The majority of analysts in the past decade have focused on basic performance measures designed to monitor and improve revenue and conversion rates. Experiences from these projects showed that carefully designed interactive, real-time data systems can and do make a major difference to the bottom line.
In today’s post we are going to discuss Behavioral KPIs and their effect on analysis results. Behavioral KPI is a subset of behavioral analytics which is used to conduct complex analysis, such as cohort and data mining. Behavioral KPIs are used to measure trends and are there to put your company on the right track when it comes to conversion rates, user experience and retention. These KPIs are divided into two: frequency, which is the number of actions, and recency, which is the remainder. In other words your KPIs help you extricate the delta, which is your success rate.
The best way to really understand behavioral KPIs is to see how they work on different widgets.
Let’s say we want to try and get our users on certain paths, for example, paths with the highest subscription rates, or paths with the highest revenue. We can direct our users to these paths by using behavioral KPIs and then see if our traffic grows as predicted.
We measure retention rates with the use of cohorts by taking a chart with last week’s results (behavioral KPI’s) and comparing it with a chart illustrating this week’s results. This analysis creates our delta, which defines our success rate. What we learn from this example is that Behavioral KPIs actually measure your success rate.
With behavioral KPIs you can also measure the rate of traffic between different segments. Let’s use an example from the world of gaming; say we have two segments, one is a segment of users that are “one game users” and the other is a segment of “multi-game users.” We want to get the “one game users” to become “multi–game users”. How do we do this?
By using Behavioral KPIs on certain paths we can measure the traffic between the different segments and analyze the user’s trends. This way we can determine which paths are more likely to lead to “muli–game” segments and then direct our “one game users” to these paths. In this case our Behavioral KPIs can range from conversion rates, user acquisition costs, session times, intensity (number of clicks per session), drop off rates by level and so on.
One of the most interesting topics Behavioral KPIs deal with are predictions. Predictions are mainly used to recognize churn paths. We can use certain actions to decrease churn rates with the help of Behavioral KPIs. So how do we do this? We measure a period of time and predict our churn rates on a certain path. Then we take another cohort on the same path and pick an action, such as an ad campaign. Now we can compare the two navigations and see how our users responded to that action, or in other words, what were the action’s success rates?