Measuring What Matters: The CPG Attribution Problem in a Cookieless, Multi-Retailer World

CPG has always presented measurement challenges because brands often don’t control the final transaction. A consumer may see a CTV ad at home, search for the product later on a computer or mobile device, encounter it again on social media, view a sponsored placement on a retailer website, and ultimately buy it during a weekly grocery trip.

Each interaction may influence the purchase, but no individual platform has a complete view of the journey.

The problem gets more complicated when the same product is sold across Amazon, Walmart, Target, grocery chains, delivery platforms, convenience stores, club stores, and other retail environments. Each retailer has its own customer data, reporting standards, attribution rules, and media ecosystem.

At the same time, privacy changes have reduced the availability of third-party signals and individual-level tracking. More platforms now depend on aggregated or modeled conversion data to fill gaps in what they can observe.

Different platforms may also use different attribution windows and definitions of conversion. One retailer might credit a sale seven days after exposure, while another uses a longer window. One may count online purchases, while another incorporates certain store transactions.

This creates an important reality for CPG marketers: no single measurement system has a complete view of the purchase journey. 

Retail Media Network Reporting

Retail media networks give CPG marketers valuable access to commerce data close to the point of purchase. Platforms such as Amazon, Walmart Connect, Target Roundel, Instacart, and other retailer media networks can often connect advertising activity to transactions within their own environments.

This can help marketers understand retailer-specific sales, SKU and category activity, shopper behavior, and campaign performance. That proximity to purchase makes retail media reporting especially useful for evaluating activity within a specific retailer.

But the same strength creates a limitation. Retail media platforms primarily see what happens inside their own ecosystems. They do not necessarily know how exposure on another platform, a CTV campaign, paid search, social media, or an offline campaign influenced the final purchase.

Measurement methodologies also vary across networks, making direct comparisons difficult. Most importantly, attributed revenue is not the same as incremental revenue. A retailer may accurately report that a consumer purchased a product, but that doesn’t necessarily mean the advertising caused the sale.

Retailer data provides valuable evidence, but it shouldn’t be treated as a complete view of media performance.

Multi-Touch Attribution

Multi-touch attribution (MTA) helps connect individual marketing interactions across the customer journey and assign credit to the touchpoints that contributed to conversion. Its greatest strength is detail.

When sufficient data is available, MTA can help marketers evaluate audiences, creative, campaigns, devices, and digital channels at a tactical level. That makes it useful for understanding how observable digital interactions contribute to performance. The challenge is that fewer consumer journeys are fully observable.

Privacy restrictions, device fragmentation, walled gardens, offline purchases, and separate retailer environments create gaps in the user-level data MTA depends on. Those gaps are particularly significant for CPG brands because final purchase activity often occurs outside the brand's own website.

MTA remains useful for tactical digital analysis where enough signal exists, but it is much less reliable as a complete explanation of everything that caused a CPG sale.

Marketing Mix Modeling

Marketing mix modeling (MMM) takes a broader approach. Rather than following individual consumers, MMM analyzes aggregated media, sales, and business data over time to estimate how different investments contribute to outcomes.  

Because it does not depend on individual identifiers, MMM is well suited to an environment where user-level tracking has become more limited. It can evaluate online and offline media together while accounting for factors such as seasonality, pricing, promotions, and distribution. It can also help marketers understand channel contribution, saturation, budget allocation, and potential future scenarios.

MMM typically operates at a broader level than a retail media dashboard or campaign reporting system. It can help determine how investment should shift across channels, but it isn’t designed to tell a media buyer which individual creative variation should receive more budget tomorrow.

That matters because different measurement methods should support different types of decisions. 

Attribution Windows and Conversion Definitions

Measurement systems frequently use different rules for determining which sales count. One platform may credit a purchase that occurs within a certain number of days after an impression or click while another platform may use a different lookback window.

Systems may also define revenue, conversions, new customers, or modeled sales differently. Two platforms can therefore analyze the same campaign and produce different results without either calculation necessarily being incorrect within its own methodology.

Overlapping Credit Across Platforms and Retailers

A consumer rarely interacts with only one form of advertising. Someone might see a social ad, watch a CTV commercial, click a search ad, view a sponsored product on Amazon, and purchase the product at Walmart. Several of those systems may claim some level of credit.

This creates a common problem for marketers. Platform-reported results cannot simply be added together to determine total media impact. Doing so can result in the same sale being counted multiple times across different environments. 

Different Levels of Measurement

Retail media dashboards, MTA, and MMM also operate at different levels.  

• Retail media reporting evaluates performance within a specific retailer environment.  
• MTA evaluates observable consumer-level digital interactions.
• MMM evaluates broader patterns between media investment and business outcomes.

The numbers should not always match because the systems are looking at performance through different lenses.

Attribution vs. Incrementality

One of the most important distinctions in media measurement is the difference between attribution and incrementality.

• Attribution asks which marketing interaction should receive credit for a conversion.
• Incrementality asks whether the marketing activity created an additional outcome that would not have happened otherwise.

That difference matters because consumers may already be likely to purchase a product before they see an ad. A campaign can therefore report strong attributed sales without producing the same level of incremental growth.

For CPG brands evaluating where to increase or reduce investment, understanding incremental impact is often more useful than simply knowing which platform claimed the conversion.

Retail media makes the attribution challenge even more complex. A CPG brand may simultaneously receive reports showing strong performance from Amazon, Walmart Connect, Instacart, Target Roundel, and other networks.

Each report may be valuable within its own environment, but the reported performance cannot automatically be combined and treated as total incremental business impact.

Retailers use different attribution windows, methodologies, audience definitions, and standards for metrics such as new-to-brand sales. Consumers also move between retail environments.

Someone may research a product on Amazon, see the brand again through Instacart, and purchase it at a physical Walmart location. No single retailer has a complete view of how those exposures worked together.

This makes comparison difficult not only between retail media networks, but also between retail media and channels such as search, social, CTV, programmatic, linear television, and out-of-home.

Retailer data is an essential part of CPG measurement, and becomes even more valuable when evaluated within the context of the full media mix. 

Trying to create one universal performance number can create more confusion rather than less. A more practical approach starts with the decision the marketing team needs to make.

Are you deciding whether retail media investment should increase?

Are you comparing investment across retail media networks?

Are you trying to determine whether paid social is creating incremental demand?

Are you deciding where the next media dollar should go?

Are you evaluating whether a channel is becoming saturated?

Different decisions require different evidence.

Match the Measurement Method to the Decision

Retail media reporting is useful for retailer-specific execution, shopper behavior, SKU performance, and campaign activity.

• MTA can support tactical analysis of digital journeys where sufficient user-level data remains available.
• Incrementality testing can help determine whether an investment created additional business results.
• MMM can support broader cross-channel budget allocation, scenario planning, and longer-term investment decisions.

This creates a clearer role for each methodology. The objective is not to make every system produce the same number. It is to use the right evidence for the decision being made.

Align Around Common Business Outcomes

Measurement also becomes more useful when teams agree on business outcomes that extend beyond individual platform metrics. Depending on the brand, those outcomes may include total sales, incremental sales, revenue, household penetration, customer acquisition, repeat purchase, market share, or profitability.

This gives teams a common way to evaluate performance even when the underlying platforms use different reporting methods. It also reduces the risk of treating a platform-specific metric as the final definition of success. A high ROAS inside one platform may look strong, for example, but the bigger question is whether that investment is contributing to additional growth.

Validate Platform Results Independently

Retailer and media platform reporting should be treated as valuable inputs, but those results can be strengthened by independent validation. Incrementality tests, geo experiments, lift studies, and MMM can help determine whether attributed performance is translating into additional business impact.

When these methods support one another, marketers can invest with greater confidence. When they disagree, the gap itself can reveal where further investigation is needed.

Evaluate Retail Media Within the Full Media Mix

Retail media budgets don’t operate in isolation. Search can capture active demand, CTV can build awareness and consideration, social can introduce a product or reinforce its relevance, and retail media can influence the decision close to purchase.

Evaluating those channels separately makes it harder to understand how they work together. CPG brands need a broader view that compares retail media with the rest of the media mix and evaluates where additional investment can create the greatest impact.

• What data can it see?
• What attribution window does it use?
• Does it include offline sales?
• Is it measuring attributed performance or incremental impact?
• Does it account for pricing, promotions, distribution, seasonality, or other external factors?
• What type of decision is the result supposed to support?

A retail media dashboard may be the appropriate tool for adjusting a retailer-specific campaign. MMM may be more useful for deciding whether overall retail media investment should grow relative to CTV or paid social. Understanding those roles helps teams interpret differences instead of spending time debating which dashboard has the correct number. 

Media execution is also becoming more automated. Platforms increasingly use automated bidding, predictive audiences, machine learning, and AI-driven campaign management to decide where and how advertising appears.

These tools can improve execution and help marketers respond quickly to changing conditions. They also make independent measurement more important.

When a platform controls more of the bidding, targeting, placement, and optimization process, marketers need reliable ways to evaluate whether those automated decisions are actually contributing to growth.

Platform reporting can explain what happened inside the platform. Independent measurement helps determine whether those results translated into incremental value. 

USIM evaluates retail media as part of the complete media mix rather than as a separate investment that should be judged solely by retailer-reported performance. Our channel-agnostic approach helps marketers compare retail media, search, social, CTV, programmatic, traditional media, and other investments against the business outcomes each is expected to support.

Retail and audience data help inform planning. Marketing mix modeling, incrementality testing, and cross-channel analysis provide additional ways to evaluate performance independently. Measurement is also connected to ongoing budget decisions. Rather than relying on a single report at the end of a campaign, marketers can use performance evidence to determine where investment should increase, where it may be reaching saturation, and where another channel may create greater impact.

The goal is a clearer understanding of what is contributing to growth and where investment should move next, not just another dashboard. 

CPG measurement will continue to be complex because consumers do not shop within a single platform, retailer, device, or channel. Perfect attribution is unlikely to resolve the issue. A better measurement approach gives marketers a clear understanding of what each system can tell them, where its limitations begin, and when additional validation is needed.

Retail media reporting, MTA, incrementality testing, and MMM can all provide valuable information when each has a defined role. When CPG teams align those methods around shared business outcomes, measurement becomes less about deciding whose number is right and more about making better investment decisions.

USIM helps CPG marketers connect retail media, cross-channel performance, and independent measurement to understand what is driving growth.

Connect with USIM to build a measurement strategy that gives your team greater confidence in where to invest next.