July & August: Ask Us Your Digital Identity Questions
We’ve received loads of questions from customers, partners, and occasionally from friends, about a myriad of things relating to Digital Identity.
We’re sure they aren’t the only ones with questions – so we’ve decided to launch a monthly mailbag where our Director of Strategic Accounts & Client Quality Services, Tom Canfarotta, will answer your questions.
In July & August, we received many questions about what happens when someone can’t be verified and what to do in escalation. Here are our answers.
What can I do to reduce friction in my new customer onboarding processes when verifying identity?
Wow, this is a big topic! I will do my best to give you the 30,000-foot view.
In today’s online world there are many methods used to automate identity verification in your new customer onboarding process; however, most all methods rely on just two sources: identity data and identity documents.
The identity data methods are based on a comparison of the personal data you collect from a new customer against one or more trusted and verified public and/or non-public data sources. Provided you can find and associate the personal identity information you have collected, these methods are usually considered very low friction. Most data verifications can be performed behind the scenes in just a few seconds without the new customer’s further engagement.
The use of trusted and verified data sources can also be used to generate “out-of-wallet”, Knowledge Based Authentication (KBA) challenge questions. While these KBA questions create a bit more friction, they can be used to add an additional layer of assurance that your customer is who they say they are.
The other commonly used methods to reduce friction are based on the collection of identity documents. Drivers licenses, passports, and other identity documents can be collected online or in-app and authenticated in real-time. While these methods can also include the automated collection of identity data for use in a “form-fill” or for further data verification, most will use these methods to verify identity solely by authenticating the document.
Based on my experience, the most important factor in reducing friction with an automated onboarding identity process is choosing an identity vendor that provides professional services that include ongoing assistance with your web-facing integration. In the end, choosing an experienced identity verification vendor that delivers multiple methodologies based on both data and documents in real-time is the best way of achieving a low friction verification. Preferably the solutions should allow you to deploy them either stand-alone or in tandem, as this will offer the best chance of achieving a balance of an excellent customer experience (low friction) and mitigation of risk.
What can I do to increase the number of successful verifications on customers with a “thin file” data history?
~Nina, Washington DC
OK, I am pretty sure there may be readers of this post that may not be familiar with the term “thin file”, so let me quickly address this first. I mentioned in my previous answer that a common method used for online identity verification relies on a data comparison of personal data you collect from a new customer against trusted and verified non-public data sources. The term “thin file” is an industry term that refers to those individuals where we find little or no information about them in those data file sources, making verification difficult or impossible.
Individuals can fail a data verification for all types of reasons, perhaps they have never established credit or a bank account or maybe they just moved and haven’t updated all their records yet.
There are several things you can do to increase successful verifications where your customer demographic includes individuals who don’t have much information in their data file.
First, engage a verification vendor that understands this challenge and can help with strategies to address it. Your vendor should be able to review your onboarding process and identify how best to maximize your verification onboarding to increase successes. Make sure your vendor includes a broad spectrum of data sources that will increase your chances of finding more individuals with a data check.
Just because an individual’s identity could not be verified with identity data, doesn’t mean you have to reject a transaction or risk losing a sale. Another excellent way of increasing the number of successful verifications when a thin file is the culprit, is to deploy an automated online document verification. This approach will guarantee your “thin file” online customers a better chance to be verified and guarantee your organization a way to avoid unnecessary transaction abandonment.
How do I know that my customers are not uploading identity documents that belong to someone else?
~Jackson, New York
This question is tricky because it gets to the heart of all online identity questions: “how can I be sure you are who you say you are?” To answer your question, we really need to answer two questions: first, is this identity document valid/real and the second, is the person presenting this document online, associated with the document.
When it comes to answering questions like these, the saying “there is no silver bullet” always seems appropriate. We always recommend for any online verification process, that you consider a customized, layered approach based on your specific application and tolerance for risk. So, with this in mind, let’s take a shot at answering the questions.
First, is the document valid/real?
Of course, the challenge here is that the fraudsters have access to sophisticated technology allowing them to create counterfeit documents that are pretty convincing. That said, with the right online document verification services, you can verify a document with reasonable assurance using a number of different available methodologies, each of which may vary the level of assurance and efficacy.
The standard online document verification services generally verify against real document libraries to measure various aspects of the document layout and fonts against a real document. To add further assurance for high-value transactions, you can up your game and deploy an online service that delivers a manual forensic review of the document. You will need to keep in mind this method is expensive and time-consuming.
The second question is whether or not the person presenting the document is actually associated with it.
One way to assure the person presenting the online document is in fact associated with that document is to add a facial recognition (selfie) component to the transaction. This technology compares a photo ID with a real-time selfie providing some assurance that a person associated with the document is actually the same individual.
Another way for us to accomplish an association is to add a data component to triangulate the transaction, such as an SSN or KBA question set. The addition of these can increase your certainty that your customers are presenting a document that they own and not one belonging to someone else.