How to calculate return on ad spend (with real examples)

Picture this: a business owner checks their Google Ads dashboard and sees a 4x ROAS. That looks strong. They assume the campaign is profitable, leave it running, and increase the budget. Three months later, margins have eroded and they cannot work out why. What the dashboard did not show was the 12% refund rate, the cost of goods they forgot to subtract, or the fact that Google and Meta were both claiming credit for the same 200 conversions simultaneously. The headline ratio was technically accurate. The campaign was not profitable.

The ROAS formula itself is genuinely simple: revenue divided by ad spend. Where businesses go wrong is in what they feed into that formula and how they read the result. Juicy Designs reports a 4.8x average return on ad spend across 38 active ad accounts, about 67% higher than the widely cited global industry average of 2.87x. That gap comes from measuring ROAS correctly from the start: with proper attribution, margin adjustments, and platform-level scepticism built in from day one.

What ROAS actually means (and how the formula works)

Return on ad spend measures how much revenue you generate for every rand spent on advertising. The formula is: ROAS = Revenue attributed to ads ÷ Ad spend. If you spent R10,000 on a campaign and it generated R40,000 in attributed revenue, your ROAS is 4x.

“Revenue attributed to ads” is the critical phrase. It means only the revenue that can be traced back to a specific campaign, not your total monthly turnover or your organic sales. If your Google Shopping campaign generated R90,000 in confirmed orders during the attribution window, that is the figure you use. The R200,000 your business made in total that month is irrelevant to the ROAS calculation for that specific campaign.

For a deeper explanation of what this metric measures and how platforms define it, see our guide to ROAS meaning.

Ratio vs percentage: which format should you use?

Most ad platforms display ROAS as a ratio or multiplier: 4x, 4:1, or the decimal 4.0. Financial reports and some agency dashboards express the same number as a percentage (400%). Both describe identical performance. The problem only arises when a team mixes formats in the same report without labelling clearly, because a 400% ROAS and a 4x ROAS look very different at a glance. Pick one format, document it, and apply it consistently across all campaign reviews.

How to calculate return on ad spend: three worked examples

The formula stays the same regardless of business model. What changes is which revenue figure you use on the top line.

Ecommerce calculation

An online retailer spends R20,000 on Google Shopping ads in a month. During that same attribution window, the campaign drives R90,000 in confirmed orders, verified against the order management system.

If your store processed R300,000 in sales but only R90,000 was trackable to the specific campaign, R90,000 is the correct input. Using total store revenue inflates ROAS and makes campaigns appear more efficient than they are.

Lead generation calculation

Lead generation businesses do not record a direct sale at the moment of conversion, so you assign a monetary value to each lead based on your close rate and average deal value.

A 1.5x ROAS in lead generation is not necessarily a poor result. Revenue from those leads often arrives weeks or months after the initial conversion, and the closed deal value is typically far higher than what the ROAS formula captures at campaign level. Lead-gen ROAS runs lower by design. The metric becomes meaningful when you track it consistently over time and compare it against your break-even threshold, not against an ecommerce benchmark.

Subscription business calculation

Subscription businesses face a specific ROAS distortion: if you only count first-month revenue, the number looks weak.

Subscription models must layer in customer lifetime value to make ROAS a meaningful decision-making tool. Cutting a campaign with a 1.58x short-term ROAS that produces a 12.64x lifetime ROAS would be a costly mistake.

ROAS vs ROI: why these two metrics tell different stories

ROAS and ROI are not interchangeable. ROAS measures the revenue return on your ad spend. ROI measures the net profit return on your total investment, including cost of goods, salaries, tools, agency fees, and fulfilment costs. A campaign can show a strong ROAS and still be running at a loss once all costs are factored in.

Use ROAS to optimise individual campaigns and channels quickly. It is the right tool for day-to-day bid adjustments and budget allocation between ad sets. Use ROI for strategic planning and profitability assessments at the business level. Platforms report ROAS because they have access to revenue data and ad spend in real time. They do not have visibility into your COGS, your salary costs, or your fulfilment expenses.

How attribution models change the ROAS number you see

Attribution is the process of deciding which ad gets credit for a conversion. The model your platform uses by default can dramatically change the ROAS reported for any given campaign, without any change in actual performance.

Different attribution models distribute revenue in very different ways. Last-click gives 100% of the revenue credit to the final ad a customer clicked before purchasing. First-click gives all the credit to the first ad that introduced them to the brand. Multi-touch models are more nuanced:

• Linear: equal credit to all touchpoints in the journey
• Time-decay: more credit to touchpoints closer to the conversion
• Position-based: extra weight on first and last touch, with the middle split between them

The same R50,000 in revenue can be distributed very differently across campaigns depending on which model applies, and that changes every ROAS figure in the report.

Why your platform ROAS and your actual ROAS rarely agree

Google Ads defaults to data-driven attribution, while Meta Ads uses a 7-day click and 1-day view window by default. These windows overlap heavily. A customer who clicked a Meta ad on Monday and a Google Search ad on Wednesday, then purchased on Friday, will be counted as a conversion by both platforms simultaneously. If that order was worth R5,000, both Google and Meta report R5,000 in attributed revenue from the same transaction.

The fix is to reconcile platform-reported conversions against actual backend orders regularly, using a neutral analytics source rather than trusting both dashboards simultaneously.

Adjusting ROAS for profit: the calculation most businesses skip

Standard ROAS ignores the costs that sit between revenue and profit. A more complete picture comes from profit-adjusted ROAS, which subtracts all variable order costs before dividing by ad spend.

Formula: (Attributed Revenue − COGS − Shipping − Fees − Refunds) ÷ Ad Spend

A campaign that appeared highly efficient on the dashboard is generating thin returns once real costs are applied. Decisions made on the 5x headline would lead to scaling a campaign that is barely breaking even.

Calculating your break-even ROAS

Break-even ROAS is the minimum return required to cover all non-ad costs without losing money. The formula: 1 ÷ Net Contribution Margin. If a product has a 30% contribution margin after COGS and fulfilment, the break-even ROAS is 1 ÷ 0.30 = 3.33x. Any campaign running below that figure is losing money, regardless of what the dashboard shows.

Every South African business owner running paid ads should calculate this number before setting a target ROAS in their campaigns. A 4x ROAS target is meaningless without knowing whether 4x covers your costs. A business with a 60% margin can scale profitably at 2.5x. A business with a 20% margin needs 5x just to break even. Calculate your break-even first, then set your target ROAS above that floor with a meaningful profit buffer built in.

What a “good” ROAS looks like by industry

Global benchmarks provide a useful reference point, but they should never be used as your primary target. Current benchmarks across the key sectors most relevant to South African advertisers are shown below. Figures drawn from published industry reports including WordStream and Tinuiti; treat these as directional, not prescriptive.

Those figures come from North American and European datasets, and the South African context diverges in ways that matter. Local CPMs tend to be lower than US equivalents, but conversion rates and average order values also differ. A Gauteng-based fashion retailer faces meaningfully different cost-per-click and purchasing-power dynamics than the US or UK brands these benchmarks were built on. Use the table as a rough orientation point rather than a target.

A blanket rule that “a good ROAS is 4x” does not account for whether your margins are 15% or 60%. Margin-based targets outperform benchmark-based targets every time, and that is especially true in South Africa where local cost structures differ from the markets where most benchmarks are published.

Common mistakes that make your ROAS misleading

The following three patterns consistently distort ROAS reporting, and all are avoidable with the right approach to measurement.

Trusting platform-reported ROAS without questioning the attribution. Ad platforms are incentivised to show your results in the best possible light. Overlapping attribution windows, view-through conversions counted at full value, and broad match keywords triggering on loosely related queries all inflate reported ROAS. Compare the number of conversions your platform reports against confirmed orders in your CRM. If the platform claims 150 conversions and your backend shows 90 orders, the reported ROAS is overstated.

Ignoring the cost structure behind the revenue number. A 6x ROAS on a product with a 12% margin is a loss. A 2x ROAS on a product with an 80% margin can be very profitable. The ROAS formula tells you the revenue multiple, not whether the campaign is making money. Without layering in margin, a strong headline ratio will lead a business to scale a losing campaign.

Comparing ROAS across channels as if they measure the same thing. Google Search ROAS and Meta awareness campaign ROAS are fundamentally different metrics. Search captures purchase intent at the bottom of the funnel. Social builds brand awareness at the top. Expecting both to produce identical ROAS is like comparing a car’s top speed to its fuel efficiency: they serve different purposes. Each channel needs its own benchmark tied to its specific role in the funnel.

What plain-language ROAS reporting should look like

A useful ROAS report contains more than a headline ratio. It should include attributed revenue by channel with the attribution window clearly stated, gross ROAS alongside profit-adjusted ROAS, your break-even ROAS as a reference point, and a reconciliation of platform-reported conversions against verified backend orders. Without that context, the ratio alone does not tell you whether your ad spend is working.

If your current agency sends a monthly report with a ROAS figure and no further context, ask for all of the above. The businesses that make the best budget decisions are the ones that can read their ROAS clearly, in plain language, without needing an account manager to translate the dashboard into something actionable.

Juicy Designs reports on the numbers that matter: profit-adjusted ROAS, break-even thresholds, verified conversions, and blended channel performance. If you are unsure whether your current campaigns are being measured and reported in a way that reflects real profitability, that is worth a conversation. Get in touch with Juicy Designs for a plain-language ROAS review of your active campaigns.

Frequently asked questions

How do you calculate return on ad spend?

What is a good ROAS in South Africa?

What is break-even ROAS?

What is the difference between ROAS and ROI?

Why do Google and Meta report different ROAS numbers for the same spend?