Xyologic has, up till now, been best known for its app statistics service – but there’s a smell of revolution in the air. In line with today’s announcement of a Facebook app store, the startup has big plans to change the way people find mobile apps, moving away from categories and ‘Top 25s’ and towards search.
The sea change in app discovery is necessary, the three Xyologic founders told Silicon Allee, because the current standard is not good enough. And Zoe Adamovicz, Marcin Rudolf and Matthäus Krzykowski hope their approach, coming from data, will prove to be the answer.
Matthäus said: “None of the current ways how users discover apps really work – for the users, for the app publishers for and the platform providers. It’s not an accident that Apple bought Chomp, our biggest competitor. We are creating a much better way to find mobile apps.”
The team has spent the last two years going through a trial and error process involving statistical modelling, testing various machine learning algorithms and gathering tons of data. Marcin, Xyologic’s CTO, said: “We discovered a lot of information hidden in app stores during these two years. Now, however, we have figured out out how apps are intrinsically connected. We will be able to utilise this complex structure to make app discovery finally work.”
As Zoe was keen to point out, Xyologic is, at heart, a technology and big data company.
So what is so wrong with how app discovery works currently? Take the Apple App Store, for example. The buttons along the bottom represent the four main methods for users to find an app. They can look at apps featured by the platform provider, in this case Apple, they can view the most downloaded apps within categories or in the top 25 or they can perform a cold search.
Both ‘Categories’ and ‘Top 25’ are key, Matthäus said, as they have so far been at the core of how publishers market their apps: “If you talk to folks in advertising, say for example Admob, Quattro, Millenial, Inmobi or the local Madvertise, then app download advertising is up to 40 percent of their revenue.”
A second form of app marketing, so-called incentivised downloads, is apparently even more popular. This tactic sees publishers offer their apps in other apps and pay for downloads when players install their titles for virtual currency. A Berlin example of a provider of such services is SponsorPay, which focuses on Android.
All these tactics are focused on pushing apps into the leader boards of ‘Categories’ and the ‘Top 25.’
“This is what we are trying to change,” Zoe said. “What we see is that this system works for less and less apps. For example, hardly any new apps get into the top 25 anymore.”
In actual fact, 0.1 percent of apps are responsible for more than half of all downloads. Zoe added: “We believe that on Apple’s platform these days only a quarter of all apps ‘make it’ compared to a year ago. On Android the situation is slightly better, but is also moving in the wrong direction.”
The Internet succeeded because Google’s web search created a long tail, according to Marcin. “Such a long tail does not exist in the current app store system,” he added.
And so the trio firmly believe that the whole industry needs to move the app discovery experience beyond ‘categories’ and the ‘Top 25.’ Apple’s acquisition of Chomp will merely serve to reinforce the trend of moving towards app search, just like digital books and music mobile apps have become an important part of the digital lifestyle.
“Where we are really different is that we came to app search from data,” Marcin said, “so everything we do is based on hard evidence. During this time we have gathered information on how users search for apps, what queries they type; how they behave. Hence we are not only exploring how apps are related to each other or how to find a good app among hundreds of bad ones. With the technologies we develop we are able to determine an overall application’s quality, thus we are aiming to address the real app search behaviour we have seen over the past two years.”