Time-Based Subscription Analysis

Overview:

KittyCarts is a fictional eCommerce company that sells monthly subscription boxes containing feline care products. In the accompanying KittyCarts app, subscribers can track and customize their monthly boxes, view their box history, and purchase specific products they liked from their boxes.

Recently, KittyCarts expanded their business by offering a line of branded webcams called KittyCams that allow customers to monitor their cat’s behavior throughout the day. They’ve expanded the functionality of the existing KittyCarts app to accommodate this feature. Prior to the KittyCam’s launch, the app was only available to subscribers with a monthly box plan. Now, KittyCarts wants to explore how many non-box subscribers who installed the app for KittyCam went on to become subscribers of a KittyCarts monthly box.

_____________________________________________________________________________________________________________

Question: At what rate do non-subscribers who installed the app convert to box subscribers over time?

 

Open the Cohort Tool: Let’s click the Cohort icon in the Navigation Bar.

x16.gif

Complete the Cohort Wizard: In Cohort, the Cohort Wizard guides us through the creation of our initial analysis.

  1. Define a Cohort: Let’s select the event “App Install”. This includes any user who installed the app in our cohort.
  2. Select a Behavior: Let’s select the event “Subscribe”. This will show us who among our cohort of app installers went on to subscribe.
  3. Select a Breakout: Let’s select the Time Generation “Day of App Install”. This will break out our results by grouping users together based on the day they installed the app.

replacement_1.gif

We now have our basic cohort analysis results displayed as a table view.

  • Looking at the rows, we can see the breakout of users grouped by day of app install.
  • The columns tell us how long it took users to complete the target action after app installation. Since our time interval is “Daily”, the bolded numbers above each column represent how many days post-install the subscription was started. Above the columns and numbers is a miniaturized bar chart visualization displaying the average of results for each column.
  • The bubbles indicate the percentage of a user group that completed the target action (Subscribe) and how many days after the cohorting action (App Install) they did it.

We can modify our analysis using the options in the grey toolbar at the base of the query builder, as well as by using additional events or properties to create clauses.  Let’s modify our cohort.

 

Modify the Cohort: Once the basic cohort analysis has been created, we can go back into the query builder and modify it to fine-tune our results. Currently, our cohort includes all users who installed the KittyCarts app. However, to answer our initial question “At what rate do non-subscribers who installed the app convert to box subscribers over time?”, we only need to see non-subscribers who took this action. We can achieve this by adding a “Where” clause to the first row in the query builder, which defines our cohort of users.

x18.gif

  • Let’s go to People Properties in the Data Panel and select the property “Subscriber?”, then drag it into the “Where” dropzone next to the event “App Install”
  • Next, let’s finish out our parameters by setting the “Where” clause to “is equal to no

Our new group only includes users who did an “App Install” and who were not subscribers.

 

Select Between First Time and Recurring: For the purpose of our analysis, we only need to know about the first time a user subscribed after installing the app. Selecting “First Time” will limit our results to only show the first time this event was performed.

x19.gif

View Results: Our cohort analysis now shows us the rate at which non-subscribers who installed the KittyCarts app are converting to subscribers over time.

 

Observation

While there are strong conversion numbers within the first day of app installation, conversion drops dramatically after the first day. KittyCart should create ongoing promotions for new app users to incentivize box subscriptions.

 

We are currently viewing our analysis results as a non-cumulative percent. There are a few other options we can use to view our results:

  • Cumulative Percent: In this view, percentages represent a running total of the percentage of the cohort that subscribed over time.


x20.gif

  • Non-Cumulative Count: This view shows us the same results as our default view. However,  the numbers displayed represent a raw count of users who subscribed on a given day, rather than a percentage.

x21.gif

  • Cumulative Count: In this view, results are viewed as a running total of the raw count of users who subscribed over time.

x22.gif

0 users found this helpful