ANAHEIM, Calif.–Forty-four percent of consumers say they are more likely to be a repeat customer/member if they experience a “personalized” relationship. And therein lies the rub, because personalizing that transaction requires using data, but data can be complex, costly and lead to false positives.
During a session at the CUNA Lending Council Conference here, Jason Dietrich and Matt Ehrlich of Experian walked attendees through some strategies for striking that balance and more effectively using member-decisioning technology to effectively personalize the experience.
“One third of customers who abandon a business relationship do so because a personalized experience is missing,” said Dietrich, citing Experian research. “Three-quarters of customers at financial institutions expect personalized treatment.”
Dietrich added credit unions need to recognize that personalized experiences aren’t just limited to current members and that a similar experience can be had with non-members if the right data is used.
“Personalization impacts banking across the lifecycle, from onboarding to retention to expansion of the relationship,” he said.
Where the balance must also be struck is in asking the right questions—but not asking too many questions. Dietrich pointed to data showing 87% of consumers believe it is important for companies to safeguard the privacy of their information; 73% say not being able to trust a company with information is a top source of frustration, and 58% said they would switch companies if their data isn’t safeguarded.
How to Personalize an Experience
According to Dietrich, the steps for personalizing a member experience include:
- Develop customer profiles
- Create a customer-focused vision statement
- Train employees to be customer focused
- Give customers choice.
- Develop a self-service experience
- Offer support via social media
- Empower sales and service reps with a well implemented CRM
He added it’s also critical to look beyond the CU’s own internal data, noting there is a move toward using more diverse data sources in order to look at a member “from different angles.”
“It’s not just one aspect, credit. There are other things out there that can describe in a deeper way what makes me who I am,” said Dietrich.
Those diverse data sources include traditional data such as scores, attributes, behaviorial data, trends and trade level mix. But they also include alternative financing data not typically on a bureau report, such as payday loans, title loans, rent-to-own, etc. (which can be used with no-file and thin file members).
The third part is public data and demographic data, said Dietrich. “This really isn’t credit data, but it’s things like channel preferences, general interest, lifestyle clusters, etc. The public data from a fraud and identify risk perspective can be very useful to help cross-check the other data sources.”
Dietrch said 70% of consumers are willing to provide additional financial information to a lender if it increases their chance for approval, or improves their interest rate for a mortgage or car loan. Seventy-one percent of lenders believe consumers will increasingly allow access to their data for lending decisions if they are empowered to turn it on and off.
“You need to put all this together in an analytic way, but not too analytic,” said Dietrich. “You have to think about the trade-off with the complexity. You have to think about if you put rules in place to stop fraud—and there are ways to stop ALL fraud—but the cost of preventing it would be high. So you have to balance those.”
The components of an effective decisioning engine are real-time capability, quick and simple integration, being flexible and customizable, and it being adaptable.
“Broad data and complex analytics may require an augmented decision in system that is hosted or on-site,” said Dietrich. “You have to think about the data you want to plug in and whether you have the plug-ins.”
The Fraud Conundrum
For his part, Ehrlich noted one of the things adding to complexity of overall decisioning is identity and fraud requirements. Identity-centric matters are consuming substantial resources, such as the assumption of full or partial identifies, synthetic identities, and more.
“One of the reasons synthetic identity has become a huge problem as the result of shift from POS to card not present purchases,” said Ehrlich. “The chip has removed a huge channel of fraud from the criminal rings. But the fraudsters didn’t retire; these are super innovative folks, they have found other ways to create fraud and protect their revenues. Synthetic ID is one area that has grown tremendously.”
The other issue, he acknowledged, is poor consumer experience is a big complaint from many, with too many false-positives a leading cause of complaint among consumers and merchants.
While many of the external forces playing a role in decisioning complexity are well known, Ehrlich said internal forces contributing to the problem include inconsistent channel strategies, fraud becoming a verifiable “truth” in traditional data, the focus on consumer experience over fraud prevention, and overwhelming innovation.
Some other stats of interest from Experian’s 2018 Global Fraud Report:
- 75% of businesses want advanced security measures that have little to no impact on the digital customer’s experience. “What’s interesting is consumer do want some security measures to feel comfortable; they want some visible signs of security,” said Ehrlich.
- 42% of Millennials said they would conduct more online transactions of there weren’t so many security hurdles (vs. 30% of those 35 and older)
- 67% of merchants said fraudulent transaction not detected is more expensive to them than legitimate transaction that is lost due to a false positive (33%).
The problem of synthetic ID, which half in the audience indicated they weren’t familiar with, is a hybrid identity that uses some real information and combines it with fake information. The real information is typically the Social Security number (dormant children’s numbers are frequently targeted), with other false information interwoven over that.
What is the primary takeaway from all this?
“Start with the assumption that member data has been compromised and use a combination of passive and active techniques. Then layer in fraud management so the right level of identification and fraud risk treatment is appropriate,” said Ehrlich. “While internal data might be fantastic, it won’t get the job done by itself.”