heavy sector, and the metrics are only as honest as their definitions. Here is how to read them.

GMV versus net revenue: never confuse the two

The most common sleight of hand in e-commerce diligence is quoting Gross Merchandise Value (GMV) as if it were revenue. GMV is the total value of goods transacted across the platform. For a marketplace that connects third-party sellers with buyers, revenue is only the commission - GMV might be £100m while revenue is £12m. For an owned-inventory retailer that buys stock and sells it, GMV and gross revenue are much closer, but even then GMV usually excludes returns, cancellations and discounts.

Your first task is to lock down the revenue definition and reconcile it to the audited accounts:

• GMV → gross transaction value, before returns.
• Gross revenue → what the company invoices, net of cancellations.
• Net revenue → after returns, refunds and discounts. This is the analytical top line.
• Take rate → revenue as a percentage of GMV (marketplace models). A rising take rate is a sign of pricing power; a falling one signals competitive pressure or a mix shift toward lower-margin categories.

Always build your model on net revenue, and always know which model you are looking at, because owned-inventory and marketplace businesses have completely different margin structures, working capital profiles and risk. Confusing the two is the fastest way to misprice the deal, and it is a distinction that runs right through the underlying revenue quality work.

Returns and refunds: the revenue that walks back out

In many categories - fashion especially - a large share of what ships comes back. Return rates of 20–40% are normal in apparel. Returns hit the business three ways at once: they reverse revenue, they destroy margin (return shipping, inspection, repackaging, markdown or write-off of returned stock), and they distort every ratio built on gross sales.

• Confirm returns are accrued, not just recognised when they happen. Under IFRS 15 a business should provide for expected returns at the point of sale via a refund liability and a returns asset. A company that recognises the full sale and books returns only on receipt is overstating revenue and profit in any growing period.
• Rebuild net revenue net of the returns provision and test whether the provision matches the historical return rate.
• Watch for rising return rates as a quality signal - they can indicate sizing problems, quality issues or promotional buying that customers never intended to keep.

Ask one question and watch the founder's face: "Is your headline revenue before or after returns?" If the answer is "before", the entire deck needs re-basing - and so does the multiple. In high-return categories, the gap between gross and net revenue is not a rounding item; it is the difference between two different businesses.

Owned-inventory versus marketplace models

The business model determines almost everything else, so establish it early.

A hybrid that does both - its own products plus a third-party marketplace - must be split, because a blended gross margin is meaningless. The single most important structural question in e-commerce FDD is: who owns the inventory risk? It drives the working capital investment, which feeds straight into the net working capital analysis and the peg.

Unit economics: CAC, LTV and payback

This is the heart of an e-commerce QoE. The question behind every acquisition-driven online business is simple: does a customer generate more contribution margin over their life than it cost to acquire them, and how fast? The metrics that answer it are CAC, LTV and payback period, and each is routinely gamed.

• CAC (Customer Acquisition Cost) - total sales and marketing spend divided by new customers acquired. The trick to catch: dividing total marketing spend by total customers (including repeat buyers who cost nothing to re-acquire), which understates CAC. Insist on new-customer CAC using paid marketing spend only.
• LTV (Lifetime Value) - the contribution margin a customer generates over their lifetime. The trick: quoting LTV on revenue rather than contribution margin, or assuming an optimistic retention curve and discount rate. Insist on LTV built on real cohort margins, not projected ones.
• Payback period - months of contribution margin needed to recover CAC. A payback under ~12 months is generally healthy for a consumer business; beyond 18–24 months the model depends heavily on retention holding up.

A rigorous read here is really an exercise in unit economics - the same discipline you would apply to any subscription or consumer business.

Worked example: is the growth actually profitable?

The table below tests a direct-to-consumer brand claiming strong unit economics. Watch what happens when LTV is rebuilt on contribution margin instead of revenue, and CAC on new customers only.

The takeaway: a business pitched at a comfortable 3.2x LTV/CAC collapses to roughly 1.0x once returns are netted, LTV is put on a contribution basis and CAC counts only new customers. The company is buying revenue at close to break-even and calling it growth. Nothing here is fraud - it is definitional optimism, and unpicking it is precisely the value an FDD analyst adds.

Cohort retention: where the truth lives

Blended metrics lie; cohorts do not. Group customers by the period they first purchased and track their spend forward month by month. This reveals whether the business retains customers or simply refills a leaking bucket with paid acquisition.

• Flattening retention curves (a cohort that stabilises at, say, 25% of month-one revenue) indicate a genuine repeat base - high quality.
• Curves that decay toward zero mean the business has no organic engine and lives or dies on marketing spend - low quality, high risk.
• Improving newer cohorts suggest product-market fit is strengthening; deteriorating ones are a serious warning that recent growth was bought, not earned.

Cohort analysis also lets you split revenue into acquired-this-period versus repeat, which is the online equivalent of separating recurring from non-recurring income. A business whose growth is almost entirely first-time buyers acquired at rising CAC is far more fragile than the headline growth rate suggests.

Deferred revenue: gift cards and prepayments

E-commerce carries a specific deferred-revenue item: gift cards and store credit. When a customer buys a £50 gift card, the retailer receives cash but has delivered nothing - it is a liability, not revenue, until redeemed. Two things to check:

• Is the gift-card liability recognised correctly, and is revenue only booked on redemption? A business recognising gift-card sales as revenue on issue is overstating the top line and understating a real obligation.
• Breakage - the portion of gift cards never redeemed - can legitimately be recognised as revenue over time, but the estimate is judgemental. An aggressive breakage assumption inflates current profit.

Any subscription element (a "buy monthly" box, a loyalty membership fee) creates further deferred revenue that must be recognised over the service period, exactly as in the broader deferred revenue treatment. Prepaid subscriptions also flatter working capital - the customer funds the business in advance - so treat that cash carefully when building net debt.

Fulfilment, shipping and the real gross margin

The reported gross margin of an online retailer is almost never the economic one, because a large chunk of the cost of serving a customer sits below the gross-profit line or gets buried in operating costs. To get to a true contribution margin, you must load in:

• Fulfilment - pick, pack, warehousing, 3PL fees.
• Outbound shipping - often subsidised or offered "free", meaning the retailer absorbs it.
• Returns processing - inbound shipping, inspection, restocking, write-offs.
• Payment processing - card and gateway fees, typically 1.5–3% of every transaction.

A "50% gross margin" business is really earning ~21% contribution per order here - and it is that contribution, not the gross margin, that funds marketing and overhead. Presenting the buyer with a proper contribution bridge is one of the most valuable things an e-commerce QoE delivers, and it belongs in the EBITDA adjustments analysis when shipping subsidies or fulfilment costs are misclassified.

In the interview

E-commerce is a favourite because it rewards commercial fluency. A typical prompt: "A DTC brand is growing revenue 60% a year and the founder says unit economics are strong. What do you look at?"

"Growth of 60% tells me nothing until I know how it's funded, so I'd go straight to unit economics on a cohort basis. First I'd fix the definitions: net revenue after returns and discounts, not GMV, and new-customer CAC using paid marketing spend only rather than blended across repeat buyers. Then I'd rebuild LTV on contribution margin - loading in fulfilment, shipping, returns processing and payment fees - not on revenue, and test the LTV-to-CAC ratio and the payback period. Anything under about 3x with payback beyond 18 months makes me nervous. The clincher is the cohort retention curve: if older cohorts flatten out and keep spending, there's a real repeat base and the growth is durable; if every cohort decays to zero, the 60% is bought with marketing and it stops the moment spend stops. I'd also check the returns provision, gift-card deferred revenue and whether prepaid subscriptions are flattering working capital. My output would be an adjusted contribution margin, honest unit economics and a clear split between organic repeat revenue and paid-acquired revenue - because those deserve very different multiples."

The strength of that answer is that it refuses the founder's framing and rebuilds the metrics from accounting first. If you want to drill answers like this until they are automatic, the TS interview preparation plan is designed for it.

Common mistakes and red flags

• Quoting GMV as revenue. Always identify the model and work on net revenue; take rate is the bridge for marketplaces.
• Revenue booked before returns. In high-return categories, an unaccrued returns provision overstates revenue and profit throughout a growth phase.
• CAC on total customers. New-customer CAC on paid spend only is the honest number; blending in free repeat buyers understates acquisition cost.
• LTV on revenue not contribution. Only contribution-margin LTV can be compared to CAC; revenue LTV is meaningless.
• Ignoring fulfilment and shipping. Reported gross margin overstates real economics; build a contribution bridge per order.
• Trusting blended growth over cohorts. Decaying cohorts refilled by rising marketing spend is the classic leaky-bucket red flag.
• Gift cards recognised on issue. They are a liability until redeemed; aggressive breakage assumptions inflate current profit.
• Missing prepaid working-capital benefit. Subscription prepayments fund the business in advance and can flatter both working capital and cash.

The bottom line

E-commerce diligence is a translation exercise. The founder speaks in GMV, blended CAC and revenue-based LTV; the buyer needs net revenue, new-customer CAC and contribution-margin LTV, tested against a cohort curve that shows whether customers stay. Do the translation honestly and one of two things happens: you confirm a business with a real repeat engine and defensible economics, or you reveal that the growth was rented from an ad platform and stops the day the budget does. Online, the metrics are a story the company tells about itself - your job is to check whether the customers agree.

The Transaction Services Interview Programme (€119.99, one-time) includes a full e-commerce unit-economics module covering GMV-to-net-revenue reconciliation, cohort-based LTV/CAC, contribution bridges and gift-card deferred revenue, with modelling drills and interview cases. Enrol today.