Andy Carvell
Jul 10, 2017 · 1 min read

Lee,

If you want to *primarily* do investigative analytics and visualization (segmentation, funnels, event-based cohorts, etc.), then my opinion is that this activity is best served by a dedicated tool (either home-grown or licensed). Strong solutions for this kind of behavioural analytics deep-diving include Amplitude, Mixpanel, etc. It’s true that attribution providers also have event tracking capabilities, and often pretty decent cohort analysis too, but imho the use case here is primarily Marketing and lack the depth and flexibility provided by dedicated analytics platforms.

There’s a lot of convergence in the tools space, with marketing analytics tools building out deeper behavioural analytics and the behavioural analytics solutions adding marketing or even attribution capabilities in some cases. Unfortunately, there’s still no killer solution that covers the whole stack, which is why most companies end up building some of their own tech, or patching together various 3rd party tools using their APIs, or with data layers such as MParticle, Segment, etc.

Some folks build dataviz solutions (Tableau, Chart.io, etc.) over their own data warehouse. I see this working better for dashboards than investigative analytics, though once again the line is blurred and perhaps there’s a very flexible setup in this direction that I just don’t happen to have seen yet (if so, I would love a demo!). Hope this helps.

Andy Carvell

Written by

Co-Founder at Phiture Mobile Growth Consultancy — we help apps grow. Formerly Mobile Growth @ SoundCloud, living in Berlin, listening to techno.