This article covers one of the core issues in the Android application ecosystem, application discovery.
The exponential boom of smartphones makes you realise that you have entered the smartphone era. Eighty per cent of the world’s population owns a mobile phone, and of the approximately 5 billion mobile phones in the world, 1.08 billion are smartphones around 20 per cent! Currently, smartphone purchases are outnumbering normal mobile phone purchases, day by day.
With a smartphone, the inevitable need is smart applications. If we go by the numbers, 74 per cent of data usage activity on Android and 79 per cent of that on iOS is related with mobile application downloads. A recent news release stated that Google Play has over 700,000 Android applications. If we believe predictions made by experts, mobile apps will grow from a US$ 6 billion industry today to US$ 55.7 billion by 2015. Users log an average of 77 minutes per day using apps on their smartphones, and hence there are a large number of mobile applications being launched every month. That’s fine, but why am I stating all these numbers?
There are a lot of Android application distribution platforms, Android application discovery platforms and a lot of third-party App stores that claim they can help you in discovering the Android application you may need. But they are still far away from doing so. As I have already mentioned, Google Play has already crossed 700,000 apps; which is very encouraging for app developers, considering that just 1 billion or so people have smartphones. But the app stores and discovery platforms we have in place are not well suited to showcase such a large number of apps. The current system of having five to 20 top applications in various categories such as Top Paid, Top Free, Top Grossing, Staff Picks, Recommended For You, Trending, etc, leaves room for only a few thousand applications to be discovered. If we closely analyse Google Play and other third-party Android app stores, they are not perfect mobile app discovery platforms yet. In the current system, developers upload app screenshots and app videos to showcase their applications on the app market. Based on the visual appeal and the description, products are sold. Now this certainly is not the perfect way of buying new apps. I mean, if you ask me to explain the current app store policy in simple terms, I would say they are selling us clothes based on their description, pictures, colours, designer, brandand we can’t try them on unless we buy them. Does that make sense? As a substitute to this, Google Play provides a window of 15 minutes for users to try the app and cancel the purchase if required. But this includes downloading the app, which is time-consuming.
Another issue with the current mobile app ecosystem is converting Web leads into actual downloads. On an average, 160 minutes/day are spent on the Internet using laptops and desktops, which is approximately 150 per cent of the mobile Internet usage. But the current app store system fails to use this platform efficiently. As explained in [image 2], app sharing on the Web is broken. We can see this if we compare the app-sharing lifecycle to the YouTube video-sharing lifecycle. How many of us actually go on YouTube searching for a video? To go by stats, 500 years of YouTube video are watched every day on Facebook, and over 700 YouTube videos are shared on Twitter each minute. A 100 million people take a social action on YouTube (such as likes, share, comment, etc) every week, out of a total 200 million users. So almost50 per cent of the traffic is via social networking sites. Imagine going on YouTube and searching for the Korean popstar Psy’s, Gangnam’s video based on a tweet or post saying, “hilarious dance steps guys, please go and download this video!” We are still lagging behind in the social recommendation model for mobile applications. Link sharing is similar to videos, but then actual consumption of content is broken until we download it. In other words, Web links fail to convince users effectively.
But there are quite a few products that try to solve this problem of app discovery; I have mentioned a few of these in Table 1. They add their own algorithms in addition to the Google Play algorithm, with extended social sharing features. That makes app discovery a bit hassle-free, and they do some kind of personalisation to the suggestions and ranking algorithm they use while showing the results to you. But the core problem remains unresolved users have to trust the content and static images.
|App.net||Static custom pages for apps
Download via SMS and email
Better social sharing
|AppHero||Good social recommendation
Personalised app recommendation
Personalised app recommendation
Good app search
|BuzzDoes||Social and personal recommendation
Gives incentives to the users
Developer needs to integrate their SDK
The Amazon app store solves the problem of app discovery with an additional feature to its ranking algorithm, called Test Drive. This enables apps to be tried on the Amazon app store Web page and mobile app itself. It allows users to consume apps without having to download them onto their device. So it doesnt matter which way users are browsing the market they will be able to try the app and if they like it, they can download it. This makes app discovery complete. However, this Test Drive feature is available only for apps hosted on the Amazon app store, and only in the United States; and usable only in Web browsers with Flash support. Moreover, they lack social features, as it is region-specific.
In spite of all these products in the market, app discovery still remains an unsolved mystery. While concluding, I would love to mention that we at AppSurfer try to overcome these limitations. We are trying to combine the remedies to the app discovery issue, so as to make the Android application ecosystem more mature and complete. We run Android applications in the browser, so users can try apps before they download them. We have social recommendations in place, as well as widgets that make your Android applications run on blogs, on Web pages and in the Facebook feed itself.
All in all, app discovery is a well-known problem, but I feel that if tackled rightly, it is solvable!