Cohort analytics via Data Locker

At a glance: Cohort analytics in Data Locker contains aggregated data of all elements in the advertising hierarchy. Reports contain data of users attributing in the previous 365 days.This report is available to advertisers who have raw data or Cohort API, or both as part of their subscription. It means that a Data Locker subscription isn't required for this report.

Cohort analytics in Data Locker

In some cases, attribution raw data in AppsFlyer is restricted due to media source data sharing policies or your privacy-preserving policies. The restriction includes all attribution-related fields like campaign and adset. Cohort reports provide you with aggregated data that can be used as an alternative to raw data. 

See attached example report below.

Enable Data Locker

The report is available to Data Locker, Cohort API, and raw data subscribers. To enable the report in Data Locker, complete one of the following procedures. 

Enable Data Locker
Do you currently get data via Data Locker?  Procedure

 

Yes 

AppsFlyerAdmin_us-en.pngTo add the Cohort reports to Data Locker:

  1. In AppsFlyer, go to Integration > Data Locker.
  2. Select all the reports in the Cohort analytics section. 
  3. Click Save configuration
    The reports will be available on the next day by 10:00 UTC. 
No

AppsFlyerAdmin_us-en.pngTo set up Data Locker:

  1. Complete the first-time Data Locker ( advertiser | partner ) setup.
  2. Add the Cohort reports to Data Locker:
    1. In AppsFlyer, go to Integration > Data Locker.
    2. Select all the reports in the Cohort analytics section. 
    3. Click Save configuration
      The reports will be available on the next day by 10:00 UTC.

Related reading: Differences between Cohort flavors | Cohort dashboard | Cohort API

You can load the data into your BI systems and use them as part of your campaign performance and optimization processes.

Metrics available

The metrics available are revenue, unique users performing an event and the number of event occurrences. To calculate cost-related metrics like ROI, and ROAS you need both revenue and cost metrics. Revenue metrics are in Cohort, and cost metrics are provided by Xpend Cost ETL.

Facts

Reports available

The following reports are available for download. The reports types are described in more detail in Cohort dashboard.

  • User acquisition: attributed to a user acquisition media source including LTV occurring during re-engagement windows.
  • Retargeting: attributed to a retargeting media source for events occurring:
    • During a re-engagement window
    • As a result of re-attributions
  • Unified: display data under the last touch media source bringing the user according to AppsFlyer double attribution rules. Meaning that LTV occurring during re-engagements displays under the retargeting media source and not under the UA media source. 
Reporting period Users converting during the previous 365 days. In other words, each day the reports include the user converting during the previous 365 days. 
Report structure The structure of each report is the same. See report structure.
Report delivery  See Enable Cohort reports in Data Locker.
Data freshness
  • Daily.
  • Metrics are calculated using data avaialble at midnight UTC. In other words, at the end of the day. 
  • Reports are available in Data Locker by 10:00 UTC in the h=23 folder. It means that Monday's data is available on Tuesday at 10:00 UTC in the Monday h=23 folder. 
Impact of partner retention policies

Consider that some partners implement a data retention policy. In this case, events occurring after the end of the retention period are disregarded in the Cohort reports.

Example: SRN A has a data retention policy of 180 days. User events until day 180 are attributed to SRN A. Events that occur after 180 days are disregarded.

Note: The events display in the Overview dashboard as organic.

Cohort analytics facts

Report structure

The report is comprised of dimensions and metrics.

The format of the fields is as follows:

  • Dimensions: String. The maximum length of the string is dynamic and in most instances depends on how you populate the elements of the advertising hierarchy
  • Metrics: Number. Note! The selected_currency field format is string.
Dimensions

Field name 

Description

app_id

--

media_source

--

conversion_type

Possible values: install, re-engagement, re-attribution

attributed_touch_type

Possible values: click, impressions, pre-install, unknown, tv, null

days_post_attribution

  • The number of days elapsed since the attribution event. 
  • Tip! Use this to calculate retention and KPI days

event_date 

  • The date a user performs a given event.
  • Format: yyyy-mm-dd
  • Example: The date a user performed a given in-app event. In the case of a conversion.
  • Note! If the event_name is af_conversion, then the event date and conversion date will be the same. 

conversion_date

  • The date the conversion occurred
  • Format: yyyy-mm-dd
  • Example: The install date

event_name

Identifies the event. Some event names have a specific meaning while others relate to in-app events set by the advertiser in the app. 

 

event_name

What did the user do

af_conversion User converted. Use conversion_type to identify if it is an install, re-engagement, or re-attribution.
af_session Opened the app
af_uninstall Uninstalled the app 
Advertiser defined in-app event Performed an in-app event in the app

campaign

Campaign hierarchy

Consider: Campaign name changes not supported. Consequently, multiple names can be associated with a given campaign ID. 

campaign_id

Campaign hierarchy

adset

Campaign hierarchy

adset_id

Campaign hierarchy

ad

Campaign hierarchy

ad_id

Campaign hierarchy

channel

Campaign hierarchy.

[Updated October 27, 2021] At present, Facebook doesn't populate channel in data provided via the Google Install Referrer mechanism.

site_id

Campaign hierarchy

is_primary_ attribution

Use to identify and deduplicate retargeting data.

geo

The ISO country code derived from the user IP address.

agency

  • Agency transparency supported.
  • In the case of non-transparent agencies, there can be multiple rows containing the same campaign name if the advertiser and the agency both run campaigns with the same name. Don't worry. The rows aren't duplicated.

install_app_store

Android apps only: The Android store from where the app was downloaded. Populated by advertisers implementing multi-store Android attribution. If blank, it means Google Play Store. 

keywords

Word(s) used in the user's online search. As reported by the ad network

keyword_id

[Field available starting April 27, 2022] Keyword ID returned by the ad network.

Metrics

Field name

Description Format

unique_users

Number of unique users performing the event

Number

revenue_usd

  • Amount of revenue in USD. For example, $100.56 is reflected as 100.56 
  • Maximum of 2 places after the decimal point
Number

event_count

Number of event occurrences

Number

selected_currency

3-letter currency code (USD, EUR) set by you in the app settings. Format ISO-4217. This is the same currency used to display revenue in Cohort in the user interface. 

String

revenue_selected_currency

  • Amount of revenue in the selected currency. For example, if the selected currency is EUR, then €1234.56 is reflected as 1234.56
  • Maximum of 2 places after the decimal point
Number

BI developer considerations

Scope of data in the report

The reports contain user attributions, installs, re-attributions, and re-engagements, and their related in-app events.

If you use the unified view in your data loading process, you must use logic to split the data between campaign types, which means, user attribution (installs) and retargeting (re-engagements). To do so, use the conversion_type and is_primary_attribution field. See Double attribution of retargeting events. 

Field-level considerations

  • Use post-attribution days to enable easy calculation of retention metrics.
  • Calculating unique users using campaign name and campaign ID dimensions: If you can disregard campaign name granularity, you can sum unique count on campaign ID and the metrics are correct. 
  • You can aggregate the data using the campaign hierarchy fields. 
  • Revenue in USD is calculated using the exchange rate on the event day. 
  • Ad revenue events are included where available but can update with a lag of up to 3 days.

General considerations

The data of all apps is provided in a single file. Use the App ID field to segregate data per app or set Data Locker to segregate by app.

Calculating the number of unique users per cohort 

To calculate the number of unique users per cohort by conversion type, for example, the number of users installing the app you must filter the data so that conversion_type=install and event_name=af_conversion. Sum the event_count column. You can add dimensions by including additional columns. To calculate the number of installs per day, group by conversion_date.

This is illustrated by the SQL statement and Excel visual that follows. 

SQL statement


SELECT   app_id,
         conversion_date,
         media_source,
         event_name,
         conversion_type,
         Sum(event_count) AS total
FROM     'REPLACE_TABLE_NAME'
WHERE    app_id='REPLACE_APP_ID'
AND      conversion_type='install'
AND      event_name='af_conversion'
GROUP BY 1, 2, 3, 4, 5

Excel example

CohortDL.png

Additional information

Traits and limitations

Trait
Trait  
Cost data Not available. Use Cost ETL.
Campaign name changes Not supported. 
Data freshness Daily UTC. Reports are in the Data Locker h=23 folder of the reporting day by 10:00 UTC of the next day. 
Ad revenue Available
Currency  USD and app-specific currency are both avaialble per row
Organic data Available
Days post-conversion (install, re-attribution, re-engagement)

365 days.

 

Agency transparency
  • Supported 
  • Twitter and Facebook data is always transparent
App segregation Supported
SKAN data Not included. Meaning, the data provided by iOS postbacks.
Reinstalls Not included.