Return on ad spend, or ROAS, has become the default metric for evaluating advertising performance in digital marketing. It is simple to understand, appears precise, and provides an obvious basis for comparison across campaigns, channels and time periods. A campaign with a 4x ROAS generated four pounds in revenue for every pound spent. A campaign with a 2x ROAS generated two pounds. The logic of using this as the primary measure of success seems straightforward. The problem is that ROAS, used in isolation, gives a distorted and frequently misleading picture of what advertising is actually doing for a business.

The useful detail here is that campaign performance is rarely caused by one setting inside an ad account. The offer, creative, landing page and follow up all shape the result. That is why our paid advertising service connects media buying with content and website decisions, as seen in our work with Pro Project Promotions.

The distortions are not random. ROAS tends to systematically overvalue certain types of activity and undervalue others in ways that, if acted upon, push businesses toward advertising decisions that look sensible on a dashboard but produce worse commercial outcomes over time. Understanding why this happens does not require abandoning ROAS as a metric. It requires treating it as one input into a more complete evaluation rather than the number that everything else is judged against.

The Attribution Problem

Every advertising platform attributes revenue to itself by default using whatever attribution model is most favourable to its own measurement. Meta's default attribution model will claim credit for a sale if someone saw or clicked your ad within seven days of making a purchase, regardless of whether any other marketing touchpoints were involved in the decision. Google's last click attribution gives all credit to the final ad someone clicked before converting, ignoring every previous touchpoint that may have built awareness and intent.

Phoenix Health and Safety provides health and safety training and needed Meta Ads that reached employers and HR teams, not just individuals. Adjusting the targeting by job title and company size brought the cost per lead down considerably and made the enquiries they got far more relevant.

In a world where most customers encounter a business multiple times across multiple channels before making a decision, these single channel attribution models produce a fundamentally inaccurate picture of what drove the outcome. The ROAS figure you see on any individual platform dashboard is that platform's self reported contribution to your revenue. It is not an objective measure of the actual value that campaign generated, because it does not account for the role other channels played in the same conversion.

This matters practically because acting on single platform ROAS figures often means cutting the spending that looks least attributable even when it is doing genuine work. Brand awareness campaigns, which build familiarity and intent over time rather than driving immediate conversions, consistently appear to underperform on ROAS metrics because the revenue they contribute is credited to the last click channel rather than to the awareness channel that created the demand in the first place. Cutting awareness spend on the basis of its ROAS figure while maintaining conversion spend is a common and costly mistake.

What ROAS Does Not Account For

ROAS measures revenue, not profit. Two campaigns with identical ROAS figures can have dramatically different profitability if the product mix or average order value differs. A campaign driving sales of a high margin product at 3x ROAS is more profitable than one driving sales of a low margin product at 5x ROAS, but the ROAS figures alone do not tell you this. Making decisions on the basis of revenue return without understanding margin contribution leads to optimising for the wrong thing.

Customer lifetime value is another dimension that ROAS flattens. Advertising that acquires customers with high retention and repeat purchase rates at a relatively modest first purchase ROAS may be significantly more valuable than advertising that drives large first purchases from customers who never return. The only way to see this is to track customer behaviour beyond the initial sale and factor it into how you evaluate acquisition cost. A business that knows its average customer spends across multiple transactions over several years can afford to bid for first purchase ROAS figures that a business without this knowledge would consider unprofitable.

Incrementality is the most fundamental question that ROAS does not answer. Would these sales have happened anyway, without the advertising? For retargeting campaigns in particular, ROAS figures can look impressive precisely because they are reaching people who were already close to buying. Showing an ad to someone who was going to purchase tomorrow regardless and claiming credit for that purchase inflates the apparent return on the campaign without reflecting any incremental value the advertising actually created. Testing for incrementality, by withholding advertising from a control group and comparing outcomes, is the only way to answer this question with confidence.

Better Ways to Evaluate Advertising Performance

A more complete evaluation framework brings together several metrics rather than relying on ROAS alone. Cost per acquisition, measuring the total advertising cost divided by the number of new customers acquired rather than the revenue generated, is often a more useful operational metric for businesses trying to grow their customer base. Contribution margin per campaign, which accounts for the gross margin of the products or services sold rather than just the revenue, gives a more accurate picture of whether advertising is actually profitable. Blended ROAS, which takes total revenue divided by total advertising spend across all channels rather than attributing by channel, provides a more honest view of overall advertising efficiency.

The most important shift is thinking about advertising performance in terms of business outcomes rather than platform metrics. Are you acquiring more of the customers you want? Are those customers retained and returning? Is the cost of acquiring a customer declining over time as brand awareness builds and conversion rates improve? These questions connect advertising to commercial reality in a way that any single dashboard metric cannot.

ROAS remains a useful signal and it would be a mistake to ignore it. But treating it as the definitive measure of advertising success leads to the kind of short term, attribution distorted optimisation that looks rational on a spreadsheet and erodes the long term brand and audience building that makes businesses durable. The most successful advertisers use ROAS as one metric in a broader conversation about what their advertising is actually doing for the business.

If your campaigns need clearer commercial results, start with our paid advertising service and campaign strategy service. Relevant examples include our work with Pro Project Promotions and Aria's Bistro.