So I had designed and developed a minimum viable product for my HipHopListings iOS app. I had worked very closely with Alex the developer, designed the mobile workflow and outlining the way I expected users to interact with HipHopListings on their iOS devices. I tested, tested and tested. It led to some critical bug fixes and improvements to both the back-end and the front-end of the app.
My jaws dropped when Apple approved my HipHopListings app within less then a week after submitting it. You can find the app here and I’ve added a screenshot below (see Fig. 1). Based on my previous experiences with getting apps approved by the mighty Apple, this was super fast!! I celebrated and felt like the proudest man on the planet as I’d just launched my own app and had been able to do so within less than a month from actually designing the app screens!
However, I realised that I couldn’t just rest on my laurels and had to start measuring if and how people were actually using my app. If I wanted to learn about how to best improve my app, I really needed to start measuring things, but where to begin!? I had implemented Google Analytics but wondered what metrics to focus on, especially as there are countless analytics ratios and reports to choose from.
I decided to try and identify my key questions and assumptions that I wanted to validate first:
- How many people are using my app on a recurring basis? – My simple reasoning was that the number of people who had installed my app (which numbers are available via iTunes Connect) is pretty meaningless when those people never use the app again. I therefore wanted to be able to do some form of cohort analysis to get a better understanding of the usage patterns of new users, developing an insight as to wether my app was compelling enough for people to keep coming back to it.
- Where are most HipHopListings users based? – One of my underlying assumptions when creating the app was that the bulk of HipHopListings’ target audience were likely to be based in the UK, especially given the fact that the “Shows” screen on my app only lists UK gigs. I now wanted to find out whether this assumption was actually true. Is most of the (recurring) usage coming from the UK? If not, what can I do to attract more UK users to the app? Is there a particular reason that people use (or don’t) use the app?
- Which screens do users tend to look at most? – I felt that looking at screen views (and the duration of each view) could provide me with a better indication of where the app could be improved. When I started creating the HipHopListing app, one of my assumptions was that people would mostly use the app to view upcoming Hip Hop shows (this assumption was part of the opportunity assessment that I did before deciding to build the app). Was this actually true? If I started making slight changes to the “Releases” screen, would these have an impact on the performance of the screen that lists upcoming releases?
- What are common drop-off points in the app? Again, I believed that if I wanted to get a better insight into the understanding of my new app, I’d need to understand the common drop-off points within the app: where do users typically lose interest and why?
Main learning point: I realised as I was thinking about the kinds of things that I’d like to measure about my HipHopListings app that I was falling into a classic ‘data trap’. My thinking was very much around specific questions that I hoped would give me a better insight into the performance of my app. However, I’d forgotten to think about goals, setting some sort of benchmark or baseline of what success should look like for my app. I thus learned that it would be good to first take a step back and identify my main goals, before deciding on the right things to measure.
Fig. 1 – My own HipHopListings app in iTunes
One response to “Developing my own product – Where to start measuring?”
[…] slightly struggling with identifying the most effective measurements for my own product I’m nevertheless learning lots of new things about when (not) to use data in developing […]