Fiksu: Apple incorporating user ratings into App Store Top Charts rankings

Tools

Apple (NASDAQ:AAPL) is factoring iOS user ratings into its App Store Top Charts rankings, according to mobile user acquisition platform Fiksu, which closely monitors App Store activity.

"After some in-depth (and still ongoing) analysis by our Performance Labs team, we can conclude that Apple is testing changes to the iTunes ranking algorithm," writes Fiksu Senior Client Development Manager Tom Cummings on the firm's blog. "What we don't yet know is the entirety and permanence of these changes."  

According to Cummings, App Store download volume and velocity previously determined iOS app rankings. "In order to rise up the ranks, your app needed lots of downloads in a short amount of time," he explains. "And to maintain that rank, you needed to continue acquiring new downloads." But last month, Fiksu identified a number of apps unexpectedly shifting ranking position without a corresponding change in download activity. After further investigation, Fiksu connected the position changes directly to App Store ratings.  

"Apps with ratings of 4 or more stars received a rank boost in late July and have maintained that rank throughout August," Cummings states. "Poorly rated apps with less than 3 stars received the opposite treatment about two weeks later, dropping off precipitously in rank and staying that way. Average rated apps, between 3 and 4 stars, suffered a slight degradation in rank but have generally been less affected." Fiksu has not yet determined whether historical rankings matter, but notes some evidence that ratings for the newest version of an app have the greatest influence.

Fiksu adds that App Store positioning now updates every three hours instead of every 15 minutes. "It's worth noting that this is only the public face of the App Store: it's not yet clear if the algorithm is updating constantly behind the scenes, or if rank data available via the App Store API is updated on a different schedule," Cummings said. "We suspect [Apple is] adding a 'buffer' period to prevent rank manipulation through short download bursts often associated with robotic downloads and other actions not in the best interests of the app ecosystem. Three hours allows enough time to identify anomalies and remove apps before they suddenly appear at the top of the rankings."

Apple did not respond to a request for comment.

Fiksu notes that rumors of ratings-influenced App Store rankings date back to early 2012, when Apple acquired mobile application discovery solutions provider Chomp for a reported $50 million. Chomp leveraged a proprietary algorithm that learned the functions and topics of apps to enable users to conduct store search queries based on app function and utility, not brand name.

"We've seen no evidence, but there have been longstanding but unconfirmed rumors that other factors, including social reach and uninstall rates, could influence rank," Cummings adds. "As such, the currently observed changes could be preludes to larger changes that coincide with the release of iOS 7, but they may also be the extent of new changes." Fiksu recommends that app marketers continue to emphasize download volume and velocity, move swiftly to patch any software bugs and iterate as quickly as possible.  

The Fiksu platform is designed to help brands more cost-effectively market their iOS and Android applications, leveraging predictive algorithms to optimize advertising campaigns in real time. The firm has accumulated more than 131 billion app actions like launches, registrations, in-app purchases and real-time bidding requests.

For more:
- read this Fiksu Blog post

Related articles:
Fiksu: App marketing costs climb as brands embrace Facebook install ads
Facebook install ads galvanizing iOS app download growth, Fiksu says
Fiksu: iOS app downloads, marketing costs hold steady in March
Fiksu: iOS app downloads plummet after holiday season surge
Apple dodges App Store antitrust complaint--for now
Distimo: Top-ranking free iOS apps generate at least 72,000 downloads per day

 

Comments