ROI vs ROAS: You’re Optimising the Wrong Number

ROI and ROAS are not interchangeable. ROAS measures revenue generated per pound of ad spend. ROI measures the actual profit returned on total investment, including costs, margins, and overheads. Confusing the two is not a minor accounting error. It is how businesses convince themselves a campaign is working when it is quietly losing money.

Most marketing teams report ROAS because it is the number that ad platforms hand you by default. It is clean, it is high, and it makes the deck look good. But optimising for ROAS without understanding ROI is like celebrating a full order book while ignoring that every order ships at a loss.

What Is ROAS and Why Do Platforms Love It?

ROAS is calculated simply: revenue divided by ad spend. If you spend £10,000 on Google Ads and the platform attributes £40,000 in revenue, your ROAS is 4x. That number is easy to produce, easy to benchmark, and easy to present in a weekly report. Which is exactly why it gets so much airtime.

Ad platforms report ROAS because it reflects their contribution to your revenue, not your profitability. Google and Meta are not running your P&L. They are selling you inventory and showing you the best possible version of what that inventory produced. Attribution windows, view-through conversions, and modelled data all tilt the number in a favourable direction. I am not saying the platforms are being dishonest. I am saying their incentives are not the same as yours.

Early in my career, I was as guilty of this as anyone. I spent years presenting ROAS figures to clients as evidence of campaign success. The number looked strong, the client was satisfied, and we moved on. It took a few uncomfortable conversations with CFOs, people who wanted to see the margin contribution not the revenue multiple, to recalibrate how I thought about performance reporting. A 6x ROAS on a product with 15% net margin is not a success story. It is a slow leak dressed up as growth.

What Is ROI and Why Is It Harder to Calculate?

ROI is profit divided by total investment, expressed as a percentage. To calculate it properly, you need to know your gross margin on the products being sold, your cost of goods, fulfilment and returns costs, the fully loaded cost of the campaign including agency fees and creative production, and any overhead allocation relevant to the channel. That is a more complex calculation than dividing revenue by ad spend, which is precisely why teams avoid it.

When I was running agency teams managing significant ad budgets across retail, FMCG, and financial services, the ROI conversation was always harder to have than the ROAS conversation. Not because clients did not want to know the truth, but because the data required to calculate real ROI sat in systems the marketing team often did not have access to. Finance owned the margin data. Operations owned the fulfilment costs. Marketing owned the spend. The three rarely spoke to each other with enough frequency to produce a clean, shared view of profitability.

That structural gap is still one of the most common problems I see. The marketing team is optimising a metric that finance does not recognise as meaningful, and finance is frustrated that marketing cannot demonstrate business impact in terms they understand. Both sides are operating in good faith. They are just working from different numbers.

If you are thinking seriously about how performance metrics connect to go-to-market decisions, the Go-To-Market & Growth Strategy hub covers the broader commercial framework that makes metrics like ROI meaningful rather than abstract.

The Break-Even ROAS Problem Nobody Talks About

There is a number that sits between ROAS and ROI that almost nobody in marketing talks about: break-even ROAS. This is the minimum ROAS your campaign needs to achieve before it is generating any profit at all. It is calculated from your gross margin, and it varies dramatically by product and category.

The formula is straightforward. Break-even ROAS equals 1 divided by your gross margin percentage. If your gross margin is 50%, your break-even ROAS is 2x. If your gross margin is 25%, your break-even ROAS is 4x. If your gross margin is 15%, your break-even ROAS is 6.67x. That means a campaign reporting a 5x ROAS on a 15% margin product is operating below break-even. It is losing money on every attributed sale, and the ROAS figure is giving no indication of that whatsoever.

I have sat in rooms where a 4x ROAS was being celebrated as strong performance. For some businesses in that room, it genuinely was. For others, it was below break-even. The number meant completely different things depending on the margin structure of the product being sold, and nobody had made that calculation explicit. The benchmark had become the standard without anyone checking whether the benchmark was relevant to the business model.

How Attribution Makes ROAS Even Less Reliable

Even if you have accepted that ROAS is a partial picture, there is a second problem: the ROAS figure itself is often wrong, or at least, it is a platform’s best estimate of the truth rather than the truth itself.

Attribution in digital advertising is genuinely difficult. Customers interact with multiple touchpoints before converting. Last-click attribution overvalues the final interaction. Data-driven attribution models are more sophisticated but still operate on probabilistic assumptions. View-through conversions count impressions that may have had no causal relationship to the sale. And with the continued erosion of third-party cookies and signal loss across platforms, the data feeding these models is getting noisier, not cleaner.

I spent a long time in the performance marketing world believing that the numbers coming out of the platforms were an accurate reflection of what those platforms were producing. I do not believe that anymore. I think they are a useful signal, not a precise measurement. The distinction matters because if you treat a signal as a fact, you make decisions with false confidence. If you treat it as an approximation, you stay appropriately humble about what you actually know.

This is not a reason to abandon measurement. It is a reason to triangulate. Look at ROAS alongside revenue trends, new customer acquisition rates, and blended margin performance. No single number tells the whole story, and the businesses that perform best over time are the ones that understand that analytics tools are a perspective on reality, not reality itself.

For a broader look at how growth measurement connects to commercial strategy, the BCG piece on commercial transformation is worth reading. It frames measurement in the context of business growth rather than channel performance, which is the right lens.

The Lower-Funnel Trap and What It Costs You

There is a broader strategic problem sitting behind the ROAS obsession, and it took me years to see it clearly. When you optimise hard for ROAS, you almost always end up concentrating spend on the bottom of the funnel: retargeting, branded search, shopping campaigns, high-intent audiences. These are the people already close to buying. The ROAS looks excellent because you are essentially capturing demand that was going to convert anyway.

Think about it like a clothes shop. Someone who walks in, picks something up, and tries it on is far more likely to buy than someone browsing the window. If you only ever talk to the people already inside the fitting room, your conversion rate looks extraordinary. But you are not growing your customer base. You are just efficiently harvesting the intent that already existed.

Earlier in my career, I overvalued lower-funnel performance for exactly this reason. The numbers were compelling. Clients were happy. But when I started asking harder questions, like where is the new customer growth, how much of this revenue would have happened without the paid campaign, and are we reaching audiences who did not already know us, the picture shifted. A significant portion of what performance marketing was being credited for was going to happen regardless. We were measuring capture, not creation.

This is not an argument against performance marketing. It is an argument for being honest about what it does and does not do. ROAS is a reasonable measure of demand capture efficiency. It is a poor measure of demand generation. If your growth strategy depends on reaching new audiences, not just converting existing intent, then ROAS alone will consistently mislead you about whether your marketing is working.

The Semrush breakdown of growth strategies illustrates how the businesses that scale tend to invest across the full funnel rather than concentrating everything at the point of conversion. The efficiency gains at the bottom are real but finite. The opportunity at the top is where sustainable growth lives.

When ROAS Is the Right Metric to Use

None of this means ROAS is useless. It is a legitimate channel-level efficiency metric, and there are contexts where it is exactly the right number to be watching.

ROAS is useful for comparing the relative efficiency of campaigns within the same channel, when margin is held constant. If you are running five Google Shopping campaigns across the same product category and you want to know which creative or audience configuration is performing best, ROAS is a clean signal. It is also useful for setting campaign-level targets once you have done the break-even calculation and know what ROAS you need to achieve profitability at your margin.

Where ROAS breaks down is when it is used as a proxy for business health, when it is compared across channels with different attribution models, or when it is the primary number being reported to senior leadership without any margin context. At that point, it stops being a useful signal and starts being a source of strategic confusion.

The discipline is knowing which question you are trying to answer. If the question is “which of these two ad sets is more efficient at generating revenue from this audience,” ROAS is probably the right tool. If the question is “is our paid media investment generating profitable growth for the business,” you need ROI, and you need the margin and cost data to calculate it properly.

Building a Reporting Framework That Uses Both

The practical answer is not to choose between ROI and ROAS. It is to use them at the right level of the business.

At the campaign and ad set level, ROAS is a reasonable operational metric. It is fast, it is available, and it helps you make tactical decisions about budget allocation within a channel. Set a target ROAS based on your break-even calculation, add a margin buffer for profitability, and use it to guide day-to-day optimisation.

At the channel and portfolio level, move to contribution margin. How much gross profit is each channel generating after its direct costs? This is the number that bridges marketing performance and finance, and it is the conversation that senior leadership actually wants to have.

At the business level, ROI is the verdict. Total investment in marketing, including agency fees, technology, creative, and media, measured against the incremental profit generated. This is a harder number to produce, and it requires collaboration between marketing, finance, and operations. But it is the only number that tells you whether marketing is genuinely contributing to business growth or just moving money around efficiently within the existing customer base.

When I was scaling agency teams and managing P&Ls, the shift that made the biggest difference to client relationships was moving reporting conversations from ROAS to contribution margin. It took longer to set up. It required pulling data from systems that did not naturally talk to each other. But it changed the nature of the conversation from “how did the campaign perform” to “what did the campaign contribute to the business,” which is a fundamentally more useful question.

Tools like Hotjar and platforms that track behavioural data can add useful context to this picture, particularly for understanding where in the funnel customers are dropping out and what the quality of traffic from different channels actually looks like beyond the conversion event. Revenue attribution is the start of the analysis, not the end of it.

The Crazy Egg overview of growth strategy makes a similar point about the danger of optimising for surface-level metrics when the underlying business mechanics tell a different story. The channel number and the business number are not the same thing, and treating them as interchangeable is where growth strategies go wrong.

What This Means for How You Brief Agencies and Platforms

If you brief an agency or a platform to maximise ROAS, that is exactly what they will do. They will concentrate spend on high-intent audiences, suppress upper-funnel activity that depresses short-term returns, and deliver a number that looks strong in the monthly report. You will have asked for ROAS and received ROAS. Whether you received profitable growth is a different question.

The brief needs to reflect the business objective. If profitable customer acquisition is the goal, the brief should specify a target cost per acquired customer at a given margin, not a ROAS target. If incremental revenue growth is the goal, the brief should include a requirement to demonstrate incrementality, not just attribution. If brand-building alongside performance is the objective, the brief needs to ring-fence budget for upper-funnel activity and accept that the blended ROAS will be lower as a result.

I have judged the Effie Awards, which measure marketing effectiveness in terms of business outcomes rather than channel metrics. The campaigns that win are almost never the ones with the highest ROAS. They are the ones where the marketing strategy was clearly connected to a commercial objective, and the results were measured against that objective rather than a platform-generated efficiency score. That gap between what gets celebrated in effectiveness circles and what gets reported in weekly performance reviews is worth thinking about.

The Vidyard research on pipeline and revenue potential highlights a related issue in B2B contexts: the gap between pipeline metrics and actual revenue contribution. The same logic applies to ROAS. Measuring the activity is not the same as measuring the outcome.

If you want to build marketing that connects to commercial outcomes rather than channel metrics, the Go-To-Market & Growth Strategy hub covers the frameworks and thinking that make that connection possible. It is where measurement, strategy, and commercial reality meet.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions