Live vs Digital onboarding: why trusting your eyes is not enough to prove the identity of your customers

Share on facebook
Share on twitter
Share on linkedin

A few days ago,  I was invited to a webinar (available here in replay for those who’d like to mock my Spanish :)) organized by my friends from Kiban to debate how lending companies and banks can accelerate the release and launching of a digital product.   

During the webinar, one of the participants asked for my opinion on the following: “what do you think of the robustness of digital verification solution compared to traditional face to face verification, for example in a bank branch?”.

As I was processing the question and thinking of my answer, the following struck me:

I came to the realization that not only digital online solutions are probably more robust, but also that there is no reason why these cannot be used in face to face interactions as well.

Let me roll back and set the stage here.

For a long time, there has been this assumption that face to face user verification was safer than having an online solution. Meaning that if I go to a bank branch to open an account for example, it is easier for a banker to identify me as being who I say I am. On the other hand, it would be easier for me to sign up on an online banking platform while stealing someone’s identity. After all, the internet is super dangerous, right?

Now, this is certainly true if the online platform does not have the necessary tools in place. However, my point here is that if it has those tools, then it is actually more robust against identity fraud, than in face to face interactions. Why? Simply because technology is now better than the human eye.

For the ones that are not familiar with current recognition technology, it may seem like a bad science fiction pitch. But let me dig a bit more here and try to prove my point to you. First of all, let’s consider the actual recognition of the person. On one side, we have a human agent comparing with his own eyes the person in front of him, and the picture on the ID that person presents. On the other, we have a biometric system that compares the biometric points between a selfie or a liveness of the user and the picture on the ID document uploaded. The solution then returns a similarity score.

I believe there is no doubt here the biometric solution is giving you a much more precise answer to that question than the human eye. I mean, come on, how many of us haven’t gotten into clubs or bought alcohol using our older siblings or cousins IDs? Plus, you can potentially set the similarity score that you deem acceptable for your business. Biometrics checks actually function by identifying specific biometric markers, meaning that this still works when you age or go for the hipster beard look.

Now how about if the picture actually is the same person, but it is still identity theft. Indeed, there is an increasing amount of high-end fraud where you see people using the base of a real ID and altering it by placing their picture instead. In this case, we are facing high-end fraud, and it would both fool the biometric solution and the human eye.

But this is where again a digital solution might get the upper hand because, after the biometric step, a good digital solution will also be able to check for alterations on the document uploaded (including but not limited to detecting a modified picture). These alterations would not be picked up by the human eye (unless you’re an expert in altered ID documentation, and trust me if that’s the case, your eyes have been trained during a loooong time) but a digital solution might. More, you might have a specific ID reader in the branch, but these typically focus on extracting data from the bar codes to see if the ID is indeed present in the government database, but it would not necessarily tell you about an altered picture.

This leads me to argue that digital onboarding might be safer than live onboarding when dealing with identity fraud. But not only that I also do not see a reason why this technology would not be used in branches. 

Indeed, the branch agent could take a picture of both the person in front of him and of the ID presented using some hardware made available on site. I doubt this would make the process any longer as you can now get the results in 2 minutes or less. Actually, with the proper training, they would also be able to make sure they accurately take pictures  (no blur, correct lighting, etc…) making the results of these even better.

This approach has already been adopted by some department stores offering credit products, where they use specific booths to collect the biometric information and the ID picture. Maybe it is time for more business to jump on the bandwagon.

Share this post with your friends

Share on facebook
Share on twitter
Share on linkedin

1 comments On Live vs Digital onboarding: why trusting your eyes is not enough to prove the identity of your customers

Leave a reply:

Your email address will not be published.

Subscribe to our Newsletter