Marketing mix modeling (MMM) is having a resurgence due to increasing privacy regulations, a fast-approaching post-cookie world, and marketers struggling to make sense of this complex ecosystem, said Chris Penn, co-founder and chief data scientist at Trust Insights.
“MMM allows marketers to export day-level data from across channels and use it to find which variables are moving the needle,” helping them get a clearer idea of how their marketing efforts are improving the bottom line, Penn said.
Here are three ways marketers can experiment with MMM.
Do you have the data? “That’s step one, making sure your marketing activity data is in a format that you can use,” Penn said. “And that ideally should be day-level data, because you want more granularity so that you can build a more robust model more quickly.”
It’s also important to identify a metric of success not tied to sales performance. “If you have an ineffective sales team, your marketing model isn’t to blame for not making sales,” he said.
Instead, find a metric that you know marketing has a direct effect on and use that to define success, like brand awareness or audience engagement.
Marketers should pair MMM with attribution modeling to get a full view of the customer journey.
“Marketing mix modeling is not the new answer,” said Penn. “It’s just one part of the analytics puzzle that helps you see the big picture.”
MMM offers a top-down view of marketing activities and how they contribute to desired outcomes. Attribution modeling offers a bottom-up view, focusing on the touchpoints an individual consumer engages with along the customer journey.
“You need both to be able to answer what happened and why,” said Penn. “And obviously attribution modeling is getting harder because you have less consumer data to work with, but there is still invaluable insight that can be gathered from things like focus groups, customer advisory boards, [and] one-on-one interviews.”
TikTok recently shared insights from its partnership with Nielsen to help brands with their MMM efforts. And the results were promising:
Penn, however, advised marketers to take this data with a grain of salt. “It’s always been the case with ad tech vendors—whatever models they use will inevitably favor them,” he said.
Rather than partnering with a vendor that could omit important data from their model to make it more favorable to them, Penn recommends marketers invest in creating a proprietary MMM strategy.
And while that may take some investment upfront, it’s worth it for a more accurate look at which marketing initiatives are working and which are not.
This was originally featured in the eMarketer Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.