A new study from Prohaska Consulting, commissioned by Rakuten Rewards, is challenging one of the marketing industry’s most trusted measurement frameworks. The report argues that traditional Marketing Mix Modeling (MMM) systematically undervalues affiliate marketing—potentially leading brands to misallocate budgets and lose competitive ground.

For decades, Marketing Mix Modeling has served as a cornerstone of enterprise marketing analytics, helping brands allocate budgets across channels like TV, search, and digital media. But according to new research titled “The Next Frontier of Measurement: Fair Evaluation of Affiliates in Marketing Mix Models,” the framework may be failing to keep pace with the evolving dynamics of performance marketing.

The report, based on interviews with more than two dozen senior marketing leaders and input from leading MMM providers, concludes that affiliate marketing—despite being widely adopted—does not fit cleanly into traditional measurement models. As a result, its true contribution to revenue is often undercounted.

Affiliate marketing, at its core, is a performance-driven channel where publishers earn commissions based on completed transactions. This structure makes it highly efficient. Data cited in the report shows that more than 80% of marketers use affiliate programs, with platforms like Rakuten Rewards delivering returns as high as 17 times the cost of commissions. Yet MMM frameworks frequently misinterpret this performance.

The issue stems from how MMM models attribute value. These models are designed to analyze correlations between marketing inputs—such as impressions, clicks, and spend—and business outcomes. Affiliate marketing, however, operates closer to the point of conversion. Because its activity aligns directly with sales, MMM often assumes affiliates are merely capturing existing demand rather than generating new demand.

This structural bias has real financial implications. In one case study highlighted in the report, two brands under the same parent company ran affiliate campaigns simultaneously. The actively managed program, which included dynamic cashback incentives and time-sensitive promotions, achieved up to 25 times higher return on ad spend compared to a passive approach. Yet without granular modeling, such performance differences can remain invisible.

Another example underscores the risk of misinterpretation. A major retailer paused its affiliate program after MMM analysis questioned its incremental value. The result was a sharp decline in customer volume—over 50%—as shoppers shifted to competitors. Even after reinstating the program, the retailer struggled to recover lost market share.

These findings point to two fundamental flaws in how MMM evaluates affiliate marketing.

First, there is a lack of standardized data. Affiliate marketing encompasses a wide range of sub-channels, including cashback platforms, coupon sites, influencer partnerships, and content publishers. However, MMM models often group all these under a single “affiliate” category. This aggregation obscures performance differences and prevents marketers from identifying high-performing segments.

Second, MMM frameworks rely heavily on upper-funnel metrics such as impressions and clicks—data points that are not always available or relevant in affiliate ecosystems. Since affiliates are typically compensated on conversions, they generate less of the upstream data MMM depends on, creating gaps in analysis.

The report argues that MMM alone is insufficient for evaluating affiliate marketing. Instead, it calls for a hybrid measurement approach that combines MMM with incrementality testing and granular data segmentation.

For enterprise marketing teams, the implications are significant. Misjudging affiliate performance can lead to underinvestment in a channel that reaches high-intent consumers at the moment of purchase—an increasingly valuable capability in a privacy-first digital landscape.

Industry leaders are beginning to recognize the need for change. Platforms across the martech ecosystem—from Google and Microsoft to Adobe and Salesforce—are investing in more advanced attribution and data modeling tools. These systems aim to integrate multiple data sources and provide a more nuanced view of customer journeys.

The Prohaska report outlines several practical steps to address the issue. Marketers are encouraged to break affiliate marketing into sub-categories within MMM datasets, track promotional variables such as cashback rates over time, and calibrate models using controlled experiments like geo-based holdouts.

Publishers, meanwhile, are advised to provide richer datasets—including impression and click data—and develop testing capabilities that align with MMM requirements. The report also calls for industry-wide standards to classify affiliate sub-channels, enabling more consistent and accurate measurement.

Ultimately, the research highlights a broader shift in marketing analytics. As performance channels become more complex and data ecosystems more fragmented, relying on a single measurement framework is no longer sufficient.

For CMOs and marketing analysts, the message is clear: understanding the limitations of MMM is just as important as leveraging its insights. In an environment where every budget decision carries strategic weight, measurement accuracy can determine whether brands capture growth—or cede it to competitors.

Market Landscape

The findings arrive at a time when marketing measurement is under intense scrutiny. According to Gartner, nearly 60% of CMOs report dissatisfaction with their current attribution models, citing challenges in cross-channel visibility and data fragmentation. Meanwhile, McKinsey & Company estimates that companies using advanced analytics and attribution frameworks can improve marketing ROI by 15–20%.

As privacy regulations limit access to user-level data, MMM has regained popularity due to its reliance on aggregated data. However, this resurgence also exposes its limitations—particularly in performance-driven channels like affiliate marketing. The Prohaska research suggests that the next evolution of MMM will depend on its ability to integrate granular, real-time data and adapt to modern digital ecosystems.

Top Insights

• The Prohaska study finds MMM systematically undervalues affiliate marketing by misattributing conversion-driven activity, leading enterprise brands to underinvest in a high-ROI, intent-driven channel.
• Affiliate marketing delivers up to 17x return on commissions, yet structural modeling limitations prevent accurate performance measurement across cashback, content, and influencer sub-channels.
• Case studies reveal misinformed MMM decisions can reduce customer volume by over 50%, demonstrating the financial risks of incomplete attribution frameworks.
• The research recommends hybrid measurement approaches combining MMM with incrementality testing, geo experiments, and granular affiliate segmentation for improved accuracy.
• Enterprise martech platforms from Google, Adobe, and Salesforce are evolving toward integrated attribution models, signaling a broader shift beyond traditional MMM frameworks.

Get in touch with our MarTech Experts