Many of our interactions with mobile devices are spent engaging with apps. In this post, we try to highlight some of the fundamentals that make measuring and understanding behavior on websites different from apps.
Web analytics is usually a means for collecting data and optimizing conversion rates. We’d usually ask how visitors reached our website, how many completed the goal (purchase, sign-up, content consumption, or other) and where in our conversion funnels users get stuck and abandon.
With mobile apps, the first conversion step is already completed before the first interaction with it. The user, who got introduced to the app (usually through an app store) has made a decision and downloaded it to his device. We’ve acquired a user, not just a visitor.
The focus now is less on a visitor’s flow through a funnel, and more on the engagement lifecycle, with the aim of maximizing three main behaviors: recurring usage (retention), spending money (monetization), and inviting friends (virality).
If a website can be seen as a store that people visit, an app can be seen as a room in someone’s house, waiting for you to come in again and again.
Website visitors are usually recognized by cookies. Cookies fail to identify people – there may be over-counting when visitors visit a website from multiple devices or browsers, and under-counting when several users from the same household use the same PC.
Counting users on mobile apps tends to be more accurate, since a user would usually use one primary device, and that device would usually be used solely by that user (although this is less true for tablets then it is for smartphones).
The current front-line challenge of analytics tools is to track users through multiple touch-points. Google Analytics is clearly going in this direction with its Universal Analytics.
When a user visits a website, the information of where he came from is usually passed along and easily tracked – be it a referral, a search result or an ad. However, with app downloads, the source is usually lost when going through the app store mechanism in order to download the app. Several tools (Flurry, AppsFlyer) are now making it possible to tie app downloads and in-app behavior to specific campaigns, by use of vanity URL’s. However, if the user found the app not via a campaign, but via web search, app store search or a referral, the source is hard to track.
While Google Analytics is the natural choice for tracking web traffic, it has been late to adapt to the mobile analytics mindset, and is still behind some of the tools that were designed for mobile only. These tend to provide a better understanding of mobile app lifecycle, with features like out-of-the-box retention / cohort analysis.