Analyst Watch: Ratings do count: Building world-class mobile apps
By Jeffrey Hammond
November 19, 2012 —
(Page 1 of 2)
Related Search Term(s): mobile apps
At last count, there were more than 500,000 mobile apps in the Apple App Store and more than 450,000 in the Google Play store. At the same time, the average U.S. smartphone user has an average of 41 apps installed on his or her smartphone. This means you'll need every possible advantage when competing for a coveted space on consumers' mobile devices, and high ratings will significantly help you.
It is simple math: Five-star apps fly off the shelves; single-star apps sit on the shelf.
For Forrester’s Mobile App Development Playbook, we asked 13 different mobile development shops to share their processes, tools and best practices. These shops all share a common trait: They all have at least one highly rated, successful mobile app in production. But that's not all: There's also a common recurring set of best practices these shops use to deliver a useful, usable, desirable mobile experience.
Real application success comes when you create a rapid, feedback-oriented development loop that does not scrimp on quality. Forrester’s playbook outlines the following best practices to minimize time to actionable feedback:
1. Assemble small, focused development teams: Mobile applications are called "apps" for a reason; they are generally smaller and less complex than traditional Java, .NET or Web applications. This means that you would need to plan for more numerous and smaller teams with two to six developers; employ specialized developers; hire design talent and invest in information architecture; cross-train developers for emergency maintenance; retire the QA center of excellence (for a quicker, specialized one); and refocus your sourcing strategies. Focus on skills rather than labor costs.
2. Favor simple development tools over complex application life-cycle management: Simplify and distribute software-configuration management, use visual designs, prototypes and collaboration instead of textual requirements; test apps with emulators and personal devices; and use mock tests and Back-End-as-a-Service frameworks to manage multi-layer complexity.