mParticle FeaturesFebruary 13, 2015

Third-party segment insight with mParticle

Create more robust customer profiles by enriching them with third-party data to activate features like lookalike audiences, geolocation, and more.

The challenges

At mParticle, we’ve seen many of our clients begin to shift their focus from user acquisition to driving engagement. To optimize user engagement and retention with your app, there are two key challenges.

  • Determine which users are more likely to engage with your app so that you can create personalized experiences for them, e.g., personalized content, relevant push notifications, engage them across social media, etc.
  • Find more users like the most engaged ones to start using your app.

While analytics and marketing services tackle these problems from what users are doing inside an app, it is based on a single view of consumers, which can be somewhat limiting. While first-party data usually has the richest signal, the limited data set makes scalability difficult as effective tactics often fatigue quickly and models to increase scale generate false positives.

The solution

The foundation of success begins with a unified and holistic data approach. Our platform is designed to make this as easy as possible. Once your first-party app data is available in the mParticle data cloud, we enrich it with data from 3rd party data providers such as Datalogix to promote a deeper understanding of your app users. By appending 3rd party data to an app’s existing user data, mParticle allows app owners to discover:

  • Who their most engaged users are, including demographics, brand preferences, interests, hobbies, etc.
  • What data points are common among the most valuable users in order to improve user acquisition and retention strategies
  • Which brands app users are most loyal to in order to create sponsorship and cross-promotional opportunities
  • How to improve scalability and performance of desired KPIs

The “Third-Party Segment Insight” report can be found under the Reports tab in the mParticle platform and is automatically updated daily using the last 30 days of app data. Essentially, this report analyzes user behaviors within your app and breaks it down into third-party segments. Data filtering and smoothing techniques are used to make sure the results are statistically robust. We show three categories of performance stats in this report:

  • Overlap: This tells you the reach of a data segment within your current user base.
  • Engagement lift: We’ve picked 3 common KPIs that represent how users interact with your app in general: session per user, median session duration, and average revenue per user.
  • Goal event lift. Other than the 3 general KPIs we’ve selected, you can pick up to five custom events that you deem important, such as in-app purchases, social shares,  account creation, etc.

Each row in this tabular-form report is a third party data segment, and it shows how many app users this segment reaches as well as lift percentage on various KPI metrics. For example, the highlighted row in the screenshot shows that “Amex cardholders” segment reaches 1% of app users, and that users in this segment are likely to have 30% more app sessions, spend 190% more time within each session, and 223% more likely to perform an “app upgrade” event than average app users.

In this particular example, you can see segments that achieve good lift on median session duration metric may not have good lift on session per user metric. Therefore, as an app owner, it’s critical to identify a proper KPI that you’d like to optimize for and sort by that column to find good data segments.

One final note, if you find a data segment with good lift percentage but low overlap with your app users, you may write down an action item to consider adjusting your marketing strategy to attract more users of that data segment to your app!

third party segment insight

Final Words

There are thousands of third-party data segments available in the mParticle platform, which makes it incredibly easy to discover which types of users are the most and least engaged. As we integrate with more third-party data providers, we will provide more powerful insights on your app users.

 

AuthorYuan RenYuan Ren is a Data Scientist at mParticle.
Share

Get started with mParticle today

Connect with an mParticle expert to discuss how to integrate and orchestrate customer data the right way for your business.

Request a demo

Startups can now receive up to one year of complimentary access to mParticle. Learn more