Have you ever caught yourself looking at app analytics from a very web-centered perspective? In this case, the following very handy guide will give you the lowdown on differences between web and app analytics.
Imagine you’re just about to launch an app, exciting times! Design, app development and marketing strategy: check. App Analytics strategy and relevant metrics in place: sure thing. But wait a minute. Do I have to forget everything I know about web analytics for my app marketing strategy? No. Well, maybe a little ;-). What we oftentimes see with clients is that they still have a very web centered view when it comes to their app analytics. Spoiler alert: by the end you will see why you should use a true app analytics tool.
Mobile App Analytics, here we go!
Now, as an app marketer you will probably start out asking yourself questions such as: what is my business goal, my use case and how should users behave within my app? These questions are relevant in order to pick relevant KPIs. And measure how your users behave.
Analytics newbies should be aware that the analytics game is very different with respect to app vs. web. Familiar metrics and mechanisms from your web analytics experience might not get you far at all. We are just using mobile apps completely different than the web.
To make matters even more complicated apps run in their own sandboxed environment. That makes sharing and collaborating with other installed apps is practically impossible. Hence you are dependent on using device-specific advertising identifiers (IDFA & GPS ADID) as provided by Apple & Google. These in turn can be accessed only in mobile apps, but not in the browser. Otherwise you could conveniently implement a matching with web cookies.
That’s why so-called Mobile Measurement Partners (MMPs) such as Adjust or AppsFlyer are rubbing their hands these days. These analytics and tracking tools are booming. MMPs are bridging the gap between web and mobile app by allowing fingerprinting methods to map app installs to clicks on a web browser ad. This is the only reliable way to record the source of your app installs and attribute them accordingly.
Updates in a fragmented app world
Rolling out new updates frequently is very common for app development teams. Website relaunches or updates leave users usually with essentially the same experience. However as an app user your experience will heavily depend on the app version you are using.
So as an app marketer you are faced with the challenge that users have to update the app on their smartphone first to enjoy the new experience. But until that happens they are still experiencing the old version of your product.
Keep in mind that you might be analysing mobile data from users that are potentially having varied user experiences due to a different look and feel of your app. Throwback to the much discussed Snapchat update in early 2018:
We encourage you to always keep this his fragmentation of your user base across different app versions in mind. Without handy resources such as the tool
Apptimize it will not be possible to roll out an update for all of your app users instantly. Any regular update through the store can take several weeks depending on whether users have activated automatic updates in the store.
You might risk losing users who get stuck on older versions of your app. A fragmentation nightmare in your data analysis will happen when you e.g. change the naming convention for events between different app versions. Our tip: Keep a close account of when you rolled out what kind of changes to the analytics and with which app version.
Pageviews vs. Screenvies: How active are your users
Small but subtle difference: Analyzing the interaction of users with your app’s content happens via screen views. Whereas you look at page views when analyzing your website’s activity. Truth be told, in the end you want to understand one and the same thing: user behavior.
The user interacts with an array of screens during a session. Hence this is a good indicator for whether or not you can catch their attention with your screens and content. How long do they stay on a screen? How do they move from screen to screen? And which screen leads to a purchase? Analytics tools such as Appsee serve very nicely for data collection and visualization. And they also make you aware of potential conversion blockers.
What is a Session? And how long is a session length
If there is one topic of discussion (that takes on quite some missionary characteristics) it’s the definition of an app’s session. Adding the differences between web and mobile app you can easily lose yourself in this discussion.
For Google Analytics the matter is relatively clear: A new session starts after 30 minutes of inactivity (= session timeout). True for websites but what about mobile apps? Is a session already finished with the user leaving the app while quickly copying an address from a message and then coming back afterwards? And how do you evaluate this for different app verticals? Wouldn’t it make sense if a news app allowed a different time out than mytaxi for instance?
These are all questions without a fixed answer that should be matched with a user’s behavior and your business model. But don’t worry, we come across this question with almost every client. You are not alone!
To cohort or not to cohort?
We often come across the cohort analysis being referred to as “game changer” for mobile app marketing. But in fact, comparing and analysing user behaviors does have more value for app analytics than for analysing web pages.
Especially looking at differences of cohorts across different app versions becomes a powerful tool. Another classic is analysing cohorts by source of installation. With this analysis you would for instance look at how many purchases were done by users that you bought through Facebook vs. Twitter.
Same same but different?
Good news: despite some differences you won’t have to abandon your web analytics knowhow once you start off with app analytics. Measuring user behavior is a huge focus of app analytics. But at the same hand it’s it is a focus for website analytics as well. Event- and conversion tracking as well as campaign tracking is a part of app analytics as well.
There is no other way in saying it: An app analytics software is essential for every successful app. Google’s own suite for apps, Google Firebase, is one of the obvious and most popular tools. But is has its limitations. What we see when working with clients is that app developers often decide on Google Firebase because it seems to tick off many boxes for the marketing department. Reality and usability for marketing managers is a different story.
We don’t want you to get lost in the jungle of app analytics tools. We have helped many clients choosing the perfect tool for them as well as setting it up. Since we don’t develop our own software we do have a very objective point of view on anything that is available on the market. How can we help you? Let Christian and Raul from Customlytics know at [email protected]
Differences Between Mobile and Web Analytics (Localytics)
Web Analytics vs. Mobile Analytics: What’s the Difference?
Mobile Analytics: A Complete Guide to App Retention and Engagement (Amplitude)