New Research with eMarketer Reveals Why MMM Undervalues Affiliate Marketing And What Marketers Can Do About It

Our latest research with eMarketer reveals a critical truth: MMM frameworks aren’t built for affiliate marketing’s trust-driven, always-on nature. Many marketers see strong affiliate performance in practice, yet their models don’t fully reflect it—leading to underinvestment in a channel that drives discovery, engagement, and conversion.

_____________________________________________________________________________________________________________

Over the past year, one topic has come up in nearly every conversation with our merchant partners: Marketing Mix Modeling (MMM) and affiliate marketing’s true impact. While MMM is more relied on than ever, many marketers admit it doesn’t reflect what they know to be true about affiliate.

To explore this, we partnered with eMarketer, surveying 110 U.S. marketers to uncover where MMM falls short, why it matters, and how brands can capture affiliate’s full value. 

The Core Issue: MMM Was Built for Impressions, Not The Relationship-Driven Affiliate Channel
Despite affiliate investments rising 11.3% next year (eMarketer), MMM still treats affiliate as a lower-funnel, uniform input. In reality, shoppers interact with creators, publishers, and loyalty platforms across multi-touch journeys built on trust and repeated exposure. 

Our research reveals two consistent themes:
• Many marketers still struggle to get affiliate properly represented in their models
• Most models rely on time windows or inputs that overlook how affiliate actually works

As eMarketer Principal Analyst Max Willens noted in our conversations, affiliate often isn’t even on the radar of the teams building the models.

Slow Feedback Loops Make Affiliate Hard to Value

MMM insights often arrive long after decisions need to be made. Several marketers in our study shared that by the time results come back, budgets have already shifted or campaigns have ended—leaving affiliate at a disadvantage in planning cycles.

Carl Kalapesi, SVP at Rakuten Rewards, put it simply: if results show up months after activation, they can’t guide decisions in real time.

The result: some brands cut affiliate spend—not because it isn’t performing, but because delayed insights and traditional measurement methods undervalue its impact.

Trust in MMM Isn’t Universal — Especially for Affiliate

Even experienced marketers say they struggle to interpret MMM results for always-on, relationship-based channels, according to our research. Many aren’t confident in the inputs being used, how affiliate data is categorized, or whether the model accounts for factors like delay and decay.

Why This Matters More Now: AI Is Reshaping Discovery

Shoppers are increasingly using AI tools to compare prices, evaluate deals, and validate reviews. These behaviors are rooted in content and trust—core strengths of affiliate.

As affiliate-driven ecommerce continues to expand, the need for accurate, channel-appropriate measurement is only accelerating.

What Marketers Should Do Next

Our full report outlines several practical steps marketers can take, including:
• Improving the frequency and granularity of data inputs
• Exploring other metrics for the affiliate channel
• Leveraging AI-driven discovery trends to advocate for affiliate internally

Download the Full Report

This is a pivotal moment for performance marketers. As AI transforms discovery and MMM becomes more influential in budget decisions, brands that modernize their measurement approach now will be better positioned to capture affiliate-driven growth.

Download the full eMarketer x Rakuten Rewards report to see the complete findings and recommendations.