Identity Resolution Moves into the Contact Center

Identity resolution can also be used to determine whether additional action is warranted. If, for example, a bank understands that the customer wants to make a deposit, then it can make the caller jump through fewer hoops than if the customer wants to move money out of an account, do a wire transfer, or perform some other higher-risk transaction.

THE DATA EXPLOSION

The need for businesses to be able to compile a complete record of customers has grown sharply in the past few years because of the increasing amount of available data and connection points, McKay points out.

Just a few years ago, one customer might have held just one or two IP addresses and phone numbers and one physical address. Now companies need to draw from a growing number of contact points—including multiple phones (mobile and landline, for instance), email addresses, connected devices, laptops, tablets, desktop computers, social media accounts, and other touchpoints—for each individual customer.

In fact, American consumers are projected to own up to 13 devices and connections by 2022, according to Cisco’s most recent Visual Networking Index. As this number of touchpoints continues to grow, identity resolution will become increasingly crucial in consumer engagement strategies, experts agree.

Companies need comprehensive maps of all of their customer identifiers, including metadata from phone calls—the device used to make the call, the pathway of the call, etc.—to be able to identify the caller in advance and provide that ID to the agent in real time prior to picking up the call, according to Prugar.

“When identity resolution is invoked, it is about understanding a complete picture of the consumer so that when a brand understands their interactions with that consumer across all channels, they have a complete, accurate, and dynamic set of information about that consumer,” McKay adds.

THE TECHNOLOGY INVOLVED

At its core, identity resolution relies on an identity graph to pull together and link a wide variety of individual consumer signals and touchpoints into a singular view.

Identity resolution platforms also employ complex algorithms, pattern matching, and machine learning to link the data with a high level of confidence. Data onboarding tools capable of pulling in third-party data from marketing partners, data vendors, and data available in the public domain are also key to identity resolution.

Advanced identity resolution incorporates both probabilistic and deterministic matching algorithms in a hybrid approach, across all data sources, explains Chad Meley, vice president of marketing at Teradata, a data and analytics solutions provider. Probabilistic matching includes various customer attributes, and a threshold of how loose or tight the match needs to be drives decisions related to pulling together the records. For example, it’s probably true that Jon A. Smith at One Main St. is the same person as Jonathan Smith at 1 Main Street, even though the name and address aren’t exactly the same or formatted in the same manner.

Deterministic matching refers to a hard match on one or several attributes. For instance, if the same email address is used on two different touchpoints, it’s safe to bring those records together simply because the email address is exactly the same, Meley explains.

Some companies also add voice biometrics for identity resolution. But while voice biometrics is gaining favor as a customer identifier, it is still very early in the adoption phase. Not all customers are comfortable with it yet, so it hasn’t been broadly deployed.

When it is used, voice biometrics typically supplements other forms of authentication; it is not something that can be used by itself.