There were an estimated 1 million cyber attacks targeting online lending transactions in 2016, with total losses projected to be above $10 billion. This fear has some online lenders setting policies for new account creation that are too strict. For instance, Kabbage CEO Rob Frohwein recently shared that his company was rejecting as many as 3% of its daily online loan applicants but noted that that some legitimate borrowers were likely included in the lot. Kabbage is not alone in establishing fraud policies that are too conservative; it’s become the norm for many lenders. Fundamentally, they don’t have the confidence – even after doing identity data checks – that a borrower is who they say they are. But it’s possible to be smarter about fraud without alienating potential applicants – and sacrificing the lifetime value of a good relationship.
Sophisticated online lenders are adopting more advanced fraud prevention measures that keep fraud in check without being overly conservative, and declining good borrowers. Specifically, they are refining their identity verification strategies with more mature solutions that utilize quality data, sophisticated data science, secure pipelines capable of ingesting these quality data sets and machine learning models. These evolved practices help them expedite applicant onboarding and approvals while safeguarding against fraud.
When talking about identity verification maturity with customers, I find that it’s helpful to discuss it using a progressive 4-stage scale. As companies advance up the scale, the faster they are able to move applicants through the pipeline, with confidence they are minimizing fraud. Each stage builds on the previous one, allowing organizations to ensure higher accuracy as they move forward.
The stages are broken down like this:
- Stage #1: Not identity proofing yet. Most if not all online lenders have progressed past this stage and take at least some steps to verify the authenticity of applicants. However, in smaller transaction markets, or niche spaces, it’s common for organizations that have yet to be hit by fraud to do no identity proofing.
- Stage #2: Verifying a single identity attribute. The first step that online lenders take to protect themselves is often rather minimal, they focus on a single factor to authenticate new borrowers. Common data points to check include the age of email (to verify its been in use for a long period of time), just an address or phone number.
- Stage #3: Multiple identity attribute verification. Online lenders that verify multiple attributes are not only able to authenticate applicants more confidently, but also more readily identify good borrowers.
Stage #4: Holistic identity verification. This stage involves verifying multiple identity attributes aren’t just accurate but all link back to the applicant applying for the loan. In a single search it’s possible to verify a name matches a home address, matches a phone number, matches an email and so on. When searching for linkages between the person and the address, email, phone and related people, it is a significant signal of risk if these cannot be established.
The use cases for identity linkages are broad. Take, for example, synthetic identity theft. In this scenario, a criminal combines real (often stolen) and fake information to create a new identity, which is used to open fraudulent accounts and conduct fraudulent financial transactions. Because “real” details can be verified alone, it is imperative to verify all of an applicant’s attributes connect together.
Stage 4 also unlocks a powerful new set of opportunities to leverage machine learning, sophisticated data analysis and data science that aren’t possible with simpler methods of identity verification. When a whole identity is considered, a world of networks, history and patterns can begin to be tapped for increased speed and accuracy.
Identity networks are extremely valuable to online lenders because they can see signals across millions of transactions and multiple applicants for a real time understanding of identity element velocities, transactional frequencies, and linkage histories. Machine learning and sophisticated data science can be applied to analyze these transactions to learn and adapt to patterns across different industries.
For the fastest identity verification, some identity data services are distilling the result of their sophisticated verification processes into a single number or a score for easy, real-time rule building or integration into a risk model.
Find More Borrowers Faster with Holistic Identity Verification
In an age where borrowers have so many lenders to choose from, lenders need to have a seamless onboarding process or risk losing borrowers to the competition. Key to this ability is having confidence in an applicant’s identity. One of the simplest and fastest ways to do so is holistic identity verification.
When every applicant has the potential to become a return borrower, it’s worth taking a step back and to make sure you are not too risk averse. Having a mature identity verification practice enables lenders to provide faster and more loan approvals for legitimate applicants, reduce fraud and lower good borrower rejection so they can compete and thrive.
Tom Donlea leads the global marketing efforts of Whitepages Pro, the worldwide identity verification data provider for risk management in banking and online lending. With over ten years of online payments and risk experience, he previously was the founding executive director of the Merchant Risk Council.