Data Trends Behind Roku Adding Media-Mix Modeling To Its Partner Program

Roku has expanded its Measurement Partner Program to include media-mix modeling (MMM) that can help marketers determine how various elements contribute to campaign goal.

The news -- announced Monday in advance of this year’s upfront with launch partners Analytic Partners, Ipsos, IRI, Mediahub, and Nielsen -- aims to create a better TV ad experience for marketers by making it easier for brands to analyze performance and measure outcomes. 

“We want to help advertisers with any type of measurement they want, and that’s something you’ll hear us talk a lot about at NewFronts,” says Asaf Davidov, director, head of ad measurement and research at Roku. “This is about us providing more granular data that’s available to help lean into the MMM models that CPGs and retail advertisers have leaned into during the past few years.”

Consumer product goods (CPG) and retail advertisers have asked for MMM. Davidov believes it has grown in importance because the measurement does not require one-to-one data targeting -- rather, it is broader channel measurement.



As data tracking becomes more complex, with many rules and regulations, advertisers are reverting to measurements that do not rely on cookies in browsers.

"That doesn't mean there are not other measurement solutions like multitouch attribution that offers one-to-one that should be used in more granular-level campaign measurement, but for a broad publisher-level measurement, MMM is become more important for CPG and retail," he says. 

Davidov says Roku will add additional granular data such as timestamp and dayparting, as well as content genres, that will correlate to many different data sets for MMM measurement.

Without any addition of data, it looks at impression volume and designated market area (DMA) at a very high level.

For example, it can measure, in aggregate, whether the advertiser sees an increase in sales based on the increase in media budgets spent in Des Moine, Iowa, or an increase in sales -- not correlated one to one, but in aggregate trends.

The additional data in volume, dayparting, content genre, and in some instances, ZIP code for location will allow marketers to gain a deeper understanding.

Last week, Roku announced the addition of cleanrooms, another trend that is driving MMM.

Marketers will have an option to gain regular reporting with more representative models, and provide a complete view of media investment tying ad exposure data to outcome data on a user or household level.

“When Roku users buy and register the product, we have about 60 million active households -- we get their email address, IP address, impression data, and content that includes genre, length of session, and dayparting,” he says. “Then passively we can understand broad viewership behavior because we can see what’s being served to a Roku device.”

In theory, Roku can determine that 10 million of the 60 million households are heavy news viewers, but because the company gathers the data in aggregate, it never analyzes just specific households.

Jennifer Pelino, executive vice president of Global Media Solutions for IRI, in a statement said that with IRI Marketing Mix Modeling, brands can continually track, measure and adjust their initiatives at the campaign level and ultimately decide the best way to allocate their advertising dollars for the months ahead.

Next story loading loading..