Every retail media platform comes with a dashboard. Amazon gives you ACOS and ROAS. Walmart Connect provides attributed sales. Instacart shows conversion rates. Loblaw Advance delivers campaign performance reports.
The data is there. The question is whether it’s telling you what you actually need to know.
Platform metrics are useful for tactical optimization—adjusting bids, refining targeting, managing budgets. But they have significant limitations when it comes to understanding what’s really driving your business.
Channel-level reporting tells you what happened. It doesn’t tell you what’s working.
Key Takeaways
- • Platform metrics are optimized for platform goals, not necessarily brand goals
- • ROAS and attributed sales can look strong while overall business performance stagnates
- • Incrementality—whether retail media is driving sales that wouldn’t have happened otherwise—is the question most brands can’t answer
- • Cross-channel effects and halo impacts are invisible in platform-level reporting
- • Moving to business outcome measurement requires defining success differently and investing in unified reporting
Why Are Platform Metrics Incomplete?
Platform metrics are designed by platforms, for platforms. They’re optimized to demonstrate the value of advertising on that platform—which isn’t necessarily the same as demonstrating value to your business.
A few structural limitations:
Attribution windows vary
Amazon’s attribution window is different from Walmart’s, which is different from Instacart’s. When you compare ROAS across platforms, you’re not comparing apples to apples. A sale attributed on one platform might not be attributed on another—not because performance differs, but because measurement rules differ.
Last-click bias:
Most retail media attribution is last-click or last-touch. The platform that gets credit is the one the shopper interacted with most recently before purchasing. This systematically undervalues upper-funnel activity (awareness, consideration) and overvalues lower-funnel activity (search ads at point of purchase).
If you ran a connected TV campaign that introduced a shopper to your brand, and they later searched for you on Amazon and clicked a sponsored product ad, Amazon gets full credit. The CTV campaign shows zero attributed sales. But without it, the Amazon conversion might never have happened.
No view of organic cannibalization:
Platform reporting shows sales attributed to your ads. It doesn’t show whether those sales would have happened anyway through organic search or direct navigation. If a shopper was going to buy your product regardless, and your ad just intercepted them on the way, you’ve paid for a sale you would have gotten for free. This is the incrementality question—and most platform reporting can’t answer it.
Cross-channel effects are invisible:
A shopper sees your ad on Instacart. They don’t buy immediately. Later, they purchase in-store at Loblaw. Did the Instacart ad influence the purchase? Platform reporting says no—there’s no attributed sale. But the ad may have done exactly what it was supposed to do: build awareness and consideration that converted elsewhere.
What Should You Be Measuring Instead?
Moving from platform metrics to business outcomes requires asking different questions.
Total sales across channels, not just attributed sales
Instead of looking at each platform’s attributed sales in isolation, look at total brand sales across all channels—including those without attribution. If retail media is working, total sales should be growing. If platform ROAS looks strong but total sales are flat, you may be cannibalizing organic or shifting sales between channels rather than creating new demand.
Share of voice and share of market
Are you winning or losing ground to competitors? Platform metrics won’t tell you this directly. You need category-level data—search share, shelf share, market share—to understand whether your media investment is translating to competitive position.
Incrementality
This is the hardest metric to capture, but the most important. What portion of attributed sales would have happened without the ad? Answering this requires testing—holdout groups, matched market tests, or statistical modeling—that goes beyond standard platform reporting.
Customer acquisition vs. retention
Are your retail media dollars reaching new customers or existing ones? Most platforms don’t distinguish well between these. But the strategic value is very different. Acquiring a new customer justifies higher CPAs than retargeting someone who was already going to buy.
Blended efficiency across channels
Instead of optimizing each channel’s ROAS independently, look at blended CAC (customer acquisition cost) and LTV (lifetime value) across all commerce channels. This prevents over-investment in channels that look efficient in isolation but don’t contribute to overall business growth.
How Do You Get There?
Shifting to business outcome measurement isn’t a flip you switch. It requires investment in capabilities that most brands don’t have out of the box.
- 1. Unified reporting across platforms: Bring Amazon, Walmart, Instacart, and Loblaw data into a single view. This doesn’t require a massive CDP—even a well-structured spreadsheet or dashboard that pulls from multiple sources is a start. The goal is seeing total performance, not just channel performance.
- 2. Incrementality testing: Design and run tests that isolate the causal impact of retail media. The simplest approach is geographic holdouts—running media in some markets and not others, then comparing sales. More sophisticated approaches use matched market analysis or synthetic control groups.
- 3. Cross-channel attribution modeling: Invest in understanding how channels influence each other. This might mean working with a measurement partner, building internal models, or simply tracking leading indicators (site traffic, search volume, brand awareness) alongside conversion metrics.
- 4. Aligned KPIs across teams: If your Amazon team is measured on Amazon ROAS and your brand team is measured on market share, they’re optimizing for different outcomes. Align incentives around business results—total sales, share growth, profitability—not just channel metrics.
- 5. Regular business reviews, not just campaign reviews: Shift the conversation from “how did our campaigns perform?” to “how is our business performing, and what role is retail media playing?” This requires bringing together data from multiple sources and asking harder questions about causation.
The Bottom Line
Platform metrics are a starting point, not a destination. They’re useful for managing campaigns day-to-day, but they don’t answer the questions that matter most: Is retail media driving business growth? Are we reaching new customers? Would these sales have happened anyway?
Brands that make the shift from platform metrics to business outcomes don’t just run better campaigns. They make better investment decisions—allocating budget based on what’s actually working, not just what the dashboards say is working.
If your reporting isn’t answering the questions that matter, we can help you figure out what’s missing.
Get in touch → commercemediaagency.co
Visit commercemediaagency.co to learn more about our approach.
Published by Commerce Media Agency, powered by geekspeak Commerce - combining two decades of ecommerce expertise with deep commerce media strategy and content execution capabilities.