A session is the duration of a visit to your website/community. If you leave and come back later, that counts as a new session.
I’ve noticed a couple of organisations recently trying to measure communities by tracking session data. i.e. do people make a purchase or cancel their service after visiting the community?
This is a terrible way to measure value.
In a recent example, a customer asked the community how to cancel a service. The community did its job and provided an answer. The response was so good that it soon became one of the most visited posts in the community (thanks Google!). But on the stats this shows the community is now causing people to cancel!
That’s the problem with session data. Session data attributes the outcome to the last action in the chain of events before that outcome. This is a lot like attributing sales at a restaurant to the menu.
Not many people make a snap decision to cancel/buy a service because of a single experience on a single visit.
To really estimate value you need to build a dataset with a broader set of variables. This is likely to (at least) include:
- Member ID
- Community participant? (yes/no)
- Date of community registration
- Date of first post
- Date of last post
- No. visits/logins to community
- No. posts
- No. topics
- Date of subscription (customer)
- Date of cancellation (customer)
Once you have this your data team can run an analysis (or you can get help) and start building a meaningful picture and range of the community’s impact.
This won’t give you causational data, but it’s a start.
