Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Ask HN: Key mobile metrics I should monitor?
3 points by kevbam on April 10, 2013 | hide | past | favorite | 6 comments
Hi, I launched an app in late December(oncoassist.com) and have seen fluctuations in usage since. I am wondering, what are the key metrics I should be measuring? I am using flurry.com and don't find it very good.I am used to google analytics for websites so maybe I am a little spoiled. Any thoughts or links to good blog posts would be greatly appreciated.


You could look into integrating Google Analytics with your application if you have not already done so.

https://developers.google.com/analytics/devguides/collection...

I have not done this myself so I am not sure if Google Analytics is any better than Flurry.


Will have a look into this,thanks.


Without trying your app (not available to US users) or knowing if it's a freemium or ad model, I can't really provide too much insight.

KISS and Mixpanel are nice, but there's little to no point in using them if you don't have your KPIs prioritized, and they may be overkill. I've found it just as effective to track growth via Google Analytics and a series of SQL queries, in conjunction with a dashboarding service like chart.io. I also used Flurry for a while, but it frustrated me quite a bit (Mixpanel offers a lot of the same features, btw). And if your app is HTML5, GA event-tracking is just as effective.

Since mobile is the primary interface, month-over-month growth rate is your live-or-die 'macro' metric, calculated as simply ('users_at_end_of_last_month' - 'users_at_end_of_prior_month') / 'users_at_end_of_last_month'. As of Apr 10, users_at_end_of_last_month would be Mar 31 and prior_month would be Feb 28. Many contextualize this metric as 'churn,' for which there are numerous calculations. Shopify gives a good example. [1]

Of course, it's way more complicated than this. And other KPIs include iTMS downloads/ratings/reviews, average time-to-subscribe, average time per session, number of events per session, DAU/WAU/MAU, e-mail conversion (assuming you manage lists/campaigns), etc.

There was also a pretty good overview of how to set up cohort analysis via SQL queries on HN today [2]. Here's where you may want to start using Mixpanel, but only if you're running targeted campaigns.

Most importantly, read up on Andrew Chen's blog, starting with this post [3]. This series by the Rapportive CEO [4] is also a good model.

Hope this helps.

[1] http://www.shopify.com/technology/4018382-defining-churn-rat...

[2] http://blog.getmetrica.com/post/44824108745/calculating-roll...

[3] http://andrewchen.co/2009/01/19/how-to-create-a-profitable-f...

[4] http://www.linkedin.com/today/post/article/20121002124206-18...

EDIT: URL formatting


Thank you very much for this, very helpful. I will get reading.


Mixpanel is the top-shelf solution at the moment. There are a variety of other options, including flurry, KISSmetrics, and a new product from testflight, depending on your preferred price/features blend. Try quora for compare & contrast questions.


Thanks for your comment Greg,I will check those links out.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: