We constantly publish updates to our documentation, which may not yet be available in your language. For the most current information, use the English version.
Meta’s Limited Data Use feature helps businesses stay compliant with CCPA. It works by signaling to Meta when users have opted out of having their data disclosed or shared with third parties.
Warning:
Please be advised that Adjust is not providing, and cannot provide, legal advice regarding whether the CCPA applies to you or whether Meta’s processing activities constitute a “sale” under the CCPA. Please consult with your own legal counsel if you have questions about your CCPA compliance obligations.
App level setting
In Adjust, you can control what information is passed to Meta at an app level using the Limited Data Use (CCPA) setting. If you turn this ON, Meta will NOT claim or report any California user data for the app
To allow Meta to process California users’ data, you can turn the setting OFF. However, be aware that by doing this Meta’s processing of California users’ data may be considered a “sale” under the CCPA. As such, if the CCPA applies to you, you need to provide a means for your users to opt-out on an individual basis, such as by collecting the user opted out status on each activity. If you do not do this, you can be held liable in accordance with the CCPA.
Device level setting
With a supported Adjust SDK version, you can use the CCPA setting and also collect and share data from consenting users. This user consent overrides the app-level CCPA setting.
To set up Limited Data Use at a device level, follow the instructions for your platform:
Meta prohibits Adjust (and all other Meta Business Partners) from storing attribution data for longer than 150 days. After that, Adjust continues measuring, but automatically displays all Meta-attributed users under Expired Attributions. This change applies retrospectively, so all previously measured event and session data remains with its user, under the Expired Attributions campaign-level link.
Expired attributions are listed at the campaign level (sublevel 1) for the following sources:
Facebook Installs
Off-Facebook Installs
Instagram Installs
Facebook Messenger Installs
Expired attributions are a display modification and not considered reattributions in Adjust.
Yes! Our SDK supports deep linking on Meta. If you are planning to use this feature, we recommend disabling the Facebook SDK's deferred deep linking functionality for the following reasons:
Since Meta only attributes its own campaigns, its deep linking methods are different from those used by the Adjust SDK, which can target multiple channels. If both are used at the same time it can lead to a conflict.
Only one SDK should handle deferred deep linking. If multiple SDKs handle deferred deep linking, the setup is more difficult to maintain and troubleshoot.
Consult Meta's documentation for more information on deferred deep linking in the Facebook SDK and remove the corresponding code from your application to disable this feature.
Can I use Adjust links in my Facebook posts?
Yes! You can copy and paste your link into any Facebook status update or wall post. Make sure your URL includes all relevant link structure parameters.
Note:
We always attribute clicks on these links via probabilistic matching. Adjust’s API integration with Facebook is only used for measuring mobile app install ads, engagement, and re-engagement campaigns.
Why does Datascape show fewer installs than Meta?
There are a few reasons why you might see install discrepancies:
Different attribution windows. Meta offers multiple attribution models (click- or impression-based) and last engagement attribution windows (7-day, 28-day, etc.). However, Adjust automatically records all installs within a 7-day last-click attribution window. Confirm that the attribution windows on both platforms are the same. You can edit your Meta attribution window or Adjust attribution windows to match.
Different time zones. Confirm that the time zones on both platforms are the same.
Different conversion dates. Meta reports installs and other conversion events as the day of click. Adjust reports installs as the day of install (i.e., first app open) and reports in-app events when they happen. To view your installs by conversion date in Meta, use the Insights API.
Some installs might be attributed to other sources in Adjust. Adjust attributes your installs using engagement data from all of your campaigns. However, Meta self-attributes your installs using their engagement data only. So let’s say someone views a Facebook ad for your app, then clicks a Snapchat ad, and then downloads your app. Adjust will attribute the install to Snapchat and Meta will attribute the install to Facebook.
Why does Adjust show more installs than Meta?
This happens when you link several Meta ad accounts to the same Adjust app. Adjust combines your app's installs across all of your ad accounts.
Why are my re-engagement numbers different in Meta and Adjust?
Adjust and Meta measure re-engagement in different ways. This can sometimes lead to reporting discrepancies.
Adjust counts re-engagement with a user-based model. If an existing user clicks an ad and triggers an in-app session, we check to see if they meet our reattribution criteria. If they do, this engagement is counted as one reattribution. If the criteria are not met, it’s recorded as a session under the user’s existing attribution source. If the user engages with the same ad again and re-opens the app, this is counted as a session - not another reattribution.
By contrast, Meta measures engagement by events. This means that when a user triggers an in-app session, event or install, Meta checks to see if it falls within 28 days of an ad click. If it does, it’s reported under the date of the latest ad engagement. For installs, if no click is available they can take into account impressions that happened in the previous 24 hours.
Finally, numbers can differ due to Meta reporting all conversions under the date of engagement, while Adjust reports events / sessions / installs under the date they were recorded by our SDK. This is most obvious when comparing short time frames; for instance, over 24 hours.