Digital Advertising ROI: Why Most Dashboards Are Lying to You

Digital advertising ROI is the ratio of revenue generated to money spent on paid digital channels, but the number most marketers report bears little resemblance to what is actually happening in the business. Attribution models inflate returns, last-click logic steals credit from channels that did the real work, and platform-reported ROAS has a well-documented interest in looking as good as possible. Getting an honest read on digital advertising ROI requires more than a dashboard. It requires a framework for separating signal from noise.

I have managed hundreds of millions in ad spend across 30 industries over two decades. The single most consistent pattern I have seen is that the closer you get to the actual business numbers, the less impressive the digital advertising story becomes. That is not an argument against digital advertising. It is an argument for measuring it honestly.

Why Platform ROAS and Business ROI Are Not the Same Thing

Every major ad platform has a structural incentive to show you strong returns. Google, Meta, and the rest are selling you media. Their attribution models are built to claim as much credit as possible for conversions that were going to happen anyway. This is not a conspiracy. It is just how incentives work.

When I was running performance marketing at scale, one of the first things I learned to do was pull the platform numbers and then sit down with the finance team. Nine times out of ten, the gap was significant. A campaign reporting 6x ROAS inside Google Ads might be generating 2x when you account for returns, cancellations, and the customers who would have converted through organic search regardless. That delta is where the real conversation about digital advertising ROI has to start.

Platform attribution models, even the more sophisticated data-driven ones, are trained on platform data. They cannot see what happens off-platform. They cannot see the TV ad someone watched three days before clicking a paid search result. They cannot see the word-of-mouth recommendation that preceded the branded search. They see the click, and they take the credit.

This matters because decisions get made on these numbers. Budget allocation, channel mix, campaign scaling , all of it flows from ROAS figures that are often materially overstated. If you are optimising toward a false signal, you are systematically misallocating spend. Over time, that compounds.

The Attribution Problem Is Not Going Away

The marketing industry has been trying to solve attribution for as long as paid digital has existed. Last-click, first-click, linear, time-decay, data-driven, multi-touch. Each model is a different theory about how customers make decisions. None of them is correct. All of them are useful, in the same way that a map is useful even though it is not the territory.

The problem with attribution is not the models themselves. The problem is treating any single model as the definitive answer. I have seen marketing teams spend months debating which attribution model to adopt, as if finding the right model would solve the measurement problem. It will not. The honest position is that attribution gives you a directional view of performance, not a precise accounting of cause and effect.

What attribution does well is help you understand relative channel contribution within a controlled framework. What it does poorly is tell you whether your advertising is actually generating incremental revenue or just capturing revenue that would have happened anyway. That distinction is the most important question in digital advertising ROI, and attribution models are not built to answer it.

Understanding how digital advertising fits into a broader commercial framework matters here. If you are building out a go-to-market strategy, the Go-To-Market and Growth Strategy hub covers how to connect channel investment to business-level outcomes rather than platform metrics.

Incrementality: The Metric That Actually Answers the Right Question

Incrementality testing asks a simple question: what would have happened if we had not run this advertising? The answer is the true measure of digital advertising ROI. Everything else is a proxy.

The mechanics are straightforward. You create a holdout group , a segment of your audience or geography that does not see your advertising. You measure the difference in conversion rates, revenue, or whatever business outcome you care about between the exposed group and the holdout group. The difference is your incremental lift. That is the revenue you can genuinely attribute to the advertising.

Early in my career at lastminute.com, we ran a paid search campaign for a music festival and saw six figures of revenue land within roughly a day. It felt like magic. But the honest question, one I did not ask rigorously enough at the time, was how much of that revenue would have come through organic search and direct traffic anyway. The brand had genuine demand. Paid search was capturing some of it, but it was also taking credit for demand it had not created. Incrementality testing is how you separate those two things.

The challenge with incrementality testing is that it requires discipline. You have to be willing to deliberately not advertise to a portion of your audience, which feels counterintuitive when you believe your advertising is working. You also need enough volume to generate statistically meaningful results. For smaller advertisers, this can be genuinely difficult. But even a rough holdout test is more informative than relying entirely on platform-reported attribution.

For teams looking at how incrementality fits within a broader market penetration strategy, Semrush’s breakdown of market penetration approaches is worth reading alongside your measurement planning.

Demand Capture vs. Demand Creation: Why the Distinction Changes Everything

Most digital advertising, particularly paid search, is demand capture. It intercepts people who are already looking for something. This is valuable. But it is categorically different from demand creation, which is advertising that generates interest and intent that would not otherwise exist.

The ROI calculation looks very different depending on which one you are doing. Demand capture advertising tends to show strong short-term ROAS because it is converting people who are already close to a purchase decision. Demand creation advertising tends to look weak on short-term metrics because the conversion happens later, through different channels, after the advertising has done its work upstream.

This is one of the reasons performance marketing has a structural bias toward the bottom of the funnel. The numbers look better there. But if everyone in your category is competing for the same pool of existing demand, you are in a bidding war for market share rather than growing the market. The economics of that get worse over time as CPCs and CPMs inflate.

I spent years running agency P&Ls where the growth model was built almost entirely on paid search efficiency. It worked, until it did not. When competition intensified and CPCs rose, the margin that had looked healthy started to compress. The businesses that held up best were the ones that had invested in brand and demand creation alongside performance. They had a lower cost base of organic and direct traffic that acted as a buffer when paid costs increased.

The Vidyard research on untapped pipeline potential for go-to-market teams points in the same direction: the biggest revenue opportunity is often not in optimising existing demand capture but in identifying where latent demand exists and building toward it.

How to Build a More Honest ROI Framework

An honest digital advertising ROI framework has four components: business-level revenue tracking, platform data as one input among several, incrementality testing where volume allows, and a clear view of the full customer acquisition cost including all overhead.

The first step is reconciling platform data against actual business outcomes. Pull your platform-reported conversions for a period and compare them against your CRM or finance system for the same period. If the numbers do not match, understand why. Double-counting is common. Multiple platforms claiming credit for the same conversion is the norm, not the exception. The sum of platform-reported ROAS almost always exceeds the actual business return.

The second step is getting your true customer acquisition cost. This means including agency fees, internal team time, creative production, and technology costs alongside media spend. Most ROAS calculations only include media spend in the denominator. When you add the full cost of running the advertising function, the return looks materially different. I have seen businesses that believed they were running at 5x ROAS discover they were closer to 2x once full costs were included.

The third step is segmenting your analysis by customer quality. A campaign that drives high volume at low cost might be generating customers with poor retention or low lifetime value. ROAS calculated on first purchase can be deeply misleading if those customers churn quickly. Connecting advertising performance to cohort-level LTV is the only way to know whether you are building a business or just buying revenue.

The fourth step is running your advertising against your growth strategy, not just your channel targets. BCG’s work on go-to-market strategy in financial services makes a point that applies broadly: channel investment decisions need to be anchored in where your customers actually are in their decision experience, not where your measurement tools are most comfortable reporting.

The Role of Creative in ROI Outcomes

Most digital advertising ROI conversations focus on targeting, bidding, and attribution. Creative gets discussed as an afterthought. This is a significant mistake. Creative quality is one of the highest-leverage variables in advertising performance, and it is consistently underinvested in relative to media spend.

When I judged the Effie Awards, the campaigns that demonstrated genuine business effectiveness almost always had a clear creative idea at their centre. Not a clever execution. A clear idea. The ones that failed were often technically sophisticated in their targeting and measurement but had nothing interesting to say. Precise targeting of an uncompelling message is still an uncompelling message.

The digital advertising ecosystem has made it very easy to optimise the mechanical elements of a campaign and very easy to ignore whether the advertising itself is any good. Auto-bidding, responsive ads, and dynamic creative optimisation can all improve efficiency at the margin. None of them can fix a fundamentally weak creative concept. And a weak creative concept caps your ROI ceiling regardless of how well everything else is set up.

BCG’s research on brand strategy and go-to-market alignment highlights the connection between brand strength and commercial performance. Brands with clear positioning and strong creative tend to convert better at every stage of the funnel, which means their paid media works harder for the same spend.

What Good Looks Like: Benchmarks Without the Bullshit

The internet is full of ROAS benchmarks. Most of them are not useful because they aggregate across wildly different categories, margins, and business models. A 3x ROAS might be excellent for a low-margin e-commerce business and catastrophic for a high-margin SaaS company. The benchmark that matters is the one derived from your own economics.

The starting point is your break-even ROAS. This is the minimum return required for your advertising to cover its costs, calculated using your gross margin and your total cost of running the advertising function. If your gross margin is 40% and your advertising costs represent 10% of revenue, your break-even ROAS is 2.5x. Anything below that is destroying value regardless of what the platform dashboard says.

Once you know your break-even ROAS, you can set a target that accounts for the gap between platform-reported and actual returns. If your historical reconciliation shows that platform ROAS overstates actual returns by 40%, your target platform ROAS needs to be set accordingly higher to ensure the business is genuinely profitable after the adjustment.

Growth hacking frameworks, like those covered in Semrush’s analysis of growth hacking examples, often treat rapid scaling as the goal. The more useful lens for mature advertisers is sustainable unit economics. Scaling a campaign that looks profitable on platform metrics but is not profitable at the business level just accelerates the problem.

Connecting advertising ROI to broader growth strategy is something I cover in more depth across the Go-To-Market and Growth Strategy hub, particularly how channel investment decisions should flow from commercial strategy rather than channel availability.

The Measurement Stack That Actually Helps

There is no single tool that gives you a complete picture of digital advertising ROI. The measurement stack that works is one that triangulates across multiple data sources and treats each as a partial view of a larger reality.

Platform analytics give you granular channel-level data with known attribution bias. Google Analytics or your preferred web analytics tool gives you a cross-channel view with different but still imperfect attribution. Your CRM gives you customer-level data that can be connected back to acquisition source. Your finance system gives you actual revenue and margin. Media mix modelling, for businesses with sufficient scale and history, gives you a statistical view of channel contribution that is not dependent on individual user tracking.

The job of a marketing leader is to hold all of these views simultaneously and make judgment calls about where the truth lies between them. This requires genuine analytical literacy, not just the ability to read a dashboard. It also requires the confidence to tell stakeholders that the number they want, the clean single ROAS figure, does not exist in the way they think it does.

Early in my career, when I built my first website from scratch after being refused budget, I learned something that has stayed with me: the willingness to do the unglamorous technical work yourself gives you an understanding of the system that you cannot get from a summary report. The same principle applies to measurement. Marketers who understand how their tracking actually works, at a technical level, make better decisions than those who treat the dashboard as gospel.

Hotjar’s work on growth loops and feedback mechanisms is a useful complement here. Understanding how users actually behave on your site, not just which channel they arrived from, adds a qualitative layer to the quantitative measurement stack that often reveals why campaigns are or are not converting.

For brands working with creators as part of their paid and organic mix, Later’s resource on go-to-market campaigns with creators addresses how to think about attribution when influencer content is part of the conversion path, which is a genuine measurement challenge that most standard frameworks handle poorly.

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