Identity associations

Different Exchange endpoints can sometime use different attributes to identify the same customer. Sometimes, the endpoints use different names for the same identifying attribute. Exchange can join the different identifier attributes across different endpoints to develop a rich view of specific customer behavior over time and in different situations.

To overcome differences in how various applications identify customers and customer behaviors, Exchange examines and compares the different identifying attributes that it receives from various endpoints and published events. Over time, Exchange determines which of the different identifiers relate to the same customer. Where possible, Exchange compares identifying information across different endpoints to provide the business user with multiple ways to view and reach the customer.

Exchange requires that endpoint providers specify the attributes that it uses as customer identifier attributes. When you register an application as an Exchange endpoint, you must declare the name and type of all of the attributes that the endpoint might use as identifiers when it publishes events to Exchange. Because identifying information can vary by event, the endpoint must also specify which of the declared identifiers it provides when calls the v1/event API to publish an event.

Customer identity association with the x1Id

Exchange maintains a database to store unique customer identifiers that are discovered when the Exchange Identity Service evaluates customer identity information that is contained in incoming published events. Exchange defines an x1Id for each new unique identifier that is discovered when it evaluates incoming events.

Note: During provisioning, identity data storage can be turned off for an account, upon request.
This ensures that no identity information is stored by Exchange. However, event and audience exchanges still retain the identifier information that they are published with until the end of their lifecycle.

Because unique customer identifiers are associated with a particular individual, each x1Id is also associated with a single customer. However, a single customer can initiate events in different endpoints that publish event data to Exchange. The different endpoints often associate different unique identifiers with the same customer. In this situation, Exchange can sometimes create multiple x1Id entries for the same person.

Each time Exchange receives a published event, it compares customer identity information that is contained in the incoming event to values for all x1Id entries. Any new customer identifiers that Acoustic Exchange can correlate to existing records are added to the x1Id entry. Over time, Exchange accumulates unique customer identifiers for the x1Id entries in the identity store.

Exchange continually collects unique identifiers, and compares them to existing known identifiers. When it finds the same type of identifier, Exchange can join the x1Id records for the individual. The result is a better understanding of how the individual behaves in multiple endpoint applications. In effect, merging x1Id records gives the marketer a view of the customer in multiple commercial contexts.

Note: Exchange limits the number of identifiers that can be associates with an x1Id to 100. Additional identifiers beyond that limit are dropped and Exchange logs an error message.

Suggested unique identifier types

To improve the chances that Exchange can detect and join identifiers across different endpoints, participating applications can use commonly recognized identifiers to identify individuals who interact with each application. Exchange provides a list of suggested unique identifiers.

For ease of implementation and maximum acceptance across different businesses, the attribute that an application submits as a unique identifier must adhere to common internet standards, must be unique among other commonly recognized identifiers, and must be specific to a single individual.

Exchange suggests the following list of identifier types as typical of identifiers with an improved chance of matching across endpoints.

Tip: Always define the identifier name in lower case letters.

• android_advertising_id
• appleid
• appsflyer_user_id
• apsalar_user_id
• coupsmartid
• email
• facebook_id
• facebook_username
• gclid
• ibm_campaign_id
• ibm_da_cookie
• ibm_da_id
• ibm_engage_id
• ibm_interact_id
• ibm_mce_id
• ibm_tealeaf_id
• ios_advertising_id
• instagram_id
• instagram_username
• mediamath_id
• phone
• sms
• sohalo_rewardee_id
• sugarid
• twitterhandle
• twitter_id
• xid
• campaignaudienceid
• turnid
• mmuid
• fbid
• fbusername
• engagecontactid
• tlid
• instagramusername
• instagramid
• appleid
• [vendor]id