Companies Find Gold in Identity Resolution

Article Featured Image

“It’s one thing to connect devices, but it’s more important to know the user’s journey,” says Evan Rutchik, chief revenue officer for the United States at Ogury, a mobile marketing technology provider. “Mobile is the best starting place.”

Whether starting with fixed or mobile IP addresses, phone calls, contact center interactions, or some other point, the idea is to build a single customer view—often called a golden record—with all of a customer’s details, according to Patrick Tripp, vice president of product strategy at RedPoint Global, a provider of customer engagement and data management solutions.

“Customers use an average of four devices per day,” adds Caitlin McCulloch, head of marketing and strategy at Bazaarvoice, which provides software that allows retailers to add customer reviews to their websites. “Any time you interact with a brand through a different device, it is a way for brands to learn something about you so that they can organize your information for personalized marketing. There are more sources of data than ever, but you can overcome those challenges.”

Though challenging, solving the identity resolution riddle will present companies with opportunities to offer the personalized service that customers expect, as outlined in the RedPoint/Harris report.

“It’s an important tool to have in your toolkit,” McCulloch says, adding that knowing a customer’s different interaction sources helps provide more information than just a customer’s journey; it also gives a window into his intent. For example, a customer might look at home repair materials on a supplier’s website. If the customer has bought such materials through the site before, it might make sense to offer him a promotion. But without confirming data garnered by culling other customer communications with the company, the supplier doesn’t know whether the customer is a do-it-yourselfer or just browsing.


“Consumers interact with brands through a number of different devices. The only way for that brand to understand those preferences is to use technology,” says Jon Hyman, cofounder and chief technology officer at Braze, a company that provides hashtagged emails to help track and identify customer interactions.

But establishing the full customer identification correctly using one or more software solutions and some human intervention is only the first step. The process also requires reconciling different details gathered from different sources (knowing, for example, that Wil Smith, Will Smith, and William Smith are all the same person) and doing so while adhering to privacy and security standards and still aligning with customer preferences.

A few vendors offer identity resolution software that promises to help companies build out golden records, but the reconciliation of conflicting details is far from a solved formula.

“Companies are struggling with those areas today,” Tripp says. “Different departments have their own data sources and different customer touchpoints.”

Abbott adds that companies typically use a combination of first-party and third-party data to build out their “golden records.”

“That can make for a more relevant communication, but first-party data is much stronger in terms of authenticated personal data,” Abbott says.

Similarly, customers have their preferred methods of communications and will seldom deviate, but when they do, it is essential that they are positively identified quickly. Needing to re-identify when they use a different device or when they’re transferred to another contact center agent during the same call is annoying enough for customers to go to competitors with better identification methods.

Even if software is collecting all or most of the data, mistakes are possible. A customer or prospect might fill out a paper form to join a loyalty program only to have a machine reader mistake a 5 for a 6 in the data record. Poor penmanship can lead to a misspelling of a person’s name. Even information entered electronically can be subject to keyboard errors.

Similarly, data collected via the telephone might be slightly different than information entered via the web, so some reconciliation will be necessary.

“You need to set rules,” Tripp says. “Does the data need to be 100 percent identical, or can you be more lenient?”

Tripp recommends using a formula with weighted averages for name, address, phone number, etc., to produce an accuracy score that the company can then use to determine if Will Smith and Wil Smith are the same person. Typically, a common email will confirm whether it is the same person.

If so, the inaccurate spelling should be corrected so that future communications contain the right details.

CRM Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues