App Analytics: The Big Question of What to Track (appiterate.com)

We all know that when it comes to analysing our apps, data driven analysis beats hypothesis driven analysis any day.

But with so many metrics to choose from, how do you know which events to track?

Here’s a list of some key metrics you should track specifically.

For e-commerce apps

In e-commerce apps, the metrics you need to track should be aimed at maximizing the conversions.

1. Tracking Leads:   A user who reached the final page of purchase but hesitated and left is a lead. Such a user will revisit the app with the intention of purchasing on her/his mind. Keeping a tab of leads help in converting them, so that the next time when they visit, you have a better offer for them.
2. Drop off points:   When a user first opens the app, he goes through a journey that starts from the landing page and ends on the conversion page. In between, there are several pages that have to be visited. Tracking the page at which the users decide to leave will help you understand which stage of conversion process needs to be worked upon.  If they leave at an earlier stage, it usually means that they are not being offered enough choices. Dropping off at a later stage signifies that something is making the users hesitate before making the purchase.
3. Time Spent on a page:   This is a slightly tricky one. Less time spent on a page is generally a good sign. It means the UX is smooth. If the user spends slightly more time, it could mean two things; either the user is showing interest in the page and considering to buy or she/he is stuck with the UX and is confused about what to do next? If a very long duration is being spent on a page, the user has most certainly put the phone away and forgotten about the app. You should take a judgment call here on how to best interpret this data.
4. Switching frequency:   The switching frequency is calculated by first adding the number of switches between individual pages of the app in one session and then dividing it by the total duration of the session. The app’s switching frequency directly correlates with the efficiency of the UX.  A higher switching frequency means the user is having an easy experience with the app.

For gaming apps

A successful gaming app is one which is most engaging. So when you go about analysing the usage of a gaming app, you’d primarily want to track its engagement quotient.

1. Upgrade from free trial to paid version:   Most gaming apps operate on a freemium model, launching a free version first and then offering an option to upgrade to a paid version. It is here that the user actually converts. Never frustrate a user by repeated offer to upgrade. It might just backfire and make the user leave your app. Also, it’s vital to know at which point in time to give out the option, such that the user gets enough of the gaming experience to get hooked, but she/he still feels that there is a lot left to be discovered by shelling out a few bucks.
2. In-app Purchases:   The things your user can buy for a better gaming experience can actually account for most of the app’s revenues. The key here is to not make them seem like roadblocks to the gaming experience but an enhancement to it. That is, never make in-app purchases a must in order to continue but always keep the option around to lure the user. Specifically which features are the users buying? At which point in game are they buying? And how much time do they take in the gaming experience before making the first purchase? Tracking such events and analysing them will go a long way in ensuring that your revenues from in-app purchases show a healthy growth.
3. In-app ads:   If your app is doing well, other apps would want to bank on it. In-app advertisements of other apps will surely supplement your revenues, but it should never disturb your existing users. An annoying ad in midst of the gaming experience can cause a user to straight away close the app. Tracking how many users are returning to your app after seeing the first ad, how much time is spent on a page with the ad and how many of the users actually convert from the ad will let you achieve this goal.

For all apps

There are certain metrics you’ll always need to track, no matter which app you are analysing.

1. Device and OS being used:   What phone does a standard user own? If, for example, she/he is an android user then what device is the app being run on? The answers to these questions can give deep insights for your app. The same goes for iOS users. As an example, suppose yours is an e-commerce app and the metric shows that a major chunk of your users are i-phone owners. This can be a good indication as the majority of your users can be considered to have a higher spending power and so it would be wise to show them products accordingly. Factors like the screen size should guide the layout of your pages.
2. Demographic  Segmentation:   Knowing  the location of the users can be very helpful. Suppose users from Canada like bigger action buttons in the app, while those from Norway like them smaller. The best solution here would be to show different variations of our app to different segments of users.
3. Session Length:   Session length means both the time spent on individual pages and the total time spent in one session. If it’s a gaming app, a more lengthy session is always good news. But if it’s an e-commerce app, longer duration spent on one page may also signify that the user is stuck (granted there is no description to be read on that page).
4. Bounce Rate:   If a large chunk of the users closed the app without going beyond the landing page itself, you need to make some serious changes to the landing screen. Making the options simple and clear and highlighting the navigation buttons should reduce the bounce rate.
5. Retention:  Your retention number is key to making your app a long-term success. Retargeting  a user using push notifications and messages may convert him/her from a dormant user to an active user.

On a closing note I would remind you to never miss out on the quality data provided by user reviews in app stores as they form a direct medium of knowing what is on your users’ minds.