The ClearBrain platform doesn't just tell you who is predicted to perform a goal, but also why they are predicted to do so.

When you predict a goal in ClearBrain, the platform builds a composite profile of the users who performed the goal in the past week, evaluating each user attribute and its relative importance to the predicted goal.

In assessing relative importance of each attribute, ClearBrain goes one step further with Benchmarks. Benchmarks tell you not just which attributes are most important, but at which threshold in user activity they lead to the highest increase in likelihood to perform the goal. Consider it like finding your version of Facebook's "7 friends in 10 days".

The information provided by the Benchmarks analysis include:

  • Attribute: the respective user attribute that is analyzed
  • Benchmark: the threshold in user activity that leads to the highest statistically significant likelihood for your users to perform the goal
  • Probability: the probability that your users will perform the goal at the respective Benchmark; a range of probabilities are provided representing a 95% confidence interval
  • User Count: the count of users that performed this attribute at the respective benchmark in the previous week.

As an example, if you see a Benchmark for an action "Clicked Button Last Week --> 4 --> 13% - 20%", then that means user who did the "Clicked Button" action last week > 4 times have between a 13% - 20% probability to perform the goal.

Did this answer your question?