People-based analysis

At a glance: People Based Analysis provides a complete picture of user acquisition and conversion flows across platforms, devices, and marketing channels. You can see the full and realistic performance picture of your brand.

People-based analysis page

To open the People-Based analysis page:

  1. In Appsflyer, go to My Apps > View brand bundles.
    The list of Brand bundles displays.
  2. In the actions menu, select People-Based Analysis

People-based analytics table

The core of this page is the people based analysis table.

The table consolidates data from web and mobile app activity. This data is the center of your cross-platform marketing analysis. It contains all the data that you need to make educated decisions about brand-specific marketing strategies.

The data enables you to see core metrics for web and mobile apps side by side. In addition to traditional dimensions such as media sources and GEO, the data also includes new dimensions such as device types, operating systems, and platforms.

Time for data to appear in the table

When data about visits to your site and events sent from the SDK reaches out servers, people-based attribution runs complex processes to analyze data and match data points. Since this process takes a while and has to be accurate and reliable, it can take up to 24 hours for data to appear in the table. 


Here are some examples of how you can use the data:

  1. Your goal is to switch users from web to mobile. You can analyze the properties of users who switch to mobile and target such users
  2. You can segment media sources into those that drive web users and those that drive mobile users
  3. You can see how a single user behaves in both web and mobile

Top performing networks

The People Based Analytics page features a chart that shows the top performing networks. It gives you a high-level view that helps you understand what top networks you should focus on.



The chart is affected by filtering and date range selection.

The chart itself is cumulative as the graph shows the total number of conversions. To see the number of conversions for each network, hover over any of the dates in the chart.

To compare specific networks, you can either filter or simply remove networks from the chart. To remove networks from the chart, simply click on them.


Below is an example of comparing Google Adwords and Facebook:


Filtering and grouping

The people based analysis table allows you to filter and group data to generate a view that meets your analysis needs.

Below are the different columns that appear in the people based analysis table:

Column Description Notes
Media Source The media source running the ad Can be drilled down to the level of campaign, adset or ad
Clicks The number of clicks leading to app engagement (install or open), clicks on third-party ads or clicks leading to the website. Includes paid and non-paid clicks on the web
Installs When a user installs and then opens the app for the first time.  
Web Conversions Client defined web events that are attributed by the system Defined independently by the client via the web SDK
Installs / Click % The ratio between app installs and total clicks  
App Opens Count of installed app opens which are not first-time launch Any app open, excluding first-time launch (install) and manually launching the app
Clicks / Opens % The ratio between clicks and app opens  
Web Sessions Count of distinct user sessions on the website Measured via the web SDK
Revenue Total revenue generated via in-app purchases or web conversions. In USD by default
Top Events Count and total revenue generated by unique in-app events or web conversions By default displays top three, you can change and choose which events to display


You can filter the results based on media sources, geographic locations, and date ranges.

Media sources filter

  • Use the menu bar at the top of the page to select one or more media sources, for which you want to observe data.
  • By default, all media sources are selected.


The table only displays media sources from which AppsFlyer receives a click, impression or install, during the selected date range.

If some media sources don't appear in the list:

  • Try to change the date range
  • Try to change any applied filters

Geo filter

  • Use the Geo drop down to select one or more specific countries.
  • If no country filter is applied, all countries appear in the table.

Date range selector

  • Use the date range selector to refine your search to specific dates. You can select your own custom date range or use one of the preset date range buttons.
  • By default, the last 7 days are selected.


The grouping options allow you to further categorize the results of your filters and break all metrics results based on new dimensions. Under the Group By filter, click the dropdown to view your grouping options:

Media source/campaign

Used for comparison of acquisition sources. For example, let's assume you run the same targeting campaign in both Google and Facebook. In this case, you can view which source is more successful when a campaign is directed towards the same target audience.

In addition, by clicking on the name of the campaign, you can drill-down to ad set and then to ad level to view in-depth campaign data. This is only available to those media sources providing this information.

The value of the campaign can come from one of the two parameters:

  • The utm_campaign parameter.
  • The c parameter.

If the URL that takes the user to the site doesn't have a utm_campaign or c parameter, there is no way to determine the campaign. In such cases the campaign value is None and there is no option to drill down to ad set or ad level.

The media source value can come from either the pid or the utm_source parameter if any of these are appended to the URL. 

The pid and c parameters take precedence over utm_source and utm_campaign. If both pid and utm_source appear on the URL, the media source value is taken from pid. If both c and utm_campaign appear on the URL, the campaign value is taken from the c parameter.

Also, it's not possible to mix and match the parameters. pid and c go together and utm parameters go together. If you use pid for media source and utm_campaign for the campaign, the media source is mapped but the campaign name is not logged. Similarly, if you use utm_source for the media source and c for the campaign, the media source is mapped but the campaign name is not logged.


Allows you to see the traffic and engagement trends by country. This has a direct impact on how much you are willing to invest per country.

When you select the Geo option the table displays a breakdown of data by country.


Shows a combination of operating systems and platform. for example, Android / Web, Windows / Web, iOS / App.

This can help you analyze the performance of campaigns that target both web and mobile app users.

Device type

Breaks all metrics based on the high-level device type. This grouping Holds five basic device types: smartphone, desktop, tablet, TV, other (unrecognized devices).

Selecting groupings

You can choose one group or more by dimensions. Doing so adds the dimension column to the table with a break down of table metrics for the selected groupings.

By default, Media Source / Campaign is displayed.


You want to see clicks by Media Source / Campaign and Device Type. You apply the relevant groupings.

The table then shows you a break down of the selected metrics per grouping.

For example, you can see 150 clicks for media source A under campaign B. The device type dimension further breaks down the clicks. You see that there are 67 clicks on smartphone, 22 clicks on tablet and 61 on desktop.

Now you want to see a break down of clicks by environment. You add the environment grouping. Now the table shows the following:

  • 67 clicks on smartphone - 15 Android / Web, 12 iOS / Web, 25 Andriod / App, 15 iOS / App
  • 22 clicks on tablet - 10 Android / Web, 12 iOS / Web
  • 61 clicks on desktop - 35 Desktop / Windows, 25 Desktop / MacOS

Brand Bundle Selector 

The people based page shows data on the brand bundle level. Each bundle groups web and mobile apps (one iOS and one Android) that operate under a specific brand. The bundle selector enables you to switch between different bundles in your account.

You can select bundles according to your permissions to view and edit certain apps.

To learn how to create and edit bundles click here.

Additional functionality

Adding and removing columns

You can add and remove metric columns from the table. To do so, click on the cogwheel cogwheel.png icon in the upper right-hand side of the table. Check or uncheck the boxes next the column name to add or remove the column from the table.

Events filter

You can select one or more (multi-select) events to view in the table. By default, the table shows the top three performing Events (both in-app and web).

To add or remove events from the table:

  1. Click on the cogwheel cogwheel.png
  2. Scroll down to the list of events
  3. Check or uncheck the box next to the event name
  4. Optional - Check or uncheck the box next to Event Count or Revenue to add or remove event count and revenue metrics from the event column
  5. Click Save

There is no limit to the number of events that can be displayed in the table.

Adding and removing columns (metrics and events) allows you to customize the display according to your analysis requirements.


  • You wish to focus only on mobile app data. To do so, you remove the Web Sessions and Web Conversions columns. The generated view is a clean and noise-free dataset that focuses on mobile app data.
  • You wish to compare two purchase events. To do so, you remove all columns except the columns related to these two events. The view then allows you to compare the events in terms of their count and revenue. You can then add more columns to enrich the view. You can add the device type or environment columns. These columns can show you the breakdown of event count and revenue per device type or environmnet.


Changes to the table view are reflected in exported CSV files.

Export CSV

To export the current display table in CSV format, click on the Export CSV button. Exported data reflects the current filters, grouping, and applied selectors.



Was this article helpful?