Measuring marketing ROI is critical, but the consensus among operators is that the most common approach, tracking Return on Ad Spend (ROAS) within ad platforms, is fundamentally broken. It creates a false sense of security while often failing to measure what actually matters: incremental profit. The real task isn't just tracking revenue against spend; it's building a measurement system that reveals the true, causal impact of your marketing on the entire business.
The core problem with ROAS is that it’s an efficiency metric, not an effectiveness metric. As Dave Levett argued on eCommerce Australia, it's the "greatest myth in eCommerce" because it takes credit for all sales it touches, regardless of whether the ad actually caused the purchase. A customer who was going to buy anyway might click a retargeting ad, making the campaign look successful when it had no real impact. This issue is magnified across a business, leading to the absurd situation Andrew Lipsman and Meghan Corroon described on Ecommerce Braintrust, where the siloed ROI reports from different marketing channels can add up to more than a company's total revenue. The team at Ecommerce Playbook has taken this a step further, arguing that Meta's own attribution settings are now so divorced from reality that they've become almost irrelevant for serious analysis.
Instead of obsessing over a single, flawed number, a much better approach is to adopt what Tony Chopp from Ecommerce Playbook calls the Hierarchy of Metrics. This framework provides a structured way to diagnose business performance. It starts at the top with your ultimate goals, like net revenue and contribution margin, and works down through customer-level data (New vs. Returning, LTV), channel-level performance (like blended Marketing Efficiency Ratio, or MER), and finally to specific campaign inputs. This structure prevents you from over-optimizing one metric, like ROAS, at the expense of overall business health. It forces you to ask better questions, shifting from "What's my ROAS?" to "Did we hit our profit goal, and if not, was it a traffic, conversion, or LTV problem?"
To understand the real cause-and-effect relationship, you have to measure incrementality. This is the central challenge of modern attribution. The goal is to isolate the sales that happened only because of a specific marketing activity. The most reliable ways to do this involve moving beyond what the platforms tell you. On Ecommerce Playbook, the hosts make a strong case for using methods like geo-based split tests, ghost ads (where a control group sees a non-functional ad), and brand lift studies. For a more comprehensive view, many experts now point to Marketing Mix Modeling (MMM), an econometric approach that analyzes the contribution of all marketing channels and external factors to your sales over a longer time horizon. MMM helps you see the whole picture, not just the last click.
Putting this into practice can feel intimidating, but it doesn't require a massive data science team from day one. The first step is to change what you measure and what you value. Stop celebrating a high ROAS in your Monday meetings. Instead, build a dashboard focused on more holistic business health metrics like MER (Total Revenue / Total Marketing Spend), Customer Acquisition Cost (CAC), and the ratio of LTV to CAC. Meghan Corroon on Ecommerce Braintrust compared this to managing a stock portfolio; you need a unified view to see where you're genuinely over- or under-invested. As the team on Future Commerce puts it, "you get what you measure." If you start measuring profitability and incrementality, you'll start optimizing for them.
Ultimately, moving toward a more accurate measure of ROI is a process of graduating from simple efficiency stats to a more robust model of business impact. It begins with a healthy skepticism of platform-reported data and a commitment to building a unified view of your performance. By expanding your toolkit to include blended metrics, a clear hierarchy of goals, and a strategy for testing the true incrementality of your ads, you can stop guessing about what's working and start making decisions that drive real, sustainable profit.