Custom Targeting

Overview

Custom Targeting targets a flight or ad to a placement depending on the pairing of key/value properties.

In other words, with Custom Targeting you set up rules to target certain combinations of key/values, and then if that value for the given key is passed in the request, the ad will show only if it matches the rules you defined.

For example, if you've created a campaign to target instances where Age > 21, if you send Age = 25 in the request, the ad would be eligible to show (but not if you sent Age = 20).

To create rules, write a query using the Zerkel Language on the flight's settings page, or when using the Create Flight endpoint, add it in the CustomTargeting field.

You'll then pass the Key/Value pair in the properties object of the Decision API request.

Kevel offers two types of Custom Targeting:

  • Custom key/value pairs sent by you
  • Reserved Keys - Targeting using databases and lookups integrated with Kevel, such as device type, OS, location, url strings, other useragent info, and more

πŸ“˜

Unlike keywords, custom targeting values are not included in Kevel reporting

Custom Targeting Parameters

A key is a string, such as Age. A value can be a string, an integer (such as 1234), or an array.

Use JSON syntax for passing in values. These key/value pairs are passed by Decision API request or by modifying the ad code.

🚧

The key cannot contain -.

Custom Targeting vs. Keywords

Like keywords, custom targeting can be used to target ads based on conditions passed into the ad request. However, there are some significant differences:

Custom TargetingKeywords
ReportingMatched custom targeting criteria don't appearKeywords that match both flight settings and ad request appear
Required in RequestCan include Reserved Keys that are not required on the requestMust be passed in specifically in a request for a matched ad to serve
LogicCan perform comparisons, such as age >= 18Use a binary system - either the keyword is present in the request or it isn't
IntegrationsCan be integrated with other Kevel targeting features, like ContentDB, UserDB, keywords, etcStandalone feature
PropertiesUses key/value pairsDo not have subsets or properties