The AppsFlyer Performance Index - MWC 2016 Special Edition - AppsFlyer

The AppsFlyer Performance Index – MWC 2016 Special Edition

Being a successful digital marketer in 2016 is hard. Being a successful mobile marketer is even harder. The space is only several years old, and comparative information about the different players in the industry ranges from scarce to non-existent. With competition in the mobile app ecosystem exploding, understanding where to invest your marketing resources is paramount to success.

As the leading, unbiased mobile attribution and marketing analytics platform, AppsFlyer has a bird’s eye view of the entire ecosystem. We are therefore in a position to offer mobile marketers a credible and in-depth report card on the best performing media sources so you can make smart, data-driven decisions on your ad spend.

The AppsFlyer Performance Index MWC Special Edition celebrates the world’s biggest mobile event of the year. It covers over 2.5 Billion installs, ranking the top media sources in mobile advertising and displaying them in interactive charts, broken down by:

  • Categories: Gaming apps and non-gaming apps
  • Regions: Global, North America, APAC, Europe and Latin America
  • Operating Systems: iOS and Android

Click on the following links to jump to the relevant section:


Executive Summary
  1. Facebook is the undisputed leader in mobile advertising. A consistent placing in the global top 5 media sources for retention, coupled with its unrivaled scale, put the social giant in the #1 spot in the power rankings for every category and region among platforms included in the study.

  2. Among ad formats for app installs, video advertising was dominant in retention across the board, as video ad networks had a roughly 30-90% higher retention rate than other networks in the index.  This was especially true for iOS, which commands the high end of this range. With video ads for app installs, users are exposed to more of the gameplay / app experience and have better knowledge of what they are getting into as compared to non-video formats.

  3. Social networks were top performers in retention, especially on Android where they were in the top 3 in both gaming and non-gaming (travel, ecommerce, utilities, lifestyle, content, transportation, etc.) categories.

  4. Google has demonstrated that the power of intent delivers quality users at scale across the board.

  5. Twitter was one of the best sources of retention in Android, coming in #1 and #2 in gaming and non-gaming, respectively. However, when it comes to scale, Twitter did not fare as well, leading to a #18 and #10 ranking in gaming and non-gaming, respectively. This is likely to change once the integration with MoPub is complete and running smoothly, which will effectively extend the social network’s reach well beyond Twitter properties.

  6. The differences in retention across the top 30 media sources in iOS was much higher than the differences in Android, where the playing field was more even. However, when looking at the top 10, the gaps were smaller in both platforms, ranging between 4-10% (based on standard deviation calculation).

  7. Running on a diverse range of phones, including many on the lower cost spectrum, Android dominates the Asian market. This explains why Asian networks made up about a third of all networks in the power rankings in both gaming and non-gaming categories.



The index is based on a robust statistical comparison between organic and non-organic retention of the same apps.


  • Date range: Oct. 15, 2015 – Jan. 15, 2016
  • Installs measured: 2.5 Billion (organic & non-organic)
  • Number of apps: 5,000+ with at least 1,000 installs per OS per region throughout the date range
  • App opens: 3 billion (day 1, day 7 and day 30; excluding day 0 launches)
  • Number of media sources included in the index: Over 1,000 of AppsFlyer’s integrated networks

Indexed Retention Score

  • Step 1: We calculated post-install retention to determine the percentage of unique users who were active on days 1, 7 and 30, out of the total number of unique users who first launched the app in the selected time frame.
  • Step 2: We then compared each media source based on its performance relative to the organic retention of the same apps that ran with the network in question. Each comparison was performed on a regional basis (APAC, North America, Europe, Latin America and Global), and per days 1, 7, and 30 post install.
  • Step 3: We assigned each day a different weight in the formula as follows: day 1 was given a 20% weight, day 7 a 30% figure and day 30 a 50% value.
  • Step 4: We indexed the retention score figures, displaying their rank and enabling a relative comparison between them.

Only media sources that ran campaigns with at least 10 apps per platform per region were included (separately for gaming and non-gaming). Since the Latin American market is still in its growth stage and in some cases has yet to scale, the number of ranked networks is lower because others did not make the cut.

Power Ranking

This rank combined two factors: both the retention score and the total number of installs generated by each network (per region, per OS, per category).


Part I: Top Gaming Media Sources

The following interactive frame includes a power ranking and an indexed retention score broken down by platform and region. In addition, key findings are added to the global indexes only.



Part II: Top Non-Gaming Media Sources

The following interactive frame includes a power ranking and an indexed retention score broken down by platform and region. In addition, key findings are added to the global indexes only.


A final note

The above figures are aggregated across a massive scope of data and as such are highly credible. However, it’s important to remember that each app is a world of its own. That means that some sources that performed well in our index may not perform well for you, and vice versa. After all, there are hundreds or even thousands of media sources out there.

Therefore, we strongly recommend that you test as many different sources as possible – even ones that did not make it into the index because they are perhaps more niche – until you find your best mix.

Get this study, FREE