Multibillion-Dollar Investment Firm Betterment Says ‘Yes’ to More Good Customers Instantly with Socure’s Graph-Defined Identity Verification

“Frictionless identity verification at onboarding is crucial for business growth and protecting our customers,” said Betterment Director of Operations Mark Kassardjian. “Betterment has grown a lot over the last five years. Our old system involved a ton of manual work and wasn’t scaling with that growth. We were seeing a number of customer prospects come in the door but then drop off because they encountered minor friction.”

Company growth goals, a fast-changing fraud landscape, and a hyper-focus to deliver on its mission; with these things in mind, Betterment sought a way to quickly and accurately identify good customers, while providing a frictionless and user-friendly onboarding experience. That’s why they turned to Socure’s industry-leading identity verification and fraud prediction platform. With Socure, Betterment achieved a 30% increase in auto-approvals, a dramatic reduction in manual review rates, and more streamlined onboarding to enable accelerated, scalable growth.

The Predictive DocV solution has been a tremendous boon for productivity and has enabled more efficient resource allocation. We saw our approval rate improve by 30%. We’ve driven down rejection rates considerably while simultaneously keeping out more applications that would increase the amount of fraud we encounter.

Mark Kassardjian, Director of Operations, Betterment

A Partner Who Understands Fraud – and Business Success

Betterment was looking for a trusted partner, one known for deploying market-leading identity verification and fraud prediction solutions as a pathway to better business outcomes. A cross-functional group of internal Betterment experts started their search by reaching out to the biggest names in the industry. From the outset, they knew they could not simply add yet another rules-based point solution to their existing stack. Betterment wanted to use an ML-driven platform and layered approach for identity verification to more seamlessly onboard new customers, with the highest degree of accuracy.

The Betterment team was determined to hold their new partner to the same high standards of seamless user experience, accuracy, and quality that characterizes Betterment’s offering itself, and Socure delivered.

From Friction to Seamless Automation

The key to solving for Betterment’s goal — faster, automated, and scalable onboarding of every good customer — was rooted in Socure’s identity graph and machine learning capabilities, which are designed to persistently process, analyze, and correlate every facet of a digital identity, utilizing over 17,000 features, redundant data sources, and a consortium network of over 1,800 customers to deliver the highest-assurance assessments of new users in the market.

“Socure was the only vendor we took to the proof-of-concept stage,” said Kassardjian. “After testing, we believed there would be a substantial improvement in lift when we went live, and we were proven right. We launched the Sigma Identity Fraud and RiskScores solutions in September 2020, followed by Predictive DocV in April 2021. We haven’t looked back since.”

Socure exceeded Betterment’s expectations by providing a holistic view of the risk of an identity, with Sigma Identity Fraud, Socure KYC, Email RiskScore, Phone RiskScore, and Address RiskScore. “Socure’s correlated, focused checks proved extremely valuable,” said Kassardjian. “Getting that high level of identity coverage and incorporating consortium feedback from such broad industries has been hugely beneficial to our overall risk strategy.”

Socure’s KYC solution also proved instrumental in driving new customer revenue with unrivaled accuracy and population coverage, enabling auto-approval of more good customers, while fully satisfying compliance requirements. Patented AI/ML-driven identity clustering and resolution capabilities helped Betterment take advantage of the highest match rates accuracy in the industry, including auto-approval of more Gen Z, Millennial, and thin-file consumers, who are foundational to Betterment’s growth strategy.

Additionally, with Socure’s Predictive DocV solution, the Betterment team was able to leave manual reviews behind by taking advantage of automated machine learning and deep forensic analysis to verify the authenticity of ID documents. Predictive DocV’s computer vision technology, in particular, looks for ultra-fine details beyond the limitations of the human eye, such as optically verifiable ink, text overlays, subtle hue changes, and watermarks, along with cross-document data consistency, signs of manipulation or forgery, and overall document quality. The easy-to-use guided image capture app and enhanced biometric facial matching enables hyper-accurate selfie-to-ID photo comparison and returns a response within seconds.

“The Predictive DocV solution has been a tremendous boon for productivity and enabled more efficient resource allocation,” said Kassardjian. “We saw our approval rate improve by 30%. We’ve driven down rejection rates considerably while simultaneously keeping out more applications that would increase the amount of fraud we encounter.”

Accurate Verification for Rapid Customer Growth

Once Socure was deployed, Betterment’s onboarding quickly transitioned from manual-heavy processes to automated customer assessments based on an identity graph powered by eight billion rows of data and an ever-growing network database of 800 million known good and bad identities. For Betterment, that meant that decision-making could now be handled by an adaptive and continuous feedback loop that gets smarter through machine learning and analytics.

“Initially, it was a challenge just to define the fraud indicators that could aid us in identifying emerging attack patterns, especially as fraudsters were acquiring more sophisticated tools and technology,” said Kassardjian.

“We saw the sophisticated and highly-predictive indicators that Socure was employing, like name-to-phone correlation, name-to-email correlation, name-to-address correlation, IP-to-geolocation tracking, and distinguishing an actual IP address from a proxy,” said Kassardjian. “Those examples have served as an indicator of whether an applicant is legitimate, or an instance of potential identity theft. Just knowing that we are decisioning on some of the most powerful ML signals available allows me to sleep better at night,” said Kassardjian.

With Socure, Betterment has been able to swiftly onboard legitimate customers while preventing identity theft, resulting in an expanded user base and a healthier bottom line. As Betterment continues to grow and fraudsters continue to evolve, Kassardjian expects to implement even more Socure solutions in order to scale and meet its strategic business goals.