Advertising ROI: Why You’re Measuring the Wrong Thing

Advertising return on investment measures the revenue or business value generated relative to the cost of running advertising. The formula is straightforward: revenue attributed to advertising, minus the cost of that advertising, divided by the cost, expressed as a percentage. What is not straightforward is whether the revenue you are attributing to advertising would have happened without it.

That gap between attributed revenue and caused revenue is where most advertising ROI calculations quietly fall apart. And it is the gap that most marketers, most CFOs, and most attribution platforms have little interest in discussing honestly.

Why the Standard ROI Formula Misleads More Than It Guides

Earlier in my career, I was a committed believer in lower-funnel performance marketing. The numbers were clean, the attribution was tight, and the ROI figures looked excellent in client presentations. It took me longer than I would like to admit to recognise that much of what those channels were being credited for was going to happen regardless. Someone had already decided to buy. The paid search ad just happened to be the last thing they clicked.

This is not a niche academic concern. It is a structural problem baked into how most businesses measure advertising effectiveness. When you optimise toward attributed conversions, you naturally shift budget toward channels that are good at being present when a decision has already been made. You end up spending more and more money on the bottom of the funnel, capturing intent that existed before your advertising ran, and calling it return on investment.

The channel with the highest reported ROI is often the one doing the least incremental work. It is intercepting demand that upper-funnel activity, brand reputation, or organic word of mouth already created. Strip away that pre-existing demand and the ROI number collapses.

Attribution Is a Perspective, Not a Fact

I have managed hundreds of millions in ad spend across more than thirty industries. In that time, I have sat in more attribution debates than I can count, and the one thing they have in common is that everyone in the room treats their attribution model as though it describes reality. It does not. It describes a methodology for assigning credit, and every methodology has blind spots.

Last-click attribution rewards the channel that closed the deal. First-click attribution rewards the channel that started the conversation. Data-driven attribution uses machine learning to distribute credit across touchpoints, which sounds more sophisticated but still operates within the walled garden of whatever data the platform can see. None of these models can account for the billboard someone drove past, the podcast they half-listened to, or the colleague who mentioned your product over lunch.

This matters enormously when you are trying to calculate advertising ROI, because the denominator (what you spent) is precise and the numerator (revenue caused by advertising) is an estimate dressed up as a fact. The precision of the formula creates a false sense of certainty about a number that is, at best, an honest approximation.

If you are building your go-to-market strategy around advertising ROI figures, it is worth understanding what those figures are actually measuring. The broader context for how advertising fits into commercial growth is something I cover in depth across the Go-To-Market and Growth Strategy hub, where the relationship between spend, reach, and revenue gets the nuance it deserves.

The Incrementality Problem Nobody Wants to Talk About

Incrementality asks a simple question: would this conversion have happened without the advertising? It is the most important question in advertising ROI, and it is the one most businesses avoid because the honest answer is uncomfortable.

Think about a clothes shop. A customer who tries something on is far more likely to buy than one who is browsing from the door. But the act of trying on the garment is not what created the desire to buy. The desire was already there. The fitting room just confirmed it. A lot of lower-funnel advertising works exactly the same way. It is the fitting room, not the reason someone walked into the shop.

When I was building out performance marketing teams at iProspect, one of the disciplines I tried to instil was separating what we could prove we caused from what we were merely present for. It is a harder conversation to have with clients when you are showing them strong ROAS figures, but it is the only honest way to understand whether the advertising is actually working or whether the business would be growing at the same rate without it.

Incrementality testing, holdout experiments, and geo-based lift studies are the tools that get you closer to a real answer. They are not perfect, but they are considerably more honest than attribution models that assume every click was decisive. The growing complexity of go-to-market environments makes this kind of rigorous testing harder, not easier, which is precisely why most teams skip it.

Short-Term ROI Versus Long-Term Value: The Trade-off That Sinks Brands

There is a version of advertising ROI optimisation that looks brilliant on a quarterly dashboard and quietly destroys a brand over three years. It works like this: you cut brand-building activity because it is hard to attribute, you shift budget into performance channels because the numbers are clean, and you watch your short-term ROAS improve while your brand slowly loses the mental availability that was generating the demand in the first place.

I have seen this happen more than once. A business inherits a strong brand with genuine consumer affinity, runs it on performance-only budgets for eighteen months, and then wonders why its cost per acquisition is creeping up quarter after quarter. The brand was doing work that the attribution model could not see. When the brand investment stopped, the invisible work stopped too. The performance numbers just took a while to reflect it.

The honest version of advertising ROI has to account for time horizon. A brand awareness campaign that runs for six months may show a poor ROI at month six and a strong ROI at month eighteen, once the mental availability it built has had time to compound into purchase decisions. Measuring it at month six and calling it inefficient is not rigorous analysis. It is impatience dressed up as accountability.

This is one reason why the relationship between brand strategy and go-to-market execution matters so much. Brand and performance are not competing budget lines. They are different parts of the same commercial system, operating on different time horizons.

What a Useful Advertising ROI Framework Actually Looks Like

A useful framework for advertising ROI does not produce a single clean number. It produces a range of perspectives that, taken together, give you a defensible view of whether your advertising is working. Here is how I think about it across three dimensions.

Reported ROI

This is the number your attribution platform gives you. It is useful as a directional signal and for comparing performance across channels within the same attribution model. It is not a reliable measure of true business impact. Treat it as a starting point, not a conclusion.

Incremental ROI

This is what you get when you run holdout tests or geo-based experiments to isolate the causal effect of advertising. It is harder to produce, takes longer to set up, and will almost always show a lower number than reported ROI. It is also considerably more honest. If you can run even one or two incrementality tests per year on your highest-spend channels, you will learn more about your advertising effectiveness than years of attribution data can tell you.

Long-Term Revenue ROI

This looks at the relationship between advertising investment and revenue trends over twelve to thirty-six month periods, controlling for external factors where possible. It is the least precise of the three but the most commercially relevant for strategic decisions. Businesses that grow consistently over time tend to maintain advertising investment through downturns rather than cutting it when short-term ROI dips.

Running all three in parallel gives you something far more useful than any single metric: an honest picture of what your advertising is doing, what it is claiming credit for, and what it is building for the future.

The Growth Problem That ROI Optimisation Cannot Solve

Here is the structural limitation of optimising hard for advertising ROI: it tends to push you toward audiences who already know you. People who are already in market, already familiar with your brand, already close to a purchase decision. These audiences convert well. They produce strong ROI numbers. And they are finite.

Growth requires reaching people who are not yet aware of you, not yet in market, not yet thinking about the problem your product solves. This kind of advertising is expensive relative to its short-term return. It is also the only kind of advertising that expands your addressable market rather than just harvesting from it.

When I was growing teams at iProspect from around twenty people to over a hundred, one of the things I learned was that the clients who grew fastest were not the ones who optimised hardest for efficiency. They were the ones who maintained investment in reach and awareness even when the attribution numbers made it look inefficient. They were building the pipeline that performance marketing would later harvest, and they understood that the two activities were connected even when the measurement said otherwise.

Growth tactics that work at scale almost always involve some form of audience expansion, not just conversion optimisation. The businesses that plateau are frequently the ones that cut awareness investment in favour of performance efficiency and then wonder why their performance channels are getting more expensive over time.

How to Present Advertising ROI to a CFO Without Lying

I have presented advertising performance to CFOs and boards across a wide range of industries. The temptation is always to lead with the strongest attribution numbers because they are the most defensible in a spreadsheet. The problem is that leading with attribution numbers trains finance teams to hold marketing to a standard of precision that the discipline cannot honestly meet.

A more honest and, in my experience, more durable approach is to present advertising ROI as a range rather than a point estimate. You show the attributed number, you explain what it is measuring and what it is not, and you show what the incremental data suggests where you have it. You frame the brand-building activity in terms of the long-term revenue trajectory it is contributing to, not the conversions it directly caused.

This approach requires more confidence in the room. It means saying “our attribution model shows X, but the true incremental effect is probably closer to Y, and here is why that still justifies the investment.” Most finance directors respect that kind of honesty more than a polished number that does not survive scrutiny. The ones who do not are the ones who will eventually hold you accountable for a number you invented.

I judged the Effie Awards for several years, and one of the things that distinguished the strongest entries was not the size of the ROI claim. It was the quality of the thinking behind it. The best cases were honest about what they could and could not prove. They used multiple measurement approaches and were transparent about the limitations of each. That rigour is what builds long-term credibility with commercial stakeholders, not a single impressive percentage.

Understanding where untapped revenue potential sits in your go-to-market pipeline is a useful complement to advertising ROI analysis, because it forces the question of whether you are optimising the right part of the system.

The Channels That Report the Best ROI Are Not Always the Best Channels

Paid search consistently reports among the highest ROI of any advertising channel. This makes intuitive sense: you are showing ads to people who are actively searching for what you sell. The intent signal is strong, the conversion rates are high, and the attribution is relatively clean.

What paid search ROI figures rarely account for is how much of that intent was created by other advertising. Someone who sees a display ad on Monday, a social ad on Wednesday, and a TV spot on Friday may convert through a paid search click on Saturday. The search channel claims the conversion. The other channels get nothing. The ROI calculation for paid search looks excellent. The ROI calculation for the channels that built the intent looks poor. Budget decisions follow the numbers, and over time the channels that build demand get cut in favour of the channel that harvests it.

This is not a hypothetical. It is a pattern I have seen play out in real budgets, at real companies, with real consequences for growth. The fix is not to stop measuring ROI by channel. It is to hold those channel-level numbers with appropriate scepticism and invest in measurement approaches that can see across the full system.

If your go-to-market planning treats each channel as a standalone investment with its own independent ROI, you are measuring the parts while missing the whole. The Go-To-Market and Growth Strategy hub has more on how to think about channel mix as a system rather than a collection of individual bets.

What Good Advertising ROI Analysis Actually Requires

Good advertising ROI analysis requires four things that most marketing teams either do not have or do not prioritise.

First, it requires a clear definition of what counts as a return. Revenue is the obvious answer, but for many businesses the more relevant metric is profit, customer lifetime value, or new customer acquisition rather than total transactions. An advertising campaign that drives high revenue from existing customers at low margin may have worse ROI than a campaign that drives lower revenue from new customers at higher margin and higher lifetime value.

Second, it requires an honest accounting of all costs. Media spend is the obvious input, but production costs, agency fees, technology costs, and the internal time spent managing campaigns all belong in the denominator. Campaigns that look efficient on media spend alone often look considerably less efficient when total investment is included.

Third, it requires a time horizon that matches the business model. A subscription business with high customer lifetime value should measure advertising ROI over twelve to twenty-four months, not thirty days. A business that measures ROI on a thirty-day window and then optimises accordingly will systematically underinvest in the channels that are best at acquiring high-value customers.

Fourth, it requires some form of incrementality testing. Not necessarily formal econometrics, though that is worth investing in if the budget justifies it. Even simple holdout experiments, where you deliberately do not advertise to a matched audience segment and compare their conversion rate to the exposed group, will tell you more about true advertising impact than any attribution model can.

Rigorous go-to-market planning treats measurement design as part of the launch strategy, not an afterthought. Building the measurement framework before the campaign runs is the only way to get data that is actually useful for future decisions.

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