Everybody carries a smartphone. Where are they going? Marketers want to know. They figured out that a mobile user tipped off their next buy whenever they searched on a specific term. So send them the ad.
YouAppi, the San Francisco-based mobile marketing solution, no longer takes that approach. The company recently updated its Matching Algorithm, linking it to three years’ worth of user data, to get a better idea of who the
consumer is and what they want.
“Ad serving is dated,” said Chantelle White, chief operating officer at YouAppi. “We offer intent-based marketing.” Gauging intent requires more than one user interaction. When digital marketing was in its infancy, that single interaction was sufficient to put a user profile into a single, shallow cohort. For example a woman in New York who likes green shoes, said White.
That profile takes on a different meaning when you find out what else that woman in New York is doing on her hand held device. She still likes green shoes, but also checks real estate listings on Zillow during her morning commute, and plays Candy Crush on the same device during the evening commute home. So is that sales opportunity just limited to green shoes?
The cookie-cutter approach towards predictive analysis turns out to be wrong if one assumes that what a consumer bought before is what they will buy next time. “All of our assumptions were proven wrong,” White explained. “The rule of thumb for us is, let’s not allow our assumptions dictate where predictive will go. Let’s allow the data to speak to us before we title it and stick it in a box.” She added, “Let’s watch what happens.”
Specifically, that means watching behavior, not past purchases, and trying to spot when the same user behavior pops up in different marketing cohorts. Every time that same behavior surfaces, a sales opportunity beckons, but in a different place. For the Matching Algorithm to do this, it taps a deep database of 1.5 billion mobile users, 15,000 past ad campaigns, and 3,500 different mobile apps, all gathered from previous interactions with YouAppi-related clients. Machine learning under the hood sharpens insights as the database grows over time. The result can be outcomes no one expected. “Everything opens up,” White noted.
The Matching Algorithm, along with machine learning, are both wrapped together in YouAppi’s OneRun platform, designed to deliver insights for mobile media buying. The entire package can sift through about 250 terabytes of data daily, searching for those clients that could be most receptive to the right targeted ad.
When YouAppi started business in 2012, this concept was in its infancy. “Now it is in its adolescent stage,” White said. “In the next year or two, the power of the algorithm will continue to expand.”
Once the database reaches global proportions, machine learning and first-party data should yield an understanding of the market when the same piece of buyer data emerges in different marketing cohorts. Tracking mobile users and pitching them tailored ads will happen in real time.
“There will be no room for error,” White said.
