Appuri Documentation

Welcome to the Appuri Documentation. You'll find comprehensive guides and instructions to help you start working with Appuri as quickly as possible, as well as support if you get stuck.

Documentation

Segmenting your Users

The Segments page allows you to segment, or group, users by their properties and events. Segments are automatically recalculated and users can join or leave segments as their behaviors change, matching or dropping out of each segment depending on the Segment definition.

Users can be in multiple segments at the same time. Segments can be sorted by clicking on the column header, so clicking "Name" will alphabetize the segments, while clicking "Total Users" will sort them by user count.

Options for defining a Segment

When creating a segment, first give it a descriptive name. Any segment you create will also be seen by others in your organization, so make sure the name will be understandable to them too.

At a minimum, each Segment is defined as one or more User Properties or one or more User Events.
At a maximum, you can have as many Users Properties and User Events as you'd like, and a user must meet all of the included requirements to be included in the segment.

User Properties

User Properties are properties which are set on a user level and will remain the same until updated. Properties are set and remain the same until they are overwritten.

Common examples of User Properties:

  • Geographic - Country, State, City, Region, Timezone, Zipcode, etc
  • Unique identifiers - UDID, Address, Name, Email, username, etc
  • Monetization information - Lifetime Value, Total purchases, etc
  • Campaign or source information - Campaign ID they joined from or referral information
  • Aggrigations - _total_events, _first_event_at and _last_event_at

For more information, see User Properties.

User Behaviors (Events)

User Behaviors are events which the user has done. Events can be completed multiple times.

Common examples of User Behaviors:

  • User actions - view page, use feature, open app, click button
  • Session information - start new session, log in or log out, etc
  • Monetization information - purchase started, purchase complete, etc

Each event can have its own properties. Most commonly these will include properties such as timestamp, device firmware or maker, and application version. Some events have specific properties, such as a purchase complete event may have the properties purchase total and purchase currency.

Suggestion: To check out live User Properties and User Behaviors, head to the User Details

On the Users page, click on any user to see their User Details.
On User Details, the top will be any pinned properties, such as KPIs. On the left are user properties. On the right are chronological User Behaviors. You can click on each event in the User behaviors list to see the associated Event Properties and their values for the given event.

Segment Best Practices

Segments are highly flexible, so be sure to create the segments that matter to you.

Some good starting segments that will work on any application:

  • Users active in last 7 days
    • User Property: _last_event_at, Within Last, 7, days
    • User Event(s): None
  • Top 10% of users by total events
    • User Property: _total_events, Above Percentile, 90
    • User Event(s): None
  • Users who have been active for at least 30 days
    • User Property: _lifespan, Greater Than, 29
    • User Event(s): None

Events that start with an underscore are created by Appuri for your use

_total_events is the total number of events a user has sent.
_first_event_at is the date a user was first seen.
_last_evet_at is the date a user was last seen.
_lifespan is the number of days from the first user activity to the latest user activity.
_days_since_last_visit is the number of days since the user has had activity.

If you have marketing attribution data:

  • Users from Facebook ad campaign
    • User Property: source_campaign, contains, "facebook"
    • User Event(s): None

If you want to know many users have used a specific feature:

  • Users who have used a specific feature.
    • User Property: None
    • User Event(s):
      1) view_feature

If you have a tutorial flow:

  • Users who have completed a tutorial
    • User Property: None
    • User Event(s):
      1) signup_complete
      2) tutorial_screen_1
      3) tutorial_complete

Behavioral Steps

If you have multiple behavioral steps, users must complete all steps in order to be included within the segment.

  • Free users who have viewed an ad
    • User Property: None
    • User Event(s):
      1) Did not: buy_premium
      2) Anytime later: Did: watched_ad_complete
      2) Anytime later: Did not: buy_premium

Restrict Behavioral Steps

Behavioral steps can be restricted to Did or Did not
Note: Currently events can only be used before or after other events, the above method is a workaround for a Did Not scenario.

Segments are great to compare groups of users

  • Users on iOS
    • User Property: None
    • User Event(s):
      1) Did: login with event property version equals "iOS"
  • Users on Android
    • User Property: None
    • User Event(s):
      1) Did: login with event property version equals "Android"

Note: Users can belong to multiple segments

If you allow users to log in with the same account on both iOS *and Android, some users may be counted as belonging to both segments!
If you want users who only signed in on iOS, you can add "Did not do login with event property "Android"

Segment KPIs

Segment KPIs, or Key Performance Indicators, are aggregate values calculated from the user properties of the entire segment. They are a great way to compare segments at a glance, such as how segments monetize or retain.

You can create a new KPI by viewing the Settings screen on the left menu, followed by clicking 'Create KPI`

KPIs are visible to all team members

KPIs created by any team member are visible to all others on the Segments page.
KPIs are limited to the current application.

KPIs can only be created from numbers

User properties that are other data types can not be used in KPIs. Only valid User Properties will be displayed on the "Create KPI" interface.

Creating a KPI

Open the "Create KPIs" popup from the Settings page

  • First, pick a user property to base the KPI on, such as _total_events or any other number, percentage, or currency based user property.
  • Next, pick how you want it to be aggregated across users, such as "average", "sum", or "count"
  • Lastly, if it is possible to have the KPI not set (User properties are set to null until set) decide if you want to include null values in the KPI.

Option: Treat null as 0

Careful! This step is important!
By treating null as 0, users who do not have this property set will be included in the KPI calculation.
By not treating null as 0, users who do not have this property set will be excluded from the KPI calculation.

When to set null to 0... or not

Let's say you want to calculate the average Lifetime Value for your customers as a KPI this is common in some industries to calculate your Average Revenue per User, or ARPU. You set the user property LTV to be the KPI, then use "average".

  • If you set null to 0, all users, even those who have not spent money, will be included, thus the average lifetime value will be across all users. This is a standard ARPU calculation.
  • If you do not set null to 0, only users who have spent money will be included in the lifetime value calculation, raising the average by comparison.

In these examples, which setting you use depends on what you want - The average lifetime value of all users (the former), or the average lifetime value of your monetized users (the latter).

Which data types can be used to create a KPI?

Only User Properties which are numbers, percentages, or currency can be used to create KPIs. For more details, see this table.

Segmenting your Users