Aptify's CEO Blog

Subscribe to this Blog with RSS
CEO Blog Home

Nov
11
 Predict if a trial member will convert, renew and engage
  Member Engagement  Predictive Analytics  Trial Memberships  Comments (0)

A recent discussion within Aptify’s User Community on trial memberships sparked my interest in exploring this topic with a broader audience and thinking about how trials might tie into previous topics on this blog including Predictive Analysis and Member Engagement Scoring models.

Trial offers are fairly common in the membership world. But just how good are trial memberships in terms of bringing on new members? The simplest answer to this question might be sought by looking at a simple initial conversion ratio. While this number may tell the early part of the story, it is very important to evaluate the trajectory of that member over a longer period of time. As an example, how active is this new member in other areas of the association – a measure of their overall engagement. Furthermore, how likely is it that the member will renew in Year 2, 3 and beyond?

Are there patterns that can be used to predict what might make one individual more likely to convert past a trial, get active, and then becoming actively engaged? The technology is out there to help with both the predictive analytics question as well as creating a method of “scoring” engagement levels. The question of when to offer trial memberships is a great example that should leverage both of these concepts.

Consider it this way – if you can find a few defining characteristics that help assess if a prospect will go from trial to member and beyond there would be many applications. It would be great to build those characteristics into your core operating processes. Imagine a scenario where your AMS (for staff and on the web too) automatically determined the probability of “success” based on several factors and then selectively (and automatically) offered trial memberships to individuals or groups that fit the model.

As I mentioned in earlier posts, none of these modeling techniques are perfect, but they often do reveal interesting trends. At a minimum, they are worth a hard look. If one or more model is good at predictive work for your organization, find ways to use it frequently and in nontrivial ways.