Sramana Mitra: Can you talk about the other side of the differentiation coin – the data side?
Asher Delug: For the first few years of mobile, the big complaint by advertisers and the reason why mobile ad rates have stayed so low compared to desktop rates is because until very recently, mobile has not had a very established data ecosystem.
On the web you have cookies, and they know I am a 32-year-old man and that I have x, y and z interests. But on mobile, they don’t know anything about me. That has created a scenario where advertisers pay less on mobile than on the web. How are companies approaching this problem? There are approximately 100 mobile ad networks on the market. However, there are only a few that have a very large scale data asset. Those data assets vary by company. We are one of the few with a very large data asset.
I will go through some players with interesting data assets. Facebook is an obvious one. They have very rich profiles about their users, and that translates to their mobile user base because users log in. The minute you log in to their app, they know everything about you. A company called Flurry has another interesting data asset. They put a free analytics platform for app developers into the market, and the tradeoff was that developers could use those analytic free, but they are basically letting Flurry collect data about all their users. With that strategy, Flurry has emerged as having one of the best data assets in the industry.
Jumptap has based their data strategy on going out and buying third party-data. They raised a huge amount of venture capital and are now going out and doing deals with Nielsen and Axiom, for example, to buy third party data and also buying match keys to match that data to mobile users.
In the case of Airpush, we consider our data strategy to be one of the top three in the industry. We created an opt-in SDK from day one. So users opt in. One of the things they opt in to is letting us collect a full list of apps on their phone – anonymously – so we can send them better apps. That has amounted to a quarter of a billion devices, which is around 30% to 40%, of the total number of Android devices. We know every app on the user’s phone. We know, for example, that Cameron has five RPG games on his phone. We know that Cameron loves RPG games. If Zynga has a new RPG game they want to promote, our system is going to promote it to Cameron. We are taking the stance that the apps on your phone say a lot about you. Because of our scale of 100,000 apps, we have been able to get that app data for a quarter of a billion users, and we believe that to be our primary differentiator on the data side.
SM: My observation is that there is one category of users that are really into apps. They have tons of apps, they are very active users of apps, and they a spend very large portion of their lives on their phones. Showing lots of app ads to them is going to make them download lots of things. Then there is a large category of users who use a few apps, but they are not voracious users of apps. What is the percentage distribution of this kind of psychographic behavior?
AD: One of the underpinnings of our data assets is to be able to find that small segment of users that are voracious – voracious overall, by category, and also as far as to who is buying paid apps. As far as your question goes, we are in the middle of putting together an infographic on this issue. But your statement is correct. A small percentage of users are “uber app users.” They are by far the most valuable.
SM: And the percentage of “uber paid app users” is probably even smaller.
AD: Exactly. Those represent the most valued user segment for an advertiser.
SM: And you are saying that with the data you have on these 100,000 apps and their users, you are able to separate those kinds of users into buckets and are able to push ads to them.
AD: That is correct.