In e-commerce, revenue and advertising ROAS can rise while economic value deteriorates.
In e-commerce, revenue and advertising ROAS can rise while economic value deteriorates. The mix must connect net price, basket, acquisition cost, product margin, payment, fulfilment, delivery, returns and service. The decision unit is incremental contribution per order or cohort, not platform-attributed revenue.
Decision and method.
Scale a campaign, promotion or delivery rule only if the additional cohort covers every variable cost and remains profitable after returns. Freeze acquisition date, source, offer, device, geography and new/existing status; follow orders through their return window. Reconcile order, payment, discount, shipment, return, refund and service contact with a stable identifier, publishing unreconciled gaps. Contribution = net revenue × product margin − payment − logistics − returns − service − incremental acquisition. Add fixed costs only if the decision changes them. Calculate maximum CAC and minimum basket compatible with target contribution; test new customers, reactivation and repeat purchase separately.
Worked example.
A campaign generates 1,000 incremental orders, a €70 net basket, 45% product margin and €22,000 media spend. Payment, fulfilment, returns and service cost €1,500, €7,500, €4,200 and €0 additional. Product margin: 70,000 × 45% = €31,500. Contribution: 31,500 − 1,500 − 7,500 − 4,200 − 22,000 = −€3,700. ROAS is 70,000 ÷ 22,000 = 3.18, but does not reveal the loss. With unchanged economics, maximum acquisition cost is €18,300, or €18.30 per order.
Checks and limits.
Follow orders through return-window closure; use incremental CAC or label it as an approximation; assign payment, logistics and service costs to the cohort; calculate CAC threshold and minimum basket before growing budget. A first order understates customers who repeat; shared logistics costs can step rather than move per order; advertising attribution does not replace incrementality measurement.
Resources and sources.
Download the e-commerce cohort model. Related: calculate a promotion threshold, compare channel costs, measure incremental orders. Sources: Gupta, Lehmann and Stuart (2004); Fader and Hardie (2009); Gordon et al. (2019).