Cross-device Targeting

Cross-device targeting is the use of data to expand your custom audience and reach the same consumer on multiple devices. This tactic increases the amount of biddable inventory for your audience in programmatic auctions, expands your reporting to see all tactics that contributed to a conversion, and creates a more holistic advertising campaign for your end consumers.


What is Cross-device Targeting

For marketers looking to reach individuals on multiple devices in a comprehensive advertising campaign, the use of cross-device targeting to identify and control the end-consumer's experience can be a great solution for managing the targeting, frequency capping, and impact of a campaign.

Cross-device targeting is powered by our data partners who provide identity graphs, which can be thought of as a collection of user profiles, that match a consumer in a Choozle custom audience to those consumer’s additional devices. 

Choozle's cross-device targeting partner, Identity Alliance, uses machine learning algorithms to contextualize unstructured data points across devices, people, and households – achieving both broad coverage and a clear view of your target audience’s behavior.  

Leveraging the use of Identity Alliance, users can tap into the power of four cross-device vendors into a single and extensive device graph. Composed of Adbrain, LiveRamp IdentityLink, Oracle (formerly Crosswise), and Tapad device graphs, the system will decide which device graph to use at the time of the impressions is served based on the targeting parameters. This means one vendor might offer better-connected TV coverage, another could have a greater global reach, but the user won’t need to decide which cross-device vendor to use during ad group setup.  

Enabling cross-device targeting will incur an additional CPM between $.50 - $1.20 depending on the provider used, but the default provider will incur a $1 CPM. 

Deterministic vs. Probabilistic Cross-Device Methods

An overview of the benefits and misconceptions of both data methods and a clear solution to getting the most value from leveraging both to reach your marketing goals.

There are two popular approaches to mapping a customer profile across different devices:

Deterministic Method

The deterministic method uses data, such as a login or a persistent customer ID to map devices together. For example, if a visitor uses the same login from a mobile app and on desktop, this approach will recognize both devices belong to the same visitor with a high level of certainty.

Probabilistic Method

The probabilistic method applies algorithms to anonymous data sets, such as location data and time of day, to build a statistical model that connects devices with high levels of probability. For example, if two different devices connect to the same home Wi-Fi location every morning and evening and are often close in physical proximity to each other throughout the day, it can be inferred that those devices belong to the same user.

The differences between the two approaches come down to a tradeoff between scale and accuracy. See the chart below for a better understanding of how scale and accuracy apply to each method.

  Deterministic Probabilistic


Scale is limited as a user typically must have logged in on multiple devices which don’t always happen.

For example, 47% of Facebook users are only active on one device (mobile) and therefore also rely on probabilistic mappings.

Probabilistic data has an infinite scale as you can combine any number of data sets to find a user on multiple devices.

In theory, deterministic should be highly accurate.

In reality, accuracy rates can be highly variable depending on the source as it is hard to know if a deterministic match is real.

For example, family members may all use the same login for a Netflix account.

Probabilistic methods report back the confidence levels of a match allowing the user to adjust according to their use case.

In some instances, probabilistic can outperform deterministic data sets.

Choozle's cross-device targeting partners use a mix of both deterministic data and probabilistic data. Adbrain uses deterministic data to help train their algorithms to help make our probabilistic data even more accurate. With a combination of both deterministic data and probabilistic data you no longer need to choose between accuracy (which may be highly misleading with deterministic!) and scale but you can leverage both to take your targeting and measurement to the next level.

Leveraging Cross-device Targeting

Cross-device targeting is enabled at the campaign-level and adds an additional CPM between $.50 - $1.20 depending on the provider used, but will typically average around $.75 to all ad groups with a custom audience. You will also receive expanded conversion tracking across all ad groups to see all tactics that contributed to a conversion. For example, if a user was served an impression on their mobile device but converted on their personal laptop, the system would be able to attribute that conversion across the variable devices. 

Apply cross-device targeting to campaigns with ad groups that contain smaller custom audiences (specific third-party data segments, retargeting, CRM audiences, or geolocations). If you have ad groups targeting less qualified audiences (demographic, broad interest, or prospecting data segments), consider building these ad groups in a different campaign that does not have cross-device targeting enabled.

Cross-device targeting's conversion reporting can be leveraged in campaigns targeting ticket sales, hotel bookings, retail purchases, or geolocations where the consumer may see an ad on their mobile device but convert on their desktop computer later. Cross-device conversion reporting and attribution will be automatically enabled for all campaigns with conversions, even if cross-device targeting is not enabled. This is applied at no additional cost! 

Cross-device targeting provides expanded conversion tracking on ad groups with a custom audience. By enabling cross-device targeting, you will receive reporting for all ad groups that have contributed to a conversion. Conversion tracking is powered by the Choozle Smart Container Tag, so you will want to ensure that the advertiser has placed the Smart Container Tag before enabling conversion tracking for your cross-device reporting.

In addition, for ad groups that are targeting Connected TV devices, Household Extension will automatically be enabled.  Household Extension enables you to expand targeting to every device on a user's IP address. This broadens campaign reach while still delivering targeted impressions across multiple devices in one location, which allows you to target the same person on multiple devices or new people in the same location as your original retargeted user. There is no additional fee for Household Extension.

If you are using third-party reporting, check if your provider will account for cross-device conversions. Once cross-device targeting is enabled, your conversion reporting will take cross-device conversions into account and you may see a reporting discrepancy with your third-party reporting partner.

Implementing Cross-device Targeting

Begin by going to the campaign tool and selecting the + sign to create a new campaign. Cross-device targeting can be applied by toggling on the Cross-device Conversion button below the campaign flight dates in the General Settings of the campaign.

Once the new campaign is saved, your setting for cross-device targeting will be set for the campaign. 

Please note that for campaigns where cross-device targeting is enabled, the additional $.50 - $1.20 CPM for cross-device targeting will automatically be applied to all ad groups with a custom audience. You will see a notification next to the custom audience selection within the Targeting section of your ad group setup.

To understand the impact made by having cross-device targeting enabled, you can review the performance of the data segment listed as "AdBrain Cross Device" in the Brand Data tab of your detailed report. 


Why Use Cross-device Targeting

Cross-device targeting is the idea that a marketer may want to provide a curated or holistic experience for the targeted audience across multiple devices. Cross-device targeting can only be done in ad groups of campaigns that utilize some format of data targeting (first or third-party data).

With cross-device targeting, you can:

  • Use cross-device data and identity graphs to identify users across display, mobile, and television devices.
  • Use the identification of users across devices to expand the audience of your ad group (retargeting and behavioral targeting).
  • With users mapped across devices, create a more complete frequency cap across ad groups and devices targeted.
  • Understand a more complete view of a conversion's acquisition flow and attribution.

Additional Cross-device Targeting Providers

Additional cross-device providers are available via our Supported Solutions. To enable cross-device targeting with a provider not available in-platform, please reach out to your member of the Client Experience Team with the name of the advertiser account, campaign, and ad group in which you would like cross-device targeting applied, as well as which identity resolution provider you would like to have applied.

Leveraging additional identity resolution providers made available through Supported Services is well suited for campaigns looking to:

  • Reach consumers outside of the United States
  • Reach consumers on Connected TV 
  • Reach consumers at a larger scale
Identity Resolution Partner Identity Graph Size Accuracy Device Types
Regional Coverage
LiveRamp Identity Link more than 1 billion devices

100% accuracy due to deterministic matching methodology

desktop, mobile, tablet, and offline data United States
Crosswise (Oracle Data Cloud) Device Graph™ 900 million devices 99.00%

New Zealand
United Kingdom
United States
Tapad Device Graph™ 2.16 billion devices 91.20% Desktop
Connected TV
Gaming console
United Kingdom
United States