The objective

One of Google's key Apps, despite having high awareness still had a fairly large base of infrequent users.

Traditionally, a blanket strategy (same treatment to every user) was being used to target even the fundamentally different users and hoping to drive the business metric. In the presence of rich data, sophisticated analytical capability and technology to differently target users, a more structured approach was required to bring in efficiency and greater impact.

Our objective was to develop a smart targeting strategy using a holistic audience segmentation statistical model to help increase daily active users for this particular Google App.

This segmentation would allow us to speak to and target different types of users differently using custom messaging and channels thus improving their engagement with the app.

The heart of our segmentation strategy was using the existing proprietary data we had from existing users of our the app. We had usage data of tens of millions of users and multiple touchpoints, translating into 350 Mn data points to be analyzed.