5 Replies Latest reply on Sep 3, 2015 1:35 PM by nbussard

    Canned "Cohort Analysis" Report?

    neilmussett

      Reading Eric Ries article on "Vanity Metrics vs. Actionable Metrics" got me inspired to create a "Cohort Analysis" report for our installation of Jive.

       

      The idea is to group your users according to when they joined (week or month), then track a percentage of users still active within that group (defined as "having logged in in the past X days"). This tracks how "Sticky" Jive is within your organization.

       

      It was not too difficult, and it led to some very meaningful insights (e.g. Jive is hard to learn, and it takes about 33 weeks for people to make a place for Jive in their daily work lives). Seems like this kind of report would be a very compact way for group owners to manage their metrics.

      Detailed Cohort Analysis

      Ted Hopton's post touched on the issue: Actionable Analytics Presentation JW2010

       

      PS. The mechanics:

      1. Every week (or month), produce a report that lists each user, along with the creation date, the last logged in date, and any demographics that you want use to break down the numbers.
      2. Group users by the week (or month) they joined Jive
      3. Calculate a percentage of users in that grouping that are still active. This is your "retention rate" (and the converse is your "churn rate")