Havas Digital is mapping out a strategy to support content attribution in which data provides information on the type of message in an ad that prompts consumers to make a purchase or complete an action.
Messaging, creative ads and content have not caught up with technology and targeting, according to Michael Kaushansky, SVP of analytics and insights at Havas Digital. While technology can identify a consumer on a Web site visited milliseconds ago, and serve an ad, the ad will not add anything personal to the experience. "The industry needs to rethink ways to use data, from retroactive to proactive analytics," he said.
The industry has solved the issues of serving the ad at the correct time to the perfect customer, but not serving the right message. Proactive analytics focuses on defining actionable insights to provide a better experience for consumers.
Kaushansky said the model will roll out in pieces. It means analyzing the data to understand how to invest in media that prompts consumers to take action, not just bring awareness to the brand, then measuring the activity. Rather than using channels like display and search marketing, the strategy relies on content and the same data stored in data management platforms.
Kaushansky points to email marketing and how a brand might tie recurring purchases into making ads more personal. Someone buys fish oil from a vitamin store, which lasts about 30 days, and typically gets a personalized replenishment reminder about 21 days into the cycle.
Less than a handful of clients have moved to this model. One unnamed client experienced a 93% reduction in CPM by reducing the frequency and improving the targeting of the communication across the campaign. Performance improved by 15% over three months. "With clients more heavily leveraged in the real-time bidding space, such as a trading desk, you can probably get results in less than 15 days," Kaushansky said.
The thinking aims to address specific customer issues that improve the experience and not just provide data dashboards or attribution modeling. For the most part, attribution models are useless, Kaushansky said, because the data will tell marketers the lift, but won't make findings actionable.
Havas Digital will build out technology to automate parts of the process. The group has already automated content and search attribution and media planning daily insights, as well as constructing a real-time dashboard.