Hi!I like to group conversion funnels and cohort analysis into what I call “Positive Influence Metrics.” This is when we look at the reports generated from the data we collect about users to determine the whether they are taking the desired action we want them to. Hence we look for repeat positive signals showing users taking the actions we want them to. These metrics and reports show us whether or not our hypothesis is correct.
However there is something I call “Negative Influence Metrics.” I use it for measuring the actions I DO NOT want the users to take.
This isn’t the same as looking for negative trends in cohort analysis or conversion funnels. You may want to consider low conversions, lack of retention among cohorts, high churn, etc as “Negative Influence Metrics”, but what we really want to do is identify specific actions that a user should not take if they are happy using our product.
For example I want to know when a user disables an account, opts out of email notifications, or revokes permissions from facebook or twitter. Collecting these actions is important, but the real value comes from when you, as a company, take action on being alerted that someone has taken a negative action.
After being alerted that a user took an action I did not want them to, I want to know the reasons why they did. Was it because they were receiving too many emails, were they skeptical about how I am handling their personal data, or did they misunderstand the value proposition and realize that the service wasn’t for them?
It’s a step in opening a dialog with a customer who is heading for the door and gives me an opportunity to learn why and evaluate how to address the problems they encountered with the service. Early on I believe these are going to be some of the most valuable conversations I will have with my users.
