In-App Analytics: Any Is Better Than None!

It blows my mind when I hear app developers and publishers don’t have in-app analytics setup. It makes me feel likes it’s 2001 and we are just discovering the internet and everyone is still treating it like offline marketing and looking at server stats for insights or in our case downloads.

You need to be using in-app analytics. This is not a ‘would be nice to have’ item and there is really no excuse to not have it. Most apps analytics are free for a certain volume of users. If install analytics is all you’re using you’ll never know what’s working and what’s not, what pages are being viewed the longest, what buttons are being pressed, where the user might be having a problem, crashes etc.

Which In-Apps Analytics Package? Any, Just Pick One!

One is better than none. But Try Many.

There are many, many packages out there. What I would suggest is that you find one which you feel comfortable using i.e. easy, quick and gives you the right data.
Load a couple in your app and try them out, find the one you are most comfortable with. I know there is an argument that having lots of analytics codes in your app will slow it down and this does have merit but it’s only for a limited time and in the end you will have working experience with the analytics tool you want to use and be able to compare each platform.

I’m busy too look for one

OK Because I really want you to start using any analytics packages I’ll give you a list of places to look at:

Google Analytics Free
mixpanel Free
Kissmetrics paid
Localytics Free Trial
Flurry – Free

…the list goes on.

I tend to recommend mixpanel to clients as you can get granular data and it also has push message functionality. This means you don’t need to have two separate pieces of code in the app. Thus keep the app as fast as possible.

Google Analytics is OK and certainly better than nothing but the data it gives you is pretty limited and frankly isn’t very actionable. As you’ll be dipping your toes in the water you probably won’t want to pay for Kissmetrics…also I’m not sure how different it is from mixpanel…I’ve never used it.

What The Heck Am I Tracking?

I don’t think I’ll be going too far when I say, EVERYTHING!!!!!

It’s a lot easier to track everything in an app than it is to track everything on a website. Your app will serve a particular purpose. This means it’s got ‘limited’ functionality and, usually, a set number of buttons and actions.

Though if you’ve got an app like tinder or Stylect where a user could swipe hundreds of times in a session you might end up with a problem that you go over your usage limit and get charged, keep this in mind if you’re boot strapped.

How Would I do It?

I would make a list of all the buttons and actions on an app. Next to that list I’d put a friendly name. Then I’d give that list to developer and ask them to make a tracking event for each button and actions. How this is set up is slightly different for each analytics package but they are all well documented so your dev or yourself won’t have a problem figuring it out.

What To Track

You usually need a few to gauge your businesses success and failures. And there is no simple answer. Pages views, times in app etc aren’t really worth governing your strategy with. Unless you have a content lead app then there is an argument for that. Ideally you want more users that spend more time in your app every month. So downloads vs quality of downloads. If you run an ad that gets your 100 downloads and 90 un-install instantly, then you need to fix something or use and another channel for acquisition.

Ideally you want to pick a metric that has an solid comparable figure or rate and which makes an impact on your business.

I can’t really tell you exactly how to do this in a blog post. Every app and business is different.

Now, Go Do It!

You should now have the basics of installing in-app analytics. There will always be a few extra questions and if you have any feel free to stick them in the comments below that way everyone can see the response and won’t need to ask it again.

Over and Out!

App Overlap, Does it Matter? Can it Feed App Strategy?

There are some tools out there that allow you to investigate the other apps your users installs. At first glace some might think this isn’t very useful…I’m going to do ‘me’ and see if I can find something useful in this data set.

This might not work out the way I planned. If you’ve not seen my presentation on Windows App Store Optimisation check it out. The hyposis hasn’t worked out to date, but we’re still trucking and figuring out the key signals.

How we will test how useful this tool is is by picking 3 strongly ranking apps from the car hire industry and look to find a pattern in overlap in the Top 10 shared apps on a device. We’ll look at dates ranges from 1 year, 3 month and 1 month as this is the time frames we have available to us.

Read more