ROAS (return on ad spend)

ROAS is one of the most important control metrics in performance marketing, particularly in e-commerce and with Google Ads, Meta Ads or Microsoft Advertising. For decision-makers in mid-sized companies, it is often the metric that decides whether an advertising channel is scaled up or stopped.

How ROAS is calculated

The formula is deliberately kept simple:

• ROAS = revenue generated with advertising ÷ advertising costs

The result can be expressed in two common ways – as a ratio or as a percentage:

• As a ratio: a ROAS of 4 means that each euro invested generated €4 in revenue – often written as "4:1".
• As a percentage: the same value corresponds to 400%.

A worked example: an online shop spends €5,000 on Google Ads in one month and generates €25,000 in revenue through it. The ROAS is 25,000 ÷ 5,000 = 5, i.e. 5:1 or 500%. Each advertising euro brought back five euros in revenue.

ROAS is revenue, not profit

The most common and most expensive fallacy with ROAS: it measures revenue, not profit. A ROAS of 5:1 sounds excellent – but if the trading margin of the product sold is only 20%, the revenue generated just about covers the advertising costs and there is no contribution margin left. Only when the margin is included does it become visible whether a campaign is really profitable.

From this insight comes the concept of break-even ROAS: the ROAS value from which a campaign recoups the pure advertising costs. Simplified, it is the reciprocal of the profit margin. With a margin of 25%, the break-even ROAS is 1 ÷ 0.25 = 4, i.e. 4:1. Everything above it is profitable, everything below it makes a loss – in relation to the advertising costs. Setting a ROAS target without knowing your own margin therefore makes little sense.

ROAS versus ROI

ROAS and ROI (return on investment) are often confused but measure different things. ROAS looks exclusively at advertising costs in relation to revenue – a pure marketing efficiency metric. ROI, on the other hand, includes all costs (cost of goods, staff, logistics, advertising) and sets the profit against the total investment. A high ROAS therefore does not necessarily go hand in hand with a positive ROI.

• ROAS = revenue ÷ advertising costs → measures the efficiency of the advertising budget.
• ROI = (profit − total costs) ÷ total costs → measures the profitability of the total investment.

In short: ROAS is the tactical channel metric, ROI the strategic company metric. Together they give a complete picture.

ROAS versus CPA and CPO

Besides ROAS, many campaigns are steered by the cost per conversion – CPA (cost per acquisition) or CPO (cost per order). The difference from ROAS lies in the perspective: CPA/CPO look at the cost of a single conversion in euros, regardless of the order value. ROAS, on the other hand, takes into account how much revenue a conversion brings in.

For shops with strongly fluctuating basket values, ROAS is usually the more meaningful metric because it weights expensive and cheap orders correctly. For business models with a uniform conversion value – such as a lead campaign or a subscription with a fixed price – CPA is often the more practical control variable. In practice, both metrics are often monitored in parallel.

Target ROAS as a bidding strategy in Google Ads

Google Ads offers an automated Smart Bidding strategy called target ROAS (tROAS). The advertiser sets a desired ROAS value – for example 400% – and the algorithm controls the bids in real time so that this target value is reached on average. The prerequisite is properly set-up conversion tracking with revenue values, so that Google knows at all which click brought in how much revenue.

The practical advantage: instead of maintaining every keyword bid manually, you specify a business goal and leave the fine-tuning to the system. The disadvantage: a target ROAS set too high can choke volume, because the system only buys the safest, highest-revenue clicks and leaves many potential sales on the table. The art lies in setting the target ROAS realistically – ideally slightly above break-even. The official Google Ads Help on target ROAS describes how to set it up.

Limits and typical mistakes with ROAS

As useful as ROAS is, it has blind spots that you should be aware of:

• It ignores the margin – a high ROAS with a low margin can still be loss-making.
• It is a snapshot – customer lifetime value (repeat purchases, follow-up orders) is not taken into account, although it often far exceeds the true value of a new customer's first order.
• Attribution distorts the picture – depending on the attribution model (last click, data-driven), the same revenue is attributed to different channels, which affects the reported ROAS.
• It says nothing about scalability – an excellent ROAS on a small budget often collapses as soon as the budget is increased significantly.

An example from practice: a fashion retailer celebrates a campaign ROAS of 8:1 – until it emerges that a large part of this revenue comes from existing customers who would have bought anyway. The ROAS relating to genuine new customers is considerably lower. That is why it is worth looking at ROAS separately for new and existing customers, if the data allows.

Looking at ROAS across the entire funnel

A single ROAS value for a whole account often conceals more than it shows. Campaigns fulfil different tasks along the customer journey, and it would be a mistake to measure them all against the same ROAS target. Roughly, three levels can be distinguished:

• Top of funnel (awareness) – campaigns that address new, still undecided prospects. Their direct ROAS is naturally low because the purchase often only takes place later via other channels.
• Middle of funnel (consideration) – retargeting and brand-related campaigns that bring prospects closer to buying. Their ROAS is in the middle range.
• Bottom of funnel (conversion) – brand and basket campaigns that pick up users who are ready to buy. They show the highest ROAS but partly buy revenue that would have come anyway.

Anyone who optimises purely for ROAS automatically shifts budget towards the end of the funnel – in the short term the metric looks brilliant, in the medium term the supply of new prospects dries up. A mature approach to ROAS therefore means accepting different target values for different funnel stages and keeping an eye on the overall ROAS across all stages.

ROAS and customer lifetime value

ROAS measures the revenue of a single transaction or period. In business models with repeat purchases – subscriptions, consumables, B2B reorders – it systematically underestimates the true value of a customer won. A new customer whose first order only brings a ROAS of 2:1 can generate five times their initial revenue over two years.

Advanced advertisers therefore supplement classic ROAS with an LTV-based ROAS, which includes not only the revenue from the first purchase but the expected customer value over the entire relationship. This allows more aggressive bidding in new customer acquisition without losing sight of profitability. The challenge lies in the data basis: you need reliable cohort data to estimate the expected lifetime value seriously.

Prerequisites for a reliable ROAS

A ROAS is only as good as the data it is based on. Three technical prerequisites determine how meaningful it is:

1. Clean conversion tracking with revenue values – every conversion must report its actual order value back to the advertising system, otherwise ROAS is calculated with the wrong figures.
2. A suitable attribution model – it determines which channel a sale is attributed to. Data-driven attribution distributes the value more realistically than the simple last-click model.
3. Server-side tracking where there are consent gaps – since many third-party cookies disappeared, pure browser tracking underestimates conversions, which artificially depresses the reported ROAS. Server-side measurement closes part of this gap.

Anyone who neglects these basics is flying blind: a ROAS measured too low leads to profitable campaigns being stopped by mistake – a common, expensive mistake in mid-sized companies.

Interpreting ROAS in practice

Figures alone do not run a company – what matters is the right interpretation. An isolated ROAS value should always be read in the context of three questions: over what period was it measured? What margin lies behind the revenue? And how much new business does it actually contain? Only this classification turns a bare metric into a reliable basis for decisions.

A practical maxim for mid-sized companies: ROAS is an excellent early warning system but a poor sole decision-maker. It quickly shows which campaign is getting out of hand, but never replaces the contribution margin calculation. Anyone who combines both – ROAS for day-to-day control, margin and LTV for strategic assessment – makes the most reliable budget decisions.

Frequently asked questions about ROAS

What is a good ROAS?

That depends on the margin. As a rule of thumb, a ROAS of around 4:1 is considered solid because it is profitable with typical trading margins. With low margins, ROAS must be higher; with high margins, a lower value is enough. What always matters is the comparison with your own break-even ROAS.

How does ROAS differ from ROI?

ROAS only looks at revenue against advertising costs, ROI at profit against all costs. A positive ROAS does not automatically mean a positive ROI.

Why is a ROAS of 5:1 not automatically profitable?

Because ROAS measures revenue, not profit. If the margin is below the reciprocal of the ROAS (below 20% for 5:1), the revenue is not enough to cover the advertising costs profitably.

How is ROAS written?

Two notations are common: as a ratio (e.g. 4:1) or as a percentage (400%). Both mean the same – four euros of revenue per advertising euro invested.

Should I steer campaigns by ROAS or CPA?

With strongly fluctuating order values, ROAS is usually more meaningful; with uniform conversion values, CPA. Many accounts monitor both metrics in parallel.