Key Considerations for Successful Customer Onboarding

Using a layered defense approach, businesses can block even the most complex fraud types and reduce onboarding friction for new customers

customer onboardingImage by Biljana Jovanovic from Pixabay

Consider this, you want to try a new online service, you’ve done some research, you hit the sign-up page… and then the list of forms, fields and password complexity requirements hit you, you lose the will to proceed and you abandon.

Now, the other side of the story. On the other side of the application form that you just abandoned, is a fraud operations team inside an organization, battling with an influx of applications from opportunistic fraudsters and organized gangs using both stolen and insidious but perfectly formed synthetic identities. To put it in perspective, the U.S. Federal Trade Commission reports that synthetic identities are growing fast and often the most difficult to detect.1

In the wake of increasing fraudulent applications, digital businesses are facing enormous challenges in successfully onboarding new customers while keeping fraudsters at bay. Businesses are scrambling to acquire new customers and trying hard to make the entire online journey as seamless as possible so that customers do not abandon their applications mid-way. And, in the process, they must also ensure that fraudsters do not sneak in.

Thus, the key considerations for businesses to have successful customer experience from application to onboarding are:

  • Reducing Friction
  • Detecting Fraudulent Applications
  • Complying with Regulations

A Tight-rope Walk

Businesses are however, finding themselves walking the tight rope, trying to prevent abandonment (by reducing friction for genuine customers), preventing fraud, and staying compliant with changing regulations. As a result, when faced with a deluge of fraudulent applications, businesses often find themselves ill-equipped to strike a balance. They are not able to distinguish between a genuine application and one that is fraudulent, which may allow fraudsters to sneak into the business ecosystem and orchestrate numerous types of financial crimes such as money laundering and creating phony credit files.

Teams of reviewers spend hours reviewing applications either manually or through legacy solutions, which translates into loss of valuable time/resources, unduly slow underwriting speeds, and loss of business. In addition, businesses also run the risk of losing customer trust resulting in rather long-term reputational losses for the brand, apart from attracting penalties arising from non-compliance to regulations if crooks manage to run these fraudulent accounts to carry out scams.

Legacy Solutions Pull Businesses Back

In a scenario where it is a matter of ‘when’ and not ‘if’ a business gets targeted, it is important that businesses remain agile and deploy better gate-keeping mechanisms to sieve out fraudulent applications at the first level itself. Businesses must also make an effort to understand how identity theft and fraud works, including the building of synthetic identities. This will not only help them get a clear understanding of the continuously evolving and complex fraud techniques but also better adapt to the rapidly changing fraud landscape, foster trust among customers, and remain on the right side of the law.

Businesses can no longer rely on legacy solutions as they are woefully lagging behind. The need of the hour is advanced data-driven solutions that enable highly precise automated review of applications and an ability to filter out fraudulent applications in real time. In short, solutions that enable businesses to build a good defense using available data sources and leveraging the latest machine learning techniques.

Layered Defense Using AI and ML

Today, advances in technology allow harnessing the data efficiently such that it can be used as a potent weapon to fight fraud. Technologies such as artificial intelligence (AI) and machine learning (ML) make it possible to stitch pieces of digital intelligence together with enormous volumes of data—structured and unstructured—from disparate sources. This helps unearth subtle patterns and identity correlations to filter out fraudulent applications right at the outset.

Further, mixing the identity correlations with data from the forensic detection layer helps identify more correlations. Automating this link analysis across all combinations of attributes and all new and historical applications, incubating fraud rings as well as high-risk new applications are also filtered. Layering historical fraud labels and blacklists on to this evolving network of connections produces a powerful model that can separate the high-risk clusters from normal applications. Further, triangulating multiple ways of validating an identity allow validation of the identities themselves that may have cleared KYC checks. Finally, using orthogonal sources of information—telecom, social, blacklists—even the most sophisticated and complex fraud types can be unmasked.

Therefore, in order to sustain business growth with frictionless onboarding for genuine customers and filter out fraud, businesses must leverage the power of AI and ML. Using AI and advanced machine learning algorithms businesses can build a robust and layered defense against the most complex frauds.

To find out how Simility can help you identify fraudulent applications while reducing friction for genuine users, please download our Ebook on Onboarding New Customers: Among a Host of Stolen and Synthetic Identities, read our solution brief  to fight new account fraud, or email us at contact@simility.com.

 


1. The Changing Face of Identity Theft, Federal Trade Commission. https://www.ftc.gov/sites/default/files/documents/public_comments/credit-report-freezes-534030-00033/534030-00033.pdf
Stephen Moody

Stephen Moody

Dr Stephen Moody, Solutions Director at Simility, is responsible for developing industry focussed solutions on top of Simility's cutting-edge adaptive risk platform. Stephen is a recognised global thought leader and is involved in the development of analytical solutions to combat financial crime, fraud and cybercrime. Previously he has worked as Solutions Director at ThreatMetrix and Head of Financial Crime Solutions at BAE Applied Intelligence. Stephen is passionate about the application of the latest capabilities in AI and deep learning to the hardest problems in fraud and financial crime prevention. Stephen holds a PhD in Astrophysics from Cambridge University and when not wrangling with detection algorithms likes to dream about the fundamentals of reality, consciousness and everything in between.
Stephen Moody