Practical Data Dictionary: Few more words about your Users’ activity

This is “Chapter_02: User-types from an activity perspective” from the Practical Data Dictionary. (Note: Read the previous Chapter: here!)

In the previous chapter I wrote about the activity types of your users. Now it’s time to continue with the different user-types!

Visitor: Someone who visits the website, a potential Registered User — but not necessarily one.

E-mail Subscriber: A visitor who provides their email address.

Registered User (in short: User): The kind of Visitor who registers, so provides their email address, their Facebook account or any kind of unique identifier, for which we create a user account.

Onboarded User: A User who has gone through the so-called Onboarding-process.

Active User: This is a changing status. The kind of user who uses our product in a specific time-frame marked by us (e.g. a given month, given week, given day or given hour). Note: Again! If the user logged into her user account, it does not necessarily mean that she used our product as well. You’d actually be surprised to see the ratio of the logged-in-but-did-nothing-else user ratio on many product… It is worthwhile to link activity identification to the end of the Onboarding process: it’s often suggested to make it the very end (e.g. with an invoice issuing software: they logged into their account — » we don’t consider this activity; they sent another invoice — » this is considered activity).

Inactive User: This is a changing status. The kind of User who does not use our product for a specific time-frame marked by us (e.g. a given month, week, day or hour).

Churned-out User: This is a changing status. The kind of User who has not used our product for a specified, lengthy time-frame marked by us (e.g. the past 3 months, past 1 year, etc.).

Deleted User: The kind of User who we deleted from our system or who has deleted themselves.

Note1: If you check the process diagram again, it will be clear that the E-mail Subscriber, Registered User and Onboarded User status’ are one-time status’. The main goal is to push our Users through these — as many as possible — and to keep them as Active Users for as long as possible. This will not work with everyone of course. From this it follows that there will be relatively low Users in the E-mail Subscriber, Registered User and Onboarded User status. Most of the Users will be coming and going between the Active/Inactive/Churned-out status’. 
However, it’s still worthwhile to have the E-mail Subscriber / Registered / Onboarded categories segmented as these Users are very fresh and curious. Due to this, they are „sensitive” about many things, thus they are easy to handle, ideal Users for you.

Derivative user types from an activity perspective

During User Research, we aren’t only interested in what phase they are in now (Onboarded, Active, etc.), but also in what phase they were in before. It makes a difference whether an Inactive User — prior to Inactive status — only registered and did not try the product yet (was a Registered User), or he/she tried the product, but only once (he/she was an Onboarded User), or he/she used it often (was an Active User). It’s sometimes advised to segment the users from each other from this perspective as well.

INACTIVE USER SEGMENT

Registered-then-Inactive User: The User who after Registration immediately became an Inactive User. Comment: Another coined term is Dead-On-Arrival.

Onboarding-then-Inactive User: The User who after Onboarding immediately became an Inactive User.

Active-then-Inactive User: The User who was Active, but then became Inactive.

ACTIVE USER SEGMENT

Onboarded-then-Active User: The User who went through the Onboarding process and stayed an Active User.

Active-then-Active User: The User who was an Active User and stayed an Active User.

Inactive-then-Active User: The User who returned after Inactive User status (Win-back) and then became an Active User.

Churned-then-Active User: The User who returned after Churning status (Win-back) and then became an Active User.

Note1: It could be interesting to broaden these groups based on our own preferences. E.g. 5*Active User (the User who was an Active User 5 weeks straight), etc…

Note2: At the same time, it’s not worthwhile to create too many subcategories either as it’s easy to lose focus if we concentrate on many segments.

Note3: Since we touched on the topic of focus! It’s a basic question of strategy on which of the above categories (8 + 3 + 3 + your own subcategories = 14+) we concentrate on. A lot of literature exists on why it’s better to pay attention to the Registered Users rather than the Inactive Users, or why Win-back is more valuable than Retention. These are interesting reads… BUT! Your product, your strategy and your Users will determine who you will focus on — for this you need to analyze your data, and not follow other people’s advice. Check it out and decide what’s important for you and with measurements identify what you need to place in the center to achieve this.

User groups on a time basis

The above User groups are more easily manageable if you divide them into groups on a time basis (e.g. Daily Active Users). Based on our personal experience, it’s practical if these belong to not relatively but absolutely determined time periods. So we are not watching those who were Active Users in the past 24 hours (as this is a constantly changing group), but those who e.g. were Active Users between 2016–01–01- 00:00 and 24:00 (as this is a fixed group, once 2016–01–01 24:00 has passed, then the distribution of the group does not change).

These groups also need to be generated by you based on your needs, but here are some examples:

  • Daily Active Users (e.g. the number of Active Users on 2016–01–01 is: 352)
  • Weekly Onboarded Users (e.g. the number of Onboarded Users on W1 of 2016 is: 1.860)
  • Yearly Churned-Out Users (e.g. Churned-out users in 2015 is: 21.512)
  • etc, etc…

Want to read CHAPTER 3? Continue: here!

Want to have the full 54 pages e-book right now? Download it here: http://data36.com/datadictionary/

Cheers,
Tomi Mester