Funnels
A funnel is a data view dedicated to analyzing users' in-app behaviors. It consists of steps that mark key events users trigger as they use your app.
You can configure two types of funnels:
- Standard funnel
- Progression funnel
Standard funnel
A standard funnel can include any type of event you receive from the justtrack SDK.
Example
Imagine you've created an app where users can search for and listen to music, make playlists, and more. You might configure a funnel that tracks the following steps:
- User opens the app
- User creates an account
- User plays a song
- User adds a song to a playlist
- User is shown an ad
- User clicks an ad
- User makes an in-app purchase
With this configuration, you can analyze things like:
- How many users enter the funnel (complete the first step)?
- What percentage of users in the funnel complete each step?
- What percentage of users churn at any given step?
- and more...
Progression funnel
A progression funnel includes only jt_progression events, but provides more detailed data about each step in the progression.
Example
Imagine you've made a mobile game. You might configure a funnel that tracks player progression through the levels. With this configuration, you can analyze things like:
- How many users enter the funnel (reach the first level)?
- What percentage of users in the funnel start or complete each level?
- What percentage of users fail at a level?
- For players that completed the level, how long did it take?
- and more...
Funnel steps
Every funnel has a series of steps. Each step is associated with an event the user must trigger to be included in the funnel and counted toward the user total for the step.
For example, imagine your game has the following progression funnel:
| Step | Event | Users |
|---|---|---|
| 1 | Level 1 | 90 |
| 2 | Level 2 | 75 |
| 3 | Level 3 | 50 |
For each step, we count the unique users who triggered the event. In this example, this would be the number of users who started the level. Users who didn't trigger the event are filtered out of the funnel, and are not considered for future steps. Therefore, users counted for any given step must have matched all previous steps as well.
It's important to note that the order of funnel steps is critical in understanding user behaviors. This isn't because the users have to complete the steps in a particular order—they don't. Rather this is because the order of the steps dictates the users who are still considered part of the funnel and eligible to be counted for the next step.
Consider this change to the example given above:
| Step | Event | Users |
|---|---|---|
| 1 | Level 3 | 50 |
| 2 | Level 2 | 50 |
| 3 | Level 1 | 50 |
Because the levels are reversed, and Step 1 now looks at users who started Level 3, only those users are considered for future steps. The users who didn't start Level 3 are filtered out of the funnel immediately. Even though 75 users completed Step 2, based on the numbers we saw in the previous example, only the 50 who matched Step 1 are in the funnel and eligible to be counted for Step 2. The same logic applies for Step 3.