View profile

The Growth Weekly - Product Analytics


The Growth Grind

October 22 · Issue #2 · View online

Your update on Growth news! 🚀

Thanks for subscribing to the newsletter. Since the first issue last week, you have already more than doubled in size! 🎉
I’m very grateful for your attention. 🙌🏻
Welcome to the second issue of The Growth Weekly. If you’ve missed out on the first one, you can find it here.
If you have any questions on product growth, don’t hesitate to reply to this e-mail. And please let me know what you think about the newsletter.
Let’s go!

Digital Product Analytics
So last week we were talking about retention as a very big driver of retention. But before you can do something about your retention rates, you’ll have to be able to measure them.
Coming from web analytics myself, getting into product or app analytics was daunting. The amount of tools available on the market is enormous and they are very fragmented. In order to possibly help you out on this, I’ll outline some of the tooling categories you might need.
(Photo by Alexander Andrews on Unsplash.)
(Photo by Alexander Andrews on Unsplash.)
Behavioral Analytics
The first thing you’ll want to do is knowing what your users are (not) doing. Behavioral analytics will be able to show you just that. There’s a lot of tools that you can use to do this, but before you go around shopping, you should set up a data taxonomy.
It’s one of the most boring jobs in the world, but it really pays off to do this. Give a name to every screen, event, object or user property within your product. For example, this is the one Twitter used:
The image links to an awesome article on Twitter's Logging Infrastructure.
The image links to an awesome article on Twitter's Logging Infrastructure.
After this, you can decide on tooling. I see most people use either Firebase Analytics or Google Analytics for Mobile Apps. Mainly because they’re free. They’re alright to begin with, but they both are lacking in some areas. This guy explains the difference very well.
Firebase Analytics vs Mobile App Tracking - YouTube
Google will probably fix this in the coming years, but until then what I like to do is set up Firebase Analytics and send the data to Google Analytics through Google Tag Manager. This way I have the best of both worlds.
I have to say that making this connection has some major pitfalls you need to look out for and I’m currently writing an article on how to do it. In the meantime, hit me up if you’re experiencing any troubles.
However, the tool I greatly advise you to give a try is Amplitude. It’s a product analytics tool that is much more adapted to the current needs in event tracking and cohort analysis. They give you a lot of freedom in how you set up your event tracking and run your querying.
They also have a cool guide on mobile analytics:
The Ultimate Guide to Mobile Analytics | Amplitude | Amplitude The Ultimate Guide to Mobile Analytics | Amplitude | Amplitude
You can of course build your own product analytics by logging events, warehousing and visualizing data through other tools. It’s not something you should put time and effort in as long as the above tools can help you out at a reasonable cost.
If you decide on doing this, here’s an awesome article from Samson Hu, who built up product analytics at 500px.
Building Analytics at 500px – Samson Hu – Medium Building Analytics at 500px – Samson Hu – Medium
If you’re running campaigns to acquire new users, another thing you’ll want to know is how these different channels and campaigns are performing.
This is where attribution tools like Adjust, Branch or AppsFlyer come in.
It’s a lot harder to attribute app users to a channel because the app stores are in between the install and the channel. So mainly what these tools do, is tracking the app store url clicks and then matching them to new users.
You could solve this problem by installing the SDK of the network you’re using (e.g. Google Mobile Ads SDK, Facebook SDK, …) but it’s not really a scalable option in the long run.
So when choosing an attribution tool, make sure they have the right networks they partner with. Built-in attribution features in Google Analytics or Firebase don’t work with the Facebook network for example.
Also important in this decision is if and how they handle (deferred) deep linking. This allows you to send people directly to a certain place in the app. And if they don’t have the app yet, deferred deep linking will send them to the correct app store first before bringing them to the right screen.
What is Deep Linking? | Branch What is Deep Linking? | Branch
Qualitative Analytics, A/B Testing, ...
And then there’s all the other categories. I’m not going to go into these right now, but if you want me to, please let me know.
  • Qualitative Analytics to help you get even closer to your customer by making heatmaps, session recordings. (e.g. Appsee)
  • A/B testing to test different versions of your app. (e.g. Apptimize)
  • ASO tools measure your app store positions for important keywords or categories (e.g. AppAnnie)
  • Performance to see whether your app crashed, runs slow…
  • Tag Managers help you send data to the different tools you need it in. (e.g. Segment, Google Tag Manager)
  • Dashboards that gather data from different sources and showing the bits a team or person needs.
  • Marketing Automation that either feeds of your existing event tracking or needs its own.
Everyone gets data!
The danger that lies in this complexity of tools is that you’ll have one go-to data person or team. They specialize and ‘own’ the data. Once your data infrastructure is somewhat in place, it’s a good thing to put some effort in getting data (knowledge) to the rest of the team.
It will be an ongoing effort but in the end it will make sure people can make better decisions with more confidence. There are some great articles on this subject linked below:
At Airbnb, Data Science Belongs Everywhere – Airbnb Engineering & Data Science – Medium
Lessons From 500px on How to Setup Analytics for Scale
Minimum Viable Analytics – The Mobile Growth Stack
What's next?
Phew 😅 Managed to get through all of this?
I’d love to get to know what you think about this and what subjects you want to read about more!
Next week, I’ll talk about growth accounting and modeling.
Please let me know what you think about this issue by reacting below. 👇🏻
If you know anyone who should be on this list, please let them know about it. It would help me out a lot! 🙌🏻
Thanks for reading. Hope to see you next week!
Kind regards,
Did you enjoy this issue?
In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue
Stenemolenstraat 44, 9870 Olsene, Belgium