Analytics

Analytics is, essentially, about analysing the usage and/or performance of some website, app, or digital product by tracking a variety of interactions and datapoints such as time spent on a particular page. They can tell you what is happening in a UI, but not why.

A classic example of a user analytics tool would be the ever present Google Analytics.

The pros and cons of analytics

Areas where analytics are particularly powerful

  • Revealing potential problems reaching goals
  • Identifying potential root causes of problems
  • Quantitive data to supplement your qualitative research

Benefits

  • Quickly validate/invalidate causation/correlation theories
  • Help persuade data-orientated stakeholders

Challenges

  • Analytics can tell you what, not why.
  • Defining and scoping metrics - so much can be measured, but not everything is meaningful.
  • Getting to answers - which metrics best answer you question(s)?
  • Set up - how do you set up the system to show what you want to find out?
  • Depending on what or how much you are tracking, it's easy to cross the line into tracking data which users may not want you to know (ex: in highly regulated industries or people just not wanting you bombarding them with tracking cookies. See the Superhuman email tracking fiasco)
  • Depending on the tool and how much you're loading into the UI, you can start to negatively impact performance.

Tips

Do

  • [[!Start with why]]
  • Step back and think about how analytics data can supplement your current work and research process, instead of jumping right into the tool.
  • Remember there are multiple ways to measure. If you start with why and questions to answer, some of them might be better (or only) answered with [[!Qualitative research]] or other methods.
  • Create an Analytics measurement plan, starting from overarching goals rather than metrics.

Don't

  • Track anything that might be 'interesting' without knowing why or how you will use that information. Analytics can become a ‘distracting black hole of “interesting” data without any actionable insight’
  • Focus on the analytics tool instead of what it is there to support.
  • Start with what you can measure with analytics, measuring the wrong things or creating an easy way to say ‘oh we succeeded’.

References