ROAS and profit: when do ads actually make money?

How is ROAS calculated?

ROAS = revenue attributed to ads / ad spending. If a report shows RSD 120,000 in revenue and RSD 30,000 in spending, ROAS is 4. That means four dinars in reported revenue for each dinar of media spending. It does not mean four dinars of profit.

Before calculating, check what the reported revenue includes: tax, shipping, cancelled orders or discounts can change the interpretation. Especially with cash on delivery, a placed order is not the same as a paid sale that has not been returned.

Two stores, the same ROAS, different outcomes

The following calculation is illustrative and does not represent a particular client's results. Both stores have RSD 120,000 in revenue excluding VAT, and the full amount is attributed to ads. Both spend RSD 30,000 on advertising, so both have a ROAS of 4.

After all variable costs except advertising, store A retains 40% of revenue: RSD 48,000. Advertising leaves RSD 18,000. Fixed costs and other marketing expenses still have to be paid from that amount.

Store B retains 20%: RSD 24,000. After the same media spending, its result is minus RSD 6,000, before fixed costs. The same advertising metric can therefore describe very different business situations.

How do you calculate break-even?

If the contribution margin before advertising is m, the simplified break-even ROAS is 1 / m. At 40%, this is 2.5; at 20%, it is 5. The formula requires a consistent basis for revenue and costs. It covers variable costs and media spending, excluding fixed costs and the desired profit.

If you want another 10% of revenue left after advertising, a 40% contribution margin leaves a maximum of 30% for ads. The planned ROAS requirement is then approximately 3.33. This is a business target derived from assumptions, not a promise that this performance can be achieved.

Where repeat purchases exist, you can evaluate the cost of the first customer over a longer period. However, future customer value should be based on measured repeat purchases, timing and costs. Assuming every customer will return is not enough to justify a larger budget.

Why do Meta, GA4 and the store disagree?

Different systems may attribute the same purchase to different channels and reporting periods. Consent choices, tracking blockers, time zones and event transmission also create differences. A discrepancy is not automatically proof that one system is entirely broken.

Use fulfilled and paid orders as the business baseline. Then compare measurement with that baseline: does one order trigger multiple purchase events, is the currency correct and can test orders be identified? Do not add together revenue that multiple advertising platforms each claim as though it represented separate sales.

Alongside ROAS, track total revenue against total media spending, contribution after ads, new-customer acquisition cost, average order value and return rate. The overall revenue-to-spend ratio provides a broader picture, but does not by itself prove that advertising caused all sales.

What should you change first if revenue grows but profit does not?

Review which products the advertising sells. A campaign may favour an easy-to-sell product with a weak margin or expensive delivery. Product-level or category-level analysis is then more useful than an account-wide average.

Check discounts, returns and the share of existing customers. Then select one specific change: a different offer, a sensible product bundle, clearer delivery charges or a different budget allocation. Changing everything at once makes it harder to understand why performance improved.

Checklist

• The same revenue basis used for margins and ROAS.
• Fulfilled, cancelled and returned orders separated.
• All variable costs included.
• Media costs distinguished from total marketing costs.

What is a good ROAS?

One that leaves enough contribution for your business objective given your margins, costs and order fulfilment. A universal threshold such as ROAS 3 or 5 is unreliable without that calculation.