Flurry is an analytics system, dedicated to applications. It providers information about general use of the application, as well as data regarding actions within the app.
You can find the system at https://dev.flurry.com
Note that Flurry offers several services. Applicaster has implemented their analytics service.
When logging into the system, you can see all your applications in the Home page. In the table, you will see the following data for each app:
Unique Users - Unique New Users for the app
Active Users - Unique active users for the app
Sessions - Total number of sessions that happened in the app
Each platfrom has a seperate account, so if you have both an iOS application and an Android application, you will have 2 accounts. Clicking on the app will open the analytics section of the system.
Flurry displays the data in several tabs:
Usage: General information on the use of the application, including number of sessions and active users.
Audience: Information about the users of the application, including age and gender.
Events: Data about specific actions within the app
Technical: Technical details about the application, including what devices and operating systems used.
On the left there is a menu, with the main sections of the system. Clicking on each of these menues will open the sub-menu for that section. Clicking on an parameter in the sub menu will display the data for the parameter
At the top of the screen you can select filters for the data displayed:
Users: You can filter the data by Segments, Campaigns or Channels. Segments are pre-defined sub-sets of users. You can create sub-sets based on, for example, age or gender of the users.
For further information, see http://support.flurry.com/index.php?title=Analytics/Overview/Lexicon/Segmentation
Version: filtering by versions of the application.
Times: Filtering by dates. The default is "Across All Times", which displays data since the first time the app reported to Flurry.
Selecting Last Week will display data for the past 7 days, while selecting Last month will display data for the past 30 days.
In addition, in most screens you will have the following options:
* Clicking on the Explain button at the top of each section will display an explanation of the data displayed.
* For the Chart and the Table, you can set the zoom level you need: Hours, Days, Weeks or Months.
How will this impact the data?
For example, If you wish to look at the number of Active Users for the past month, setting the zoom to Days will show you the number of unique active users for each day in the past month. Selecting Weeks, it will display the number of unique active users per week in the time period selected/
* You can export the data easily by clicking on Download CSV.
Some basic terms used in the system:
Session: one use of the application by an end user. IT starts when the application is launched and ends when the application is terminated. If a user does not terminate the app, the session will end after 10 minutes of the app running in the background.
New User: A user that downloaded the application and opened it at least once ( note that flurry starts collecting data when the app is opened, so if a user downloaded the app but did not open it, Flurry will not count that user)
Active User: A user that had at least one session with the app. The data for Active User is unique, i.e, for each period of time, each user is counted only once, regardless of the number of time he opened the application.
Event: An action in the application. For example: Clickng on the Play button. In addition to a pre-defined set of events, it is also possible to define events unique to your application.
Note: The data in Flurry is in Pacific Standard Time.
The dashboard area allows you to set up all the data you need in one screen.
To create a dashboard, click on Create Dashboard. You can either use pre-made dashboard templates, or create your own dashboard.
For more information on creating a dashboard, see http://support.flurry.com/index.php?title=Analytics/Overview/Lexicon/CustomDashboards
Clicking on the tab will open a menu with all the data breakdown avialable in this section.
For each parameter, the data is displayed in 3 formats: Chart, Table and Map.
A new user is a user that downloaded the application and opened it for the first time.
The users are identified by a unique identifier. For iOS devices, this is the Mac address of the device. For Android, this is the Android ID.
Note that the number of New Users does not correlate to the number of downloads as provided by Apple and Google. This is because while Flurry uses the device unique identifier to track users, the app stores use the User ID. Following are a few cases where the Flurry number of New Users and the number of downloads will differ:
1. A user downloaded the app but did not open it. It will be counted by Apple/Google, but not by Flurry.
2. A user used the app off-line. It will be counted by Apple/Google, but will only be counted by Flurry when connecting to the internet.
3. Jail-broken device. In such a case, Apple will not count it in the downloads numbers, but Flurry will count it as a new user.
4. One user was used to download the app to several devices. In such a case, it will be counted as one download, but several New Users.
A user that opened the application.
Session is a single use of the application by an end user. A session starts when the application is started, and ends when the application is terminated. If the application was not terminated, the session will end after 10 minutes of inactivity in the background.
A length of the session, from the moment it starts to the moment it ends.
note that in some cases, for example, if the application crashes, there will be no end time for the session. In such a case, the length of the session will be unknown. Such cases will not be considered when calculating the average and median values of session length.
Frequency of Use
How many sessions unique users had in a specific period of time.
Rolling Retention is defined as the ratio of number of users who use the app after day N, to the number of users who could have used the app on day N (i.e sum of new users upto day N). Its basically a measure of how many users out of the total new users use the app, after day N.
Defined as users who use the app at-least once in the last 7 days. So, for last 7 days, all New Users are considered as recent users and hence the retention for last 7 days is 100%. If the app had 10,000 new users in the month of Jan, and out of that consider that 4000 users used the app at-least once in the last 7 days, the retention for the month of Jan would be 40%.
A user that started the app for the first time is considered a New User. If that user started the app again within a week from the first activation, he will no longer be considered a New User. This section shows how long after first activation, a user remains active.
The number of users for the top most popular 50 versions of the app
Measures the times users viewed pre-defined pages in the app.
This tab shows the data based on information about the users, including gender, age and geographic location.
In cases where the app does not collect this information, Flurry will use data collected about other apps in the same category (Ie: education, entertainment, sports etc).
Interests - My Apps
In case you have more than one app, this will show how many of the users of the current app also installed your other apps.
Interests - Categories
What other categories of applications ( that use Flurry) your users have.
Flurry uses all the data it collects to define "Persona", ie, a type of user. An example of a Persona can be "Sports fan".
This section displays analysis of the users of your apps, by Persona.
Segmentation of the users, based on Age and Gender. If your application does not collect this data, flurry uses its benchmark estimates, collected from its other apps that do collect this information.
Segmentation of the users based on their location.
Segmentation of the users based on the language of the device.
This tab displays information about events taking place within the app. An event is a pre-defined action in the app, including playing a video, clicking on a category etc.
Please find in the attached document a list of all the available custom events for the modular app version 2.5.1. If you are not sure which modular version you have, please contact Support.
In this area you can find data for each of the events defined for your application.
For each event, you can see general data including average duration and total number of occurrences, as well as segmentation by parameters, logs and paths. See further here for in-depth explanation of the Events section of the system.
Shows the paths the users take in the application. A triangle at the top of each event in the chart indicates that there are more events. Clicking on the event will open the events that follow it.
Allows you to create a scenario of steps in the app, and see if users follow this scenario.
For further information about using this option, see http://support.flurry.com/index.php?title=Analytics/Overview/Lexicon/FunnelAnalysis
Displays a chronological list of all the sessions in the event. For each session, it shows the events that took place.
Events and their data
In the main screen you will see all the events defined for your account.
For each event, clicking on it will open the details about the event. For each event, there can be 4 tabs:
Event Statistics: General data about the event. In this tab, you will always have the following data:
- Events Per Session -
- Unique Users Performing Event - Number of unique users who did the action.
- Total Event - the total number of times this event happened.
- Avg. Event Per Session - average number of times this event happens per session. The calculation is total number of events for a period of time / total number of sessions for a period of time
User Paths: the path the users took in the application, from the event you are looking at onwards.
Event Parameters (Optional): These are the pre-defined parameters for the event. Parameters are additional information about the event.
This is best explained using an example: The event "Play VOD Item" tracks how many people watch vod items. Every time a user clicks on an item and opens it, it is counted as a play. When this happens, there's additional information about the event: What is the name of the item? What is the name of the show it belongs to? was it a free item or not? All of these are parameters of the event.
Note that some events will have parameters, while others wil not. The parameters differ from one event to another.
When opening the Event Parameters tab, the drop-down menu on the left displays all the available parameters for that event.
Selecting a parameter from the drop down will display the data for that parameter. For each parameter, the data is displayed in a pie chart and in a list. For each event parameters, it shows the count (number of times it occurred , and it's % of the total number of time the event parameter happened. In the following example, these are parameters for the event Play Channel, which measures how many times a live channel was views. In this case, the Paid feed (ie, the feed that is defined as not being free), was watched 244,258 times in the selected period, which is 86.4% of the total number of times the event happened, while the Free feed was watched 38,529 times, which amounts to 13.6% of the total number of times the live feed was watched.
Event Log : Displays a log of all the sessions where the specific event happened. The Date/Time displayed is the data/time of your computer. Note that you can download the data, but it will only download the current page you are in.
Event Duration (Optional): This tab will be displayed only if the event is a timed event (ie: event that has a duration). It will show the following data:
- Avg. Event Duration - The average duration of the event in the time period selected. For the events "Play VOD Item" and "Play Channel", this is the average viewing time.
Calculation: Total duration of the event / number of times the event occurred.
- Event Duration Per Session: the length of the event by session. This is an indicator of
Calculation: Total duration of the event/total number of sessions
- Avg. Event Duration per User: the avarage duration of the event per user.
Calculation: Total duration of the event / total unique number of users who carried out this event
How to find:
1. Total times VOD Items were watched
For VOD Items: The event "Play VOD Items", Event statistics tab, Total Events
For Live Channels: The event "Play Channel", Event statistics tab, Total Events
2. Average viewing time:
For VOD Items: The event "Play VOD Items", Event duration tab, Total Events
For Live Channels: The event "Play Channel", Event duration tab, Total Events
3. How many times each VOD item was watched
4. How many times each live channel was watched
This tab displays information about the devices used .
The number of sessions on each device type.
The number of session by internet providers and wireless networks.
All Firmware Versions
The number of sessions for each version of operating system.
The error logs the application sent
The data is shown in Pacific Standard Time.