Mobile ad networks: in the land of the blind, the one-eyed man is king

Published on October 12, 2012 by Eric Benjamin Seufert

The mobile advertising landscape is caught in a strange, dynamic, transitionary stage. Information — about everything: mobile ad networks, platforms, advertising identifiers — becomes irrelevant so quickly that the only true expert on mobile acquisition is the person launching an app right now. No one else has accurate information about the state of the market; the tectonics simply shift too rapidly.

The dearth of actionable information about ad networks presents a challenge to the user acquisition professional; how does one prepare for a paid acquisition campaign if he doesn’t know anything about the state of the marketplace? I believe this question is answered by accepting as true a few fundamental conditions of mobile advertising:

  1. All CPA traffic is the same. Some networks employ sophisticated targeting strategies to try to pair your app with the users most likely to monetize, but I’m not convinced that these methods provide value. Mobile ad networks are serving their impressions to the same, limited group of app users (and those users are adversely selected for in the first place). If you do a search on Quora about a specific ad network, you’ll find just as many people claiming it is terrible as claim it is great. The truth of the matter is that no ad network can deliver the “best” users on a consistent basis because they’re all competing for the same people.
  2. All incentivized traffic is the same — and it’s bad. Incentivized installs are cheap, they don’t retain or monetize, but they’re available in very high volumes.
  3. The network from which a user is obtained has less of an effect on monetization than how that user is treated in-app. Upstream marketing is limited; downstream marketing is infinitely effective. Micro-managing paid acquisition campaigns yields less than strong, in-app predictive techniques that improve the user experience. Data science trumps Excel wizardry.
  4. Mobile app tracking is terrible. Tracking every user that enters an app is impossible with the tools available on the market. Large developers can try to improve upon the accuracy of those tools (which tops out at around 80%), but I think these incremental improvements aren’t worth the effort for all but the very largest developers. I don’t know why mobile app tracking is in such an appalling state, but my best guess is that the companies best positioned to fix it benefit from it being broken.

The fundamentals of user acquisition have more impact on the success of a campaign (and an app) than conditions which change on a daily basis. The only “expert” on ad network prices, fill rates, quality, etc. is the user acquisition manager in the middle of an app launch.