Return on advertising spend is calculated by dividing attributed advertising revenue by advertising expenditure. The standard formula is ROAS = attributed revenue ÷ advertising spend. If an organisation spends £8,000 on advertising and records £32,000 of revenue attributed to that advertising, ROAS is £32,000 ÷ £8,000 = 4.0. This means the measured attributed revenue equals four times the advertising expenditure.
ROAS can also be expressed as a percentage by multiplying the ratio by 100, producing 400 percent, but the multiple format is common in marketing reporting. The calculation itself is straightforward; the difficult part is determining which revenue and expenditure figures legitimately belong in the numerator and denominator. If those definitions change, the resulting ROAS can change even when the underlying campaign activity has not changed.
The revenue figure should be based on a clearly defined attribution and revenue methodology. Suppose an advertising campaign generates £75,000 in sales and £5,000 of those sales are later refunded. If the business defines attributed revenue as net revenue after refunds, the numerator is £70,000 and the ROAS against £10,000 of advertising spend is 7.0. If the platform instead reports gross transaction value of £75,000, its reported ROAS would be 7.5.
Both calculations can be internally consistent, but they answer different questions. Similar differences arise when taxes, discounts, cancellations or returns are included or excluded. Therefore, an organisation should define the revenue measure before calculating ROAS and should apply that definition consistently across campaigns and reporting periods.
The advertising-spend denominator must also be defined. If a campaign has £12,000 in media spend, the standard media ROAS calculation uses £12,000 as the denominator. If the organisation additionally wants to include £3,000 in agency fees and £2,000 in creative costs, a broader return calculation could use £17,000, but it should not be labelled as the same standard media ROAS without documenting the changed cost scope.
With £60,000 of attributed revenue, the media ROAS would be 5.0, while the broader revenue-to-advertising-related-cost ratio would be approximately 3.53. The difference illustrates why cost definitions matter. ROAS should be calculated using the cost base appropriate to the analytical question, and different cost bases should receive distinct labels rather than being silently mixed.
A step-by-step calculation begins by defining the reporting period, advertising activity, attribution model and revenue event. The analyst then totals the qualifying advertising spend and totals the revenue attributed to the same advertising activity according to the chosen rules. The revenue total is divided by the expenditure total. For example, if three campaigns spend £4,000, £6,000 and £10,000 respectively and receive attributed revenues of £20,000, £24,000 and £30,000, aggregate spend is £20,000 and aggregate attributed revenue is £74,000.
The aggregate ROAS is therefore 3.7. It should not necessarily be calculated by simply averaging the individual campaign ROAS values because campaigns have different spend levels. The ratio of total revenue to total spend is the appropriate aggregate calculation when the objective is overall campaign ROAS.
ROAS can also be converted into a percentage, but the conversion should be clearly labelled. A ROAS of 3.0 means £3 of attributed revenue per £1 of advertising expenditure. Expressed as a percentage of advertising spend, it is 300 percent. This does not mean that profit is 300 percent, nor does it mean that the business received £3 of profit for every £1 spent.
If gross margin is 40 percent, £3 of sales produces £1.20 of gross margin before other costs. The economic break-even point therefore depends on the margin and other relevant variable expenses. For a product with a 50 percent gross margin and no other variable costs considered, a simplified break-even ROAS would be 2.0 because £2 of sales produces £1 of gross margin to cover £1 of advertising cost. Real businesses often require a different threshold because additional costs must be covered.
Attribution methodology can materially change the numerator. Imagine a customer interacts with paid search and social advertising before purchasing a £500 product. If a reporting model assigns £500 to paid search, that revenue contributes £500 to the paid-search numerator. If another model divides the credit between search and social, each channel receives only part of the revenue. The total business revenue remains £500, but channel-level ROAS changes because the attributed revenue is allocated differently.
This is one reason platform ROAS figures should not automatically be compared or summed. Analysts should specify whether conversions are click-through, view-through, multi-touch, last-touch, first-touch or data-driven and should document the attribution window. Otherwise, differences in ROAS may reflect reporting rules rather than genuine changes in advertising economics.
When calculating ROAS for a business with repeat purchases, the analyst must decide whether the numerator contains only immediate revenue or revenue generated later by acquired customers. An initial-order ROAS might include only sales occurring shortly after the advertisement, whereas a lifetime or cohort-based ROAS may include subsequent purchases from the same customers. The latter can be useful for businesses where customer lifetime value is important, but it introduces a longer measurement window and potentially forecast assumptions.
For example, a subscription business might spend £20,000 to acquire customers who initially generate £30,000 in recognised revenue during the first month but are expected to generate £100,000 over a longer period. The immediate ROAS is 1.5, while a forecast lifetime revenue ROAS would be 5.0. These are distinct metrics and should never be presented as though they measure the same time horizon.
For reliable calculation, retain the raw revenue and spend totals as well as the resulting ratio. A report should state the currency, period, advertising channels, attribution method, attribution window, revenue definition, treatment of refunds and the cost categories included. If spend is zero, ROAS is mathematically undefined because division by zero is not possible. If attributed revenue is zero while spend is positive, ROAS is 0.0.
Very high ROAS values can also arise from small spend volumes, so the ratio should be considered alongside absolute revenue and expenditure. The core calculation remains attributed advertising revenue divided by advertising spend, but its usefulness depends on compatible definitions and accurate data. ROAS becomes substantially more informative when combined with contribution-margin analysis and evidence about whether the attributed revenue represents incremental business generated by the advertising.
