3 Data Tips for Marketing Mix Modeling Effectiveness

March 28, 2019 Ginna Hall

This article originally appeared in MarTech Advisor.

Businesses across industries are challenged with determining which marketing efforts are actually driving profit and which are just wasting spend. Marketing Mix Modeling (MMM) is a marketing measurement approach that can address this challenge.

MMM calculates the total effect that every marketing channel and its key dimensions, such as creative, product, and geography, have on sales and other performance metrics, while controlling for exogenous factors like weather and holidays that impact business performance.

MMM is critical for marketers who want to optimize ROI holistically across all business drivers – online and offline. The insights it produces help CMOs and marketing leaders plan both short and long term, determine how to best allocate their budgets, compare year-over-year metrics, and better understand non-direct performance measures such as brand affinity.

This measurement method can also provide recommendations for how to optimize future marketing activities in order to drive the best results.

Despite the benefits, not all MMM models are created equal. There are three critical factors that can affect the overall success of an MMM analysis, and they all lie in the data collection.

3 Critical Factors for MMM Success

Let’s take a closer look at three data inputs that can greatly influence the accuracy and actionability of MMM results:

1. Manual vs. Automated Data Collection

Marketing mix modeling requires gathering, consolidating and normalizing data from a variety of sources into a standard format. The data should provide a comprehensive look across all of a companies’ activities, including media data from TV, Print, and Digital, as well as point-of-sale data, such as pricing promotion, and distribution.

Often times, this data collection is done manually – a process that is tedious, time-consuming, and error prone.

Automating data collection can eliminate the risks of manual data manipulation while ensuring that information is collected and normalized in a format that preserves its integrity. Some marketing measurement vendors have direct access to data and/or partnerships with third-party sources to ensure data is collected in the right format.

An automated approach not only improves speed to insight, but also ensures a large amount of data can be distilled into quality analytics that marketers can use to improve campaign performance and spending.

2. Questionable vs. Validated Data

When it comes to MMM, the key to having high-quality output is having high-quality input. As the saying goes, “garbage in, garbage out.” Without quality data, it’s not possible to carry out a sufficiently accurate statistical analysis.

Therefore, it’s critical to review and validate the data for accuracy as its being collected. For instance, if there is an anomaly in the data – such as a big spike in activity – the inputs should be reviewed to ensure what is collected reflects reality.

Data errors can lead to poor outputs, so calling out any questionable data at the start of an MMM analysis is essential for getting the best results.

3. Aggregate vs. Granular Data Inputs

Modeling at the correct level of granularity maximizes model accuracy and actionability. For instance, incorporating data at the store/branch level, segment level, zip-code level, product level, and/or DMA level will enhance the precision and usability of the results over using aggregate data at a larger geographic or national market level.

Most often using highly aggregated data results in an underestimation of the true impact of marketing efforts, which leads to incorrect decisions and wasted resources.

Granular data leads to granular insights. Having information about specific campaigns, tactics, products, retailers, or geographic area enables marketers to take more specific actions to improve results.

For example, if you can’t understand the impact of TV by dimensions such as campaign, DMA, spot length, or daypart, then you can’t tease out what’s really driving sales or identify where improvements should be made.

Maximizing marketing effectiveness requires specificity. The more comprehensive a marketing mix is across potential drivers of volume, the more accurately it will attribute impact to the right touchpoints.

Marketers need a comprehensive and insightful understanding of their marketing efforts – not compromised measurement that lacks actionability, granularity and accuracy. Marketers in every vertical have adopted a marketing mix modeling approach to measurement because it serves their businesses’ operating cadence and supports strategic planning. Quality data inputs will guarantee that marketers have the best measurements to drive the greatest return on their investments.

Learn More

Download our ebook to cut through the hype and learn the differences between key measurement approaches: Untangling Attribution’s Web of Confusion: A Primer for Marketers

Request a Demo

Learn how Nielsen can help you measure and optimize your marketing and advertising: Request a demo today.

Subscribe to our Blog

Get news and information in your inbox every month: Subscribe

Previous Article
7 Lessons for Digital Marketers in Korea
7 Lessons for Digital Marketers in Korea

Korea is one of the most attractive and potentially lucrative markets in Asia. Get 7 lessons for digital ma...

Next Article
How to Lead a Successful Multi-Touch Attribution Implementation
How to Lead a Successful Multi-Touch Attribution Implementation

Implementing multi-touch attribution can seem like a daunting task. Read about Nielsen's recommended best p...



Error - something went wrong!