Stop Flying Blind: How Smarter Attribution and Incrementality are Transforming Affiliate & Influencer Marketing #67

Executive Summary

In today’s rapidly shifting marketing landscape, brands can no longer afford to “fly blind” when measuring the actual impact of their affiliate and influencer campaigns.

With rising media costs, evolving privacy regulations, and disruptive macroeconomic events—like the recent U.S. tariff hikes that sent the S&P 500 tumbling—marketers face unprecedented pressure to prove ROI and optimize every dollar spent.

Yet traditional attribution models, such as last-click and even multi-touch attribution (MTA), often fall short. They can overcredit specific channels while masking inefficiencies and undervaluing upper-funnel efforts.

This article breaks down the latest thinking and practical solutions around attribution and incrementality in affiliate and influencer marketing.

It explains why relying solely on platform-reported data is risky, how advanced statistical approaches like Media Mix Modeling (MMM) are gaining traction, and why incrementality testing is essential for understanding what drives conversions.

Drawing on real-world examples, expert insights, and the latest SaaS tools, we outline actionable strategies for brands to:

• Move beyond outdated attribution models and embrace holistic, data-driven measurement.

• Leverage MMM and holdout testing to isolate the true incremental value of each marketing channel, even in complex multi-channrel affiliate ecosystems. (Yes, it can be done with affiliate influencer despite some skeptics).

• Avoid common pitfalls—such as over-reliance on last-click or blanket dismissals of coupon and cashback partners—by adopting nuanced, brand-specific approaches.

• Implement practical steps & modern tech using tools like GA4, Impact Optimize, Northbeam, and Paramark to triangulate attribution and incrementality, reduce double-counting, and improve marketing efficiency.

Ultimately, the article empowers marketers, CMOs, and finance leaders to make smarter, more confident decisions, ensuring that every marketing dollar is accounted for and every channel’s contribution is accurately measured, even in turbulent times.

Attribution and Incrementality

Have you been in the dark on attribution and incrementality in affiliate or influencer marketing?

You are not alone.

And this is for you.

No one wants to fly blind in your marketing.

Especially when we have the technology to see how it’s performing.

Sadly, technology, tracking, intelligent people, and budget are often NOT ENOUGH!

This week’s issue of the RBL Flywheel outlines RBL's approaches to attribution and incrementality. It helps our Account Managers and clients improve their outcomes at the lowest cost and friction possible.

Hopefully, you learn something and it helps you too!

Shout out to Al Pacino. His birthday is April 25, so who better to help us ring in April’s Attribution month?

Macro Environment

The last couple of weeks were eventful, to say the least, with stock markets reeling from the recent US tariff announcements. 

President Trump announced a 10% global tariff on all imports effective April 5, 2025, with higher reciprocal tariffs (up to 50%) targeting 57 countries starting April 9.

The S&P 500 has declined by 9%, marking one of its worst weekly performances in 25 years. 

While the market rebounded with the announcement of a 90-day pause sans China, events like this underscore the importance of an effective MMM strategy and ensuring your marketing analysis considers macro factors like these.

Pranav Piyush, the CEO of Paramark, underscores this point here:

Challenge

Clients, the head of affiliate marketing, VPs, CMOs, and Finance want to understand whether the dollars they spend on a marketing channel or campaign helped drive a valued action (lead, SQL, purchase).

Sadly, affiliate marketing consultants/agencies and paid media agencies often lack data rigor.

Media prices are higher than ever due to various factors, including privacy regulations and supply-demand limits. (Yes, the recent tariff policies also brought down CPMs on Meta, as Temu has taken a big hit.)

Thanks for letting me digress on Temu.

Anyway, media cost is still high! As a result, everyone wants to get the most out of their marketing and ensure they have a clear view and better measurement, attribution, and incrementality.

Channels like Meta and Google often overcount and take more credit than they deserve. See their default settings. 

Last-click attribution skews the incentives to reward lower funnel / later funnel marketing exposure when customers are closer to the moment of purchase.

While easy to interpret, last-click attribution distorts marketing efficacy by ignoring the entire customer journey. It overcredits the final touchpoint (e.g., retargeting ads, coupon/loyalty, branded search) while undervaluing upper-funnel efforts.

This is why the question of incrementality is so critical and brands keep asking for it - rightfully so!

Enter MMM, the hottest topic in measurement right now.

MMM (Media Mix Modeling) is a statistical model (econometric modeling) that uses data over time to help brands better calculate the value of marketing channels, campaigns, and tests.

Interest in and use of this statistical model are rising, and we estimate that more brands will rely on it to make marketing decisions.

MMM uses causal inferences, regression models, and AI applied to data science, not LLM-based AI.

MMM looks at various factors like weather, macroeconomic conditions, interest rates, current events, TV ad spend, sale/promo, landing page, and creative, and looks for correlation to value the incrementality of each of these factors/levers to isolate and measure the actual value of your marketing more accurately. 

It has come back into favor lately as SaaS solutions have emerged, offering MMM at a lower cost and leveraging more AI data to make it more accessible. 

Clients have become accustomed to measuring and testing the incrementality of Search and Social via lift tests, MMM, and holdout testing.

The nature of affiliate marketing, whereby brands pay AFTER the action by definition, makes attributing causality to affiliate marketing in MMM models very difficult. Using MMM to understand the incrementality of affiliate marketing is also tricky.

MTA is different and defined in this case as Multi-Touch Attribution.

Northbeam and Triple Whale have focused on MTA as a software solution.

Definitions

• Attribution - Attribution is the process of assigning credit to specific marketing channels for their role in driving customer actions or purchases. It helps determine which channels are most effective in influencing customer decisions. 

• Incrementality - Incrementality measures whether a marketing channel or campaign caused an action or purchase. It asks if the outcome would have happened without the channel or campaign.

• MMM - MMM is a statistical approach used to analyze the impact of various marketing channels on business outcomes. It helps marketers understand how different marketing elements contribute to sales and revenue.

•  A holdout test is a simple way to check if something new works by keeping a group of people or data separate from the change, and then comparing what happens to them with those who got the change.

Common Attribution Models

1. First Click - in the path to conversion, the first click generated is awarded credit for the action/purchase.

2. Last Click - credit goes to the last click

3. Multi-Touch - credit evenly distributed across multiple touchpoints

4. Time Decay - credit distributed across touch points, with the first getting the most and the last getting the least credit 

5. Custom - you get the idea

6. One of the best write-ups on the topic of Attribution Models here.

Despite the rise of more sophisticated attribution models, last-click attribution remains the most commonly used model and is the default for most marketing platforms.

RBL Recommended Solutions for Affiliate & Influencer Clients

Vendors charge $50k annually for MMM MTA and Incrementality Tools on the market.

Meta lift tests typically require a minimum of $120k in media spending and three months of testing. 

Google and YouTube often require similar time and investment to measure accurate lift tests.

Paramark MMM requires doing some proper testing to answer the Incrementality questions. 

• 2 years of data to plug into a tool

• $2M per year in media spend (bare minimum)

• At least three channels invested in, for example, Google, Meta, Affiliate, and Influencer.

Affiliate clients will sometimes question the “incrementality” of a typical partner.

Coupon, Deal, Cash Back, and Browser Extensions are often cited as a concern in this case.

It has become fashionable to “throw the baby out with the bath water” and state that these partner types have no value. Agencies and providers have used this as a sales tactic. 

We take a more nuanced approach.

The reality is that consumers of all income levels love deals, and millions are trained to look for promo codes, get a deal, and apply cash back, airline miles, bitcoin, 529 fund dollars for their kid's college, you name it, to their digital purchases.

We do not want to work with all partners in all situations or all partners in each partner type; instead, we carefully select the right ones for each brand partner's promotional strategy and growth goals.

This is where the RBL Trust Score comes into play, ensuring the ideal match between brand and partner.

Stay tuned for more from us on “Trust Score” coming soon.

We are highly selective when matching a browser extension and an attribution strategy. For example, last-click attribution and Honey will not work well for the brand unless the affiliate network has a stand-down policy.

Here are some light ways to help brands better understand the value of their affiliate marketing:

GA4 correctly instrument GA4 with tracking tied to the affiliate network and comparing multiple attribution models to see how GA4 is crediting Affiliate partners and not relying on Google’s default attribution.

1. Look at last-click attribution

2. Look at first-click attribution

3. Look at a multi-touch attribution view where possible 

4. Ensure UTM parameters are set up properly to track accurately both within affiliate network tracking and Google Analytics (GA4).

5. Please note that this should function beyond GA4 and can work with similar tools like Adobe, Mixpanel, and Amplitude

Impact Optimize

1. This is an add-on feature to the Impact.com SaaS tool that allows brands to see the multi-touch view not just within affiliates but also within multiple marketing channels/touchpoints.

2. Impact and the brands they work with report a 20% decrease in costs after implementing Optimize, which removes double-counting and provides a more accurate view of attribution.

3. Subway stop view of the multiple touch points to conversion and improve de-duplication of actions, and improve marketing efficiency.

4. This can be compared to GA4 and other measurement tools to triangulate value with some level of objectivity and go beyond last click only.

Measurement Tools to Consider:

Northbeam

Multi-Touch Attribution (MTA): Northbeam provides almost real-time conversion attribution using first-party data, which is highly effective for understanding the customer journey across different touchpoints.

Media Mix Modeling (MMM+): Northbeam's MMM+ solution is digitally native. It focuses on optimizing channel mix and forecasting revenue. It incorporates machine learning to provide insights into complex channel interactions.

Paramark

• Essentials Plan: $6,000 monthly, ideal for startups with limited marketing channels.
  

• Advanced Plan: $12,000 monthly, suitable for scale-ups with multiple channels and significant marketing spending.
  

• Enterprise Plan: $20,000+ per month, with custom pricing for established organizations with advanced needs

Time to Get Learnings on Incrementality

The time it takes to obtain learnings from incrementality testing with Paramark depends on several factors, including the complexity of the tests, the size of the audience, and the type of incrementality testing used (e.g., conversion lift, geo testing, or time testing)

General Timeline:

• Setup and Planning: 1-4 weeks - This involves setting up the experiment, defining control and test groups, and planning the test duration.
  

• Test Execution: 2-12 weeks - The actual test period can vary significantly based on the type of test and the desired statistical significance.
  

• Analysis and Insights: 1-4 weeks - After the test concludes, it takes time to analyze the results and derive actionable insights.
  

Total Timeframe: Typically, you can expect to start seeing initial learnings within 4-20 weeks (1-5 months) after initiating the process, depending on the scope and complexity of the tests.

Do I use MTA or MMM?

A spicy take to consider:

Can Affiliate & Influencer be measured in MMM?

Some say it is impossible since affiliate payments often occur after the action (click/lead / purchase) occurs.

I sat down with measurement expert Dominic Williamson, Director of Marketing Analytics, Chime.

The good news for affiliate and influencer marketing.

ROBYN, Meta’s open-source MMM allows you to select an input variable.

As long as you have a signal tied to the actual impact.

It is possible to align spend with when the key event happened, even if it was a commission-based event. This is perhaps the most challenging part, but it is doable.

Dominic and I ran holdout tests at eBay, and he found that there are times when affiliates can be one of the easiest to read channels in MMM.

So if you gather accurate impression data, click data, and spend allocation is allocated to the action, and have a key input variable like account creation or purchase, MMM should enable marketers to read affiliate and influencer marketing like any other marketing channel.

Now that you have read this, I hope the most misunderstood and underrated marketing lever will be understood and appreciated! 🙂 

Stop flying blind with your affiliate and influencer marketing.

This issue of the RBL Flywheel reveals why old-school attribution models like last-click can mislead and how smarter, data-driven approaches like MMM (Marketing/Media Mix Modeling) and incrementality testing help brands see what’s truly driving sales.

We debunked the theory that affiliates and influencers cannot work with MMM analysis, and we did it in style, thanks to our friend Al Pacino (Happy Birthday).

Here is to no longer being “in the dark here!” 🙂