Enhancing Brand Performance: USIM's Application of Marketing Mix Modeling (MMM) in Upper Funnel Media Strategy

May 7, 2024

By Rob Jayson

Discover how USIM has used Marketing Mix Modeling (MMM) to demonstrate the true value of upper funnel awareness media for performance focused brands.

What is Marketing Mix Modeling (MMM)?

Marketing Mix Modeling (MMM), also referred to as Media Mix Modeling, is a statistical modeling technique that employs multivariate regression to construct predictive models. These models analyze media expenditures and impressions, along with non-marketing influences like economic conditions or other business-affecting factors, to gauge their contributions to outcomes like sales, traffic, customer acquisition, engagement, and branding effects. By evaluating these impacts, marketers can forecast potential results from varying levels of marketing investments or scenarios.  

MMM has been a staple in the toolkit of large brands for decades, primarily to decode the effects of substantial traditional media campaigns (TV, Print, Radio) on sales activities. It is recognized for its accuracy and effectiveness in marketing analytics.

This example output shows ‘Base’ sales or short-term sales not directly driven by media activity, and incremental sales that can be directly attributable to media activity from different media channels.  In this client case 85-87% of sales were ‘Base’ and 13-15% were incrementally added through media activity.

When to Use Marketing Mix Modeling

Marketing Mix Modeling is ideally employed when a brand seeks to identify and quantify the incremental effects of their marketing efforts across both digital and traditional media channels. Suited for complex media landscapes, MMM excels in managing marketing intricacies. It not only boosts marketing performance and efficiency but, with recent technological advancements in AI and high-speed computing, is also accessible to brands of all sizes needing sophisticated marketing analysis and optimization.

MMM Methodology

MMM utilizes a structured approach to evaluate and enhance the effectiveness of marketing initiatives on brand performance. We begin by systematically gathering historical data related to various marketing elements, such as advertising spend, pricing tactics, promotions, and product introductions. These data points are aligned with business results like sales or profits. A detailed statistical model is then developed using methods such as regression analysis or machine learning to identify connections between these factors, while also accounting for external influences. The next step involves calibrating the model using historical data, followed by rigorous testing to confirm its precision and dependability. The insights obtained from this validated model empower organizations to tailor their marketing approaches, optimize resource distribution, and improve their overall effectiveness and profit margins. By regularly updating and fine-tuning the model, businesses can respond promptly and effectively to evolving market dynamics.

How MMM demonstrates the benefits of upper funnel/awareness media activities

Upper funnel/awareness tactics are key aspects of the marketing funnel where potential customers are first introduced to a brand, setting the foundation for future engagements and conversions. Here's how MMM can illustrate the effectiveness of these upper funnel strategies:

1. Quantifying Long-term Impact: Upper funnel media, like TV ads or sponsored content, often do not lead to immediate conversions but are crucial for brand building. MMM can measure the delayed effects of these media investments by tracking how they influence consumer awareness and consideration over time, which eventually drives sales.
2. Attribution Across Touchpoints: By integrating data from various marketing channels, MMM provides a holistic view of the customer journey. It helps attribute the right value to upper funnel media, showing how these early customer interactions contribute to eventual outcomes, even when they are not the last touchpoint before a purchase.
3. Optimizing Spend Across Channels: MMM helps identify the optimal mix of marketing investments. It can demonstrate how reallocating budgets towards or away from upper funnel activities impacts overall business objectives, ensuring resources are spent where they generate the greatest return.
4. Scenario Analysis: MMM allows marketers to simulate different strategic scenarios, such as increasing or decreasing spend on upper funnel media. By predicting potential outcomes of these scenarios, businesses can make informed decisions that align with long-term growth objectives.
5. Isolating External Influences: Upper funnel activities are often influenced by external factors like economic trends or competitive actions. MMM helps isolate these effects to provide a clearer picture of how much upper funnel media itself contributes to brand metrics and business outcomes.

Case Study: How USIM used MMM to optimize upper/lower funnel media mix for a performance client

For a fast casual restaurant client who is fully focused on daily and weekly sales performance, measuring and quantifying the impact and sales effectiveness of video and other awareness driving tactics had been a challenge.  Conversion tracking for lower funnel media channels such as Search, Social and Retargeting Display attributed direct sales to those channels and they had increased their share of the media budget.

We had seen some decline in directly attributed performance and this suggested we needed to increase brand awareness and perceptions

USIM created an MMM of daily data for the 15 month period starting in January 2023.  Daily data MMM allows for us to use current data and focus on more recent brand innovations rather than rely on years old data.  The MMM model creates response curves for each media type showing the level of incremental sales revenue delivered for different spend levels.

These response curves are then incorporated into our scenario planning tools to enable us to forecast future incremental sales responses to different levels of media channel spending.

The MMM model then showed us the average return on ad spending (ROAS) for each channel across the period, clearly highlighting the vital contribution of upper funnel video media in delivering incremental sales revenue.

 This analysis was fundamental to the creation of media plans for the rest of 2024 and the incorporation of strong elements of upper funnel awareness media. Scenario planning demonstrated how incorporation of video media would deliver significantly higher ROAS levels than a 100% focus on lower funnel performance media only.