My Store Has a 4× ROAS. Why Am I Still Losing Money?

A 4× return on ad spend means an advertising system attributes four dollars of revenue to each dollar spent on ads. It does not mean four dollars of profit. The order still has to pay for the product, delivery, transaction costs, returns, and the business that makes the sale possible.

Use a per-order model to understand what is left. Then reconcile the model with actual store and accounting results.

Follow one hypothetical order

Assume net merchandise revenue is $60 after discounts, excluding taxes and shipping collected. The following costs are illustrative, not a benchmark:

The $7 still has to contribute toward fixed costs and profit. If the business incurs $1,000 of monthly fixed overhead and produces only 100 comparable orders, $700 of contribution would not cover it.

Do not double count returns. If actual net revenue already deducts refunds, the additional return allowance should cover only costs or expected losses not already included. Similarly, COGS may already include costs that another line in your model would otherwise repeat.

Calculate the advertising break-even point

In this example, pre-ad contribution margin is $22 divided by $60, or about 36.7%. Break-even ROAS before fixed overhead is $60 divided by $22, approximately 2.73×.

At 2.73×, advertising would consume the entire $22 contribution. The order would provide nothing for fixed overhead or desired profit. A practical target therefore needs room above this threshold under these assumptions.

If you require $6 per order for overhead and $4 for profit, the allowable ad cost becomes $12. The corresponding modeled target is $60 divided by $12, or 5×. This is an example of working backward from the economics, not a universal ROAS target.

Check who is buying

Ad reports may include existing customers. A low purchase CPA can look appealing while new-customer acquisition remains expensive. Separate new and returning customers where the data supports it, and distinguish ad spend per attributed order from fully loaded CAC.

Future repeat orders can improve customer economics, but they are not guaranteed. Use observed cohort contribution and payback rather than assuming a customer will buy five more times because the spreadsheet needs them to.

Check what is being credited

Different channels may each claim the same purchase. A customer could click an ad, open an email, and later return directly. Adding the platforms' attributed revenue can exceed actual store revenue.

Use order records as the transaction reference and accounting records for the broader financial picture. Compare attribution models separately instead of combining their claims into a larger fictional sales total.

Improve the constraint you find

If product margin is weak, investigate pricing, sourcing, or the product mix. If returns are high, examine fit and expectations. If fulfillment is expensive, inspect packaging and shipping patterns. If acquisition is the problem, improve traffic relevance, creative, and conversion while monitoring contribution.

Your next action is to calculate the economics of your three most important products. A storewide ROAS average can hide a profitable product funding another that loses money on every acquired order.