The effectiveness of advertising formats and networks in 2017: an analysis of 69 billion advertising hits

Appodeal released a report on the effectiveness of advertising embedded in mobile apps: from January to November 2017, we analyzed 30,000 applications, 69 billion impressions and 60 sources of advertising demand and five regions of the world. Performance Index will help developers understand which advertising formats and networks should be targeted next year.

According to the analysis, the most popular formats in mobile games and applications for 2017 were rich interstitials (non-permissive ads, which are shown for a few seconds before disappearing), as well as rewarded video. A new variation of rich interstitials - playable ads, - which allows users to play a mini-game when clicking on ads, increases the level of involvement and also claims a fairly steady trend.

In addition, it turned out that among the rewarded video, the most growing was the format when the user receives an award in the form of virtual content in a game or application. The eCPM (effective cost per mille) of this format has shown impressive growth. Facebook Audience Network and AdMob provide the most expensive advertising in this format.
')


In addition, banner advertising is still afloat. Despite the fact that this format appeared before the others and has one of the lowest eCPM figures, banners are still in second place in terms of revenue among all regions.

As a result of 2017, in all regions, AdMob showed the highest results in revenue and advertising fill rate (fill impressions) on banners and rich interstitial ads. In addition, Inyractive Marketplace from Fyber stood out in banner advertising and rich interstitial ads formats. In the US, East Asia and Western Europe, Amazon Mobile Ads showed the best banner results.



AppLovin remains the top advertising network for rewarded video and video interstitial ads. Also, the main networks for video formats are Tapjoy, Vungle, AdColony and Unity Ads.

Download the free report by the link .