I have tons of ideas on how to increase user engagement: better onboarding, more personalized email notifications, etc.
The point is: I have no clue what my user engagement is at the moment. And you can't improve what you can't measure.
Sure, I track how often people log in. But I'm sure I can do better than that.
With all the SaaS founders on IH, I'm sure some of you must have some good ideas on how to measure user engagement. And more specifically, whether that shiny new feature actually helped to improve it?
Any suggestions would help:-)
Everyone's app is different, but there are some commonalities.
You can measure where people come from, how they navigate, and how they interact. An analytics tool like Google Analytics, Mixpanel, Amplitude, Simple Analytics* or Fathom* can give you those things and make it easier to find common numbers such as MAU since it's already counting those activities on a per-person basis. (*The privacy-centric ones have some caveats).
Your business will probably have a small handful of key metrics. You can take baby steps to work backwards through your app, figure out what touch points matter for that key metric and where it's affected, and focus your measurement efforts there, using tools like the list above.
I can walk through that with an example. Suppose you're an e-commerce shop and you want to raise ARPU. Working backwards, you could a) increase the number of orders per user, or b) you could increase the amount of revenue per order.
For a), you could focus your efforts on customer retention, so you'd want to watch for (and optimize for) return visits to your shop. That leads you on a path toward retention efforts, so now you're watching Visits per User as a KPI. Are you able to capture that?
For b), you could focus your efforts on upsells, cross-sells and adding more products to cart. That probably means you'd want to capture the amount of revenue in the cart, Product Views, and maybe interactions with a product recommendation system. Are you able to capture those?
By the by - I'm writing a book that begins with #1 above and enables you to go down whatever path you want for #2. If this is sounding like your jam, check out signup.analyticsforindies.com to get notified when the book is out.
Cool, looking fwd to the book!
Struggling almost daily with this one too. Still experimenting a lot and trying different things. I'll start with the technical setup first...
I started with pure SQL queries running in SQL client. Who are the people who logged in recently. What did they do. Obviously, this became a hassle to ssh to the bastion host and then connect to the database. I then built a simple web page with that same info. But it also became a problem since my questions were changing over time and I had to spend time updating the code and re-deploying it. Currently I am using Metabase which is a free app that connects to the database and allows you to build dashboards. Pretty neat. You can share it with other team members too.
Now to the actual questions about engagement. What I have figured out for myself so far is that every app is quite specific and so metrics vary. There's a very informative video from Startup School. It is about the metrics that investors want, but this is also something that you want too.
At getshipit we help people plan their product roadmaps. How often do you need to update it? Definitely not every day, so looking at how many times people login is an absolutely worst metric. But you do need to have a look at it at least once a week or every other week, do an update to some timelines and items. And by the end of the quarter you have to work on it a lot. So we track item updates and item creation per roadmap per week. We don't care about daily metrics here, all aggregates are done per week. This is the core metric that everything is based off. It is likely to change over time as the product evolves together with my understanding, but so far it's the best there is :-)
Now how do you measure whether your onboarding had any dent? Or that new feature released last week? You grab existing users who logged in to your app since the feature was released and track their engagement using your core metrics and compare that to a period of time before the release. And then you compare that user group with the new signups. Do you see an improved engagement between your new signups and existing users? Do you see an increase of activity before and after the release?
It's much easier to analyze engagement when you have your core metrics defined. For us it's those weekly item updates and new items created on the roadmap.
As I mentioned in the beginning, I'm still experimenting a lot and trying to figure out the best ways. I am not a data scientists and that sucks :-D
Hope it helps a bit.
Metabase looks pretty cool, I'm trying it:-)
What you describe seems to be a sort of cohort analysis, and that totally makes sense. I hope Metabase can do that, if not, I'll write some custom queries:-)