App Store searching and how it can be better
Every time you search for a new app on your iPhone or iPad’s App Store, what happens behind the scenes is that Apple takes your search terms, and then compares it with the titles and keywords of the apps found within the App Store. Relevant apps are then sorted through a secret algorithm that gives them a ranking based on their keyword similarity, popularity, and rating among other statistics. This usually works fine if you are searching for a specific app that you have in mind, since the search function ranks apps that have similar keywords and titles to your search especially highly. However, this becomes much more troublesome if you are trying to search for a certain category of app. For example, as seen in the MacWorld article linked below, the writer searched up the term “movies” in the App Store, and was returned a list of apps that had Instagram as the 7th most relevant app, Pinterest as the 10th, and Doodle Jump as the 14th. This is a major problem since these three apps have absolutely nothing to do with “movies”, and are therefore muddling up the search results. To make matters worse, 47% (based on a MobileDevHQ survey) of iPhone users found the last app they downloaded by searching for it on the App Store. By refining their search results, Apple can save users quite a bit of time by removing irrelevant results and improve the general user experience by suggesting more relevant, higher-quality apps through App Store searches.
One way in which Apple could improve their App Store search is by incorporating link analysis techniques in finding relevant apps. Currently, Apple’s algorithms to calculate search relevance seem to focus too much on the number of downloads an app has. While Instagram and Pinterest may have a lot of downloads and relatively high reviews, they shouldn’t have been ranked highly on the App Store when searching for “movies”. As we learned from lecture, the reason why search engines like Google are so good at finding what you want is because of the link analysis techniques that they use to rank search relevance. While the specifics differ between the various algorithms used, all modern search engines will rank the search results based on the number of times a specific website is linked to by other websites that contain the search query, weighted by the prominence of each of the sites that link it. This usually works great for us users, as it filters out hastily made websites that are not frequently referenced while allowing prominent, frequently referenced websites with useful information to make it to the top of searches.
In many ways, the App Store’s current search function seems to resemble those of the ancient search engines of yore before the Web 2.0 era. Just like the older search engines, by placing a lot of popular search-terms within the app’s title and description page, your app will instantly be ranked higher in searches and be suggested to more users more frequently. Perhaps in the future, if Apple does implement some sort of link analysis technique, Apple can achieve better search results in their App Store and remove the irrelevant results that seem to frequently appear. While this technique may be very costly for Apple to implement, since they do not have their own search engine, competitor Google already uses their own PageRank system to determine search ranking on the Play Store depending on the number of websites that link to an app. With diehard Apple fans, like the writer of the MacWorld article linked below, admitting that the Google Play Store is significantly better at finding relevant apps through searches, maybe it won’t be too long until Apple upgrades their App Store search system.
http://www.macworld.com/article/2151387/why-apple-needs-to-fix-app-store-search.html
http://www.forbes.com/sites/johnrampton/2015/10/16/the-beginners-guide-to-app-store-optimization/
https://moz.com/blog/app-store-seo-the-inbound-marketers-guide-to-mobile