Ecommerce ads: how can you tell whether a campaign pays off?

An ad campaign can show an attractive return on ad spend while losing money. Attributed revenue does not remove product cost, shipping, payment fees, discounts or returns. Calculate contribution for each order first, then compare it with acquisition cost.

Use a business-level formula

For an order, start with cash received net of applicable tax. Subtract product purchase, packaging, fulfillment, payment processing, absorbed shipping and an evidence-based allowance for returns. Then subtract associated ad spend using a clearly stated attribution method. Keep fixed costs separate and check whether remaining contribution can cover them.

Illustrative example: one campaign produces a €100 order with €55 variable costs and €30 attributed advertising cost. That leaves €15 before fixed costs. Another produces an €80 order with €35 variable costs and €20 advertising cost, leaving €25. Higher revenue is not necessarily better economics.

Be careful with attribution

An ad platform may claim a purchase the customer would have made after a brand search or email. Look at paths, new versus existing buyers and total order trends. Where possible, compare similar periods or areas with and without the campaign while recording other changes. Do not claim precise incrementality without a suitable control.

Decide where the next euro goes

Set an acceptable contribution threshold before raising spend. If a campaign brings low-margin baskets, revise offer or landing page before increasing budget. If sales rise but support and returns surge, include those costs. The purpose of measurement is to decide whether to continue, change or stop.

Choose a formula the team understands

Start with order contribution: collected revenue minus product cost, processing, packaging, handling, shipping absorbed and expected average returns. Only then compare contribution with ad cost per order. ROAS, attributed revenue divided by ad spend, excludes product and operating costs. A high ROAS can therefore accompany a losing campaign.

In an illustrative case, a store spends €500 and the ad dashboard attributes €2,000 sales: ROAS 4. Products and variable costs take €1,400, leaving €600 contribution before advertising. After €500 ad spend, €100 remains before fixed costs. If €300 of attributed sales came from customers who would have bought without ads, incremental results may be weaker. Do not turn platform attribution into proof of causality.

Reconcile data with orders

Compare purchase counts from the ad platform, analytics and Shopify for the same period and note differences. Check currencies, time zones, cancellations, refunds and taxes. One order can appear in several systems with different attribution rules. Keep a table of paid orders and actually billed advertising. For repeat purchases, separate new from existing customers so the same relationship is not valued twice.

With enough volume, compare a region, group or period exposed to advertising with a similar one without it, keeping other changes as steady as possible. Such tests have limits but can help estimate sales added by ads. At low volume, use contribution, customer feedback and an explicit uncertainty range instead.

Decide where the next euro goes

Increase spend only when extra orders keep an acceptable contribution and the store can fulfill them. If acquisition cost rises after a budget increase, return to the earlier level and examine audience, offer and frequency. If returns rise, fix the ad promise. If carts do not become orders, test checkout and delivery fees. The right decision might be to stop the campaign, improve the page or change product; a dashboard cannot decide alone.

Questions when deciding whether ads pay

Is ROAS 4 good? It cannot be judged without margin. If 75% of revenue goes to product, processing and delivery, the threshold differs greatly from a digital guide. Calculate contribution and an acceptable acquisition cost for your range.

Should you trust platform-attributed sales? They offer a signal, but their attribution windows differ from Shopify order records. Reconcile figures, remove cancellations and refunds, and separate new from existing customers.

When should budget rise? After several orders with acceptable contribution, a reliable buying path and enough fulfillment capacity. Increase in steps to see how acquisition cost changes. A sudden jump may reach less interested audiences.

Why might a paper-profitable campaign strain cash? Ads and stock may be paid before order payouts arrive, while returns and fees show up later. Add payout timing and an incident reserve to the model.

A campaign contribution worksheet

For each paid order attributed to a campaign, record collected revenue, product cost, handling, payment fee, shipping absorbed, expected return allowance and ad spend for the period. Sum contribution before ads, then subtract the actual platform bill. Mark refunded or canceled orders separately. This yields a more useful first view than the platform's revenue divided by spend.

Next challenge attribution. Which buyers were already customers? Did an email or organic search also help? Would some orders likely have happened without the ad? You may not be able to answer exactly, so calculate a conservative range rather than inventing one precise incremental profit. Review cash timing too: supplier payments and advertising charges may precede customer payouts. If the low end of the range is negative, pause scaling and test a clearer offer or audience before increasing budget.

Related guides

• Online store pricing: calculate costs before matching rivals
• Ecommerce conversion: diagnose the journey before testing

Further reading: official documentation.