Geo-based Predictive Targeting

Geo-based Predictive Spend and Geo-based Spend History are Experience OS extensions, powered by the data of Mastercard. 

Geo-based predictive targeting enables you to target anonymous users who are likely to engage and spend on various preset categories (for example, Children’s Apparel, Jewelry, Beauty Products, and many more), based on proprietary predictive models and localized consumer spending data provisioned by Mastercard.

See the list of categories for targeting.

Geographical coverage

Geo-based predictive targeting is currently available for the following countries:

When you select a geo-based category to target in an experience, the number and list of countries where targeting is available appears.  

Click the link to see the specific countries:

How it works

The extension evaluates the user’s geographical location in real time, based on the user’s IP address and the geographies it maps to. It then evaluates whether users in that geographical location are considered highly likely to spend on any of the given categories based on a propensity scores that are calculated by Mastercard predictive models and datasets. If the propensity score for the user’s location is high for the chosen category, the user is targeted and receives one of the variations in the experience; if the score is low, the user is not targeted.

You don't need to specify any location to set up a targeted experience leveraging Geo-based Predictive Targeting conditions. The geo-based evaluation is done automatically, referencing all relevant data points to determine whether a user should be targeted.

Targeting conditions

The two extensions map to the two available condition types, which can be used separately or together, and work similarly to every other available targeting condition:

• Locations with high probability to spend: Targets users based on the geo-based predictive models.  
• Locations with high spend history: Targets users based on the historical consumer behavior observed in their zip code.

Learn more about Targeting Conditions.

Use case examples

Targeted experiences and content personalization are especially relevant to category promotions or hero banner optimization on the homepage, for example:

• E-commerce:
  Offer high-end brands for users in zip codes with high scores for luxury retail, and money-savers for users in areas with a preference for discount stores. 
• Financial institutions:
  Promote different credit cards (travel, premium) to users in areas with high scores for respective categories (international air travel, high net worth consumers). 
• Other business areas: 
  Offer tools or decor items to high-spend areas for home improvement, beauty subscriptions to high-spend areas for cosmetics, and so on (pet care, new parents, consumer electronics, and more).

Best practices

The best use of Geo-based Predictive Targeting is to target users for whom you haven't yet collected any behavioral data. This includes anonymous or first-time visitors who would otherwise default to a non-personalized, sub-optimal experience.

In practice, this means that the best way to incorporate Geo-based Predictive Targeting into a campaign is in targeted experiences, as follows:

• The top-most experiences are aimed at known users, based on behavioral targeting conditions or audience membership. 
• Next, create Geo-based Predictive Targeting experiences.
• At the bottom, place the default experience.

See this arrangement illustrated in the following screen capture, where tech enthusiasts in an audience are targeted first; next, users are targeted because they have an affinity to tech; the next tier of targeting is based on users being in a location with high tech spending, and then, finally, the default experience is served to visitors who do not match any of these criteria:

With this setup, the number of users who receive a personalized, optimal experience is maximized. Of the remaining users, the number receiving a geo-based experience is then maximized, and only users for whom there is no data at all default to a generic, non-personalized experience.

Predictive targeting category lists

Prerequisites and limitations

• To get Geo-based Predictive Spend and Geo-based Spend History, contact your account manager.
• The list of available categories might vary between accounts, based on the account’s vertical classification (E-commerce, Financial Institutions, Quick Service Restaurants, or Consumer Packaged Goods). 
• Geo-based predictive targeting currently works only for select geographies throughout the United States, Europe, Latin America, and APAC. In the near future, availability will be expanded to support additional countries, including: