Reputation Systems for Mobile App Storefronts

One of the primary draws to mobile devices such as smartphones and tablets is the ability to download and install apps on the device. The paradigm of an application storefront began in early versions of Apple’s iOS, and has since been adopted by all smart device operating systems. Now what interests me the most about these storefronts is how people use them. Specifically, how do people make decisions about which apps they should download? In many cases, users know precisely which apps they want, because of word of mouth or experience using other devices. However, sometimes users will have a specific goal, but won’t know which application will best accomplish it for them. In these instances, they will rely on the storefront’s reputation system to help them filter out the good from the bad.

In investigating the storefront reputation systems employed in three major mobile operating systems (Apple’s App Store, Google’s Market, and RIM’s App World), I noticed two distinct user groups with different requirements from the system. These groups are the end users of the app and the app’s developers. The end user’s primary objective is to use the reputation system to make an educated decision on which app will best satisfy their needs. The developers’ objective, on the other hand, is to get the best rating possible, to get their application surfaced (e.g., in a ‘featured apps’ section), and ultimately to increase the revenue generated by the app.

How can a storefront reputation system reconcile the needs of the end user and those of the developer? I reviewed some forums to get an idea of the changes people were requesting from the various storefronts, both users and developers, and it seems that there might be a common ground. Users want to be able to sort app ratings, so that they receive only the most relevant information. One way to achieve this is to allow sorting by application version, so that users can filter out only the ratings pertinent to the latest version. This is also desirable for developers, as the filtered rating will not be negatively impacted by bugs or design flaws from previous versions. That is, improvements in the current version will be reflected in the new ratings.

Similarly, end users and developers appreciate a reputation system that allows the user to rate other users’ ratings. For example, users can flag reviews as “helpful”, “unhelpful” or “spam”, and filter them as such. This way, users can filter out unhelpful reviews that are not specific to the design or function of the app, such as complaints about pricing or compatibility with specific devices.

Finally, many developers want the ability to respond to negative reviews, much like the buyer/seller feedback system on eBay. This gives the developers an opportunity to defend themselves in the midst of unwarranted criticism or trolling. It can also be beneficial for those reading the review, who may question the validity of the reviewer’s comments. A developer response system can also convey to users that the developers are engaged and listening to their feedback.

While some storefronts have already acknowledged and implemented some of the above recommendations, there is always room for improvement. Given the constant flux and innovation in the mobile space, I look forward to seeing improvements in future application storefronts and how they adapt to their users’ needs.