Retention Rates for CPE Campaign Blow Away CPI Campaign, Rival Organic

< Back to Blog

Kevin Ford

March 25th, 2013

To determine what the quality of users is from various ad types, we set-up a test and measured the absolute retention rate* on Day 7 on a Cost per Install (CPI) campaign and Cost per Engagement (CPE) campaign on a segmented audience** run consecutively. The difference between a CPI campaign and a CPE campaign is that an advertiser pays when a user installs for CPI campaign while an advertiser only pays when the user installs and completes a specific action for CPE campaign.  For example in this CPE campaign we required the user to install and complete the testimonial—the action required in this test.



To prove out our hypothesis that CPE campaigns attract and retain users better than CPI campaigns, we ran a test on a popular, free-to-play Android role-playing game , PlayCoMo’s Little Dragons. The campaign was set-up half with a straight CPI campaign and the other half with a segmented CPE campaign where the user was required to install the game and complete the tutorial. The test geo-targeted users in the US, UK, and Canada. Prior to the test, organic traffic was the primary user acquisition source for this game, generated largely from a feature on Google Play.


Here are the results:

CPI:  12.7%

Healthy retention rate for RPG


Organic:  15.8%

Out-performed CPI by 24.4%


CPE Segmented:  19.2%

Out-performed CPI by 51.2% and Out-performed Organic by 21.5%


In short, Segmented CPE campaigns drive better quality users than CPI campaigns. It also shows CPE campaigns rival organic installs at attracting and retaining quality users.

To learn about measuring the health of your mobile game, check out this post that breaks down the key metrics for engagement, retention, and monetization. If you’ve already mastered measuring the health of your game, start attracting quality users with a CPE campaign.


*Absolute retention rate—a user is only counted if they initiated a session on a specific day.  For example, if a user does not play the game on Day 7, they are not counted.  In the other of type of retention, rolling retention, a user is counted for all the days from the first session through the current session.  So if a user logs in on Day 7 they are counted for Day1, Day 2, Day 3, Day 4, Day 5, Day 6 and Day 7 even though there is no opportunity to monetize the user on Day 2 through Day 6. To learn more, read “How Do I Know If I Have a Healthy Game.”

**Segmentation included the majority (60%) of the traffic.

Kevin Ford

Marketing Manager


Leave a comment

Artificial Intelligence and the Chinese Mobile App Market: Changes on the Horizon

Kevin Ford - September 1st, 2019

Artificial intelligence (AI) is no longer something to be feared. Instead, machines are going to change how life is in […]

Read More >

4 Things Mobile Developers Can Learn About the Chinese Mobile App Market

Kevin Ford - August 12th, 2019

American companies – all western brands to be exact – are having a hard time breaking into the Chinese mobile […]

Read More >

Important Changes in Chinese Mobile App Market Development

Kevin Ford - July 24th, 2019

Although mobile app markets are indeed a global phenomenon, China is leading the way with crucial changes in how they […]

Read More >