When COVID hit, call volumes surged. Customers were calling their banks and financial institutions out of fear. Banks had to scale this volume while simultaneously transition their customer service staff to work from home.
Vijay Balasubramaniyan is CEO of Pindrop which provides many of the top banks and insurers with voice authentication and fraud detection. Pindrop has raised more than $200 million. Vijay joins us on the podcast to talk about the challenges in today’s banking call centers and how some of the top firms are improving outcomes and reducing fraud. We discuss how COVID is impacting the financial industry and what the ecosystem may look like on the other side of the crisis.
What Pindrop does
Pindrop provides voice authentication and security. When you call a financial institution, you're greeted with a whole bunch of questions: what's your date of birth? What's your mother's maiden name? We do away with all of that. We replace passwords, pins, questions -- everything that's the older form of security. We look at who you are based on your voice, what you have based on your device, and what you do based on your behavior.
I founded the company for a simple reason. This was my PhD thesis. I started working on this because of a failed suit purchase. I bought a suit in India and gave it in for alternations. I didn't think anything of it. The next day at around three in the morning I got a call from my US bank. They wanted to verify the transaction and asked for my social security number so they could authenticate me. I wouldn't give it to them. We played this stupid cat and mouse game for thirty minutes until I told them to cancel the transaction. The tailor stopped working on my suit after the transaction was canceled. Voice over the phone is over 150 years old and we still don't have a basic sense of where this call is coming from or who's on the other end.
If you flip this around and think about a bank, they get calls from everyone and don't know who's on the other end.
Use cases within an FI's call center
The number one use case is the first thing that happens when you call a call center environment -- they have to identify who you are. Either self service or an agent -- you have to be identified before you go any further. 90 percent of all interactions with a financial institution requires them to identify you as the first step.
Most of what banks use is knowledge-based authentication. They ask questions like date of birth, social security, and mother's maiden name. These questions have gotten crazier and crazier.
Moving away from knowledge based authentication
These rely on the old form of security where there's a shared secret. There's something you know and something the financial institution knows. But now, with all the data breaches we've had, every fraudster has access to this information. 93% of fraudsters beat the knowledge based authentication questions. Worse, 30% of customers don't know the answers to the questions because they're getting increasingly more complicated. As a security solution, it creates a lot of friction. It's a horrible security solution.
We passively figure out three things: Is this who you are based on your voice? Is this your device? Based on the way you enter information, is it you using your device? We're replacing the single factor of knowledge based authentication with a multi-factor authentication solution. It makes a very strong authentication solution and all this happens behind the scenes in the first few seconds of the call.
The COVID era
Now that everyone is working from home, you've pretty much lost the ability to go into physical locations. Call volumes have been surging. We saw them jump 200% and then during the peak of March, jump 600%. When the government stimulus came out, volumes jumped 1000%. Massive numbers of people were calling in. They were panicked.
Here's the kicker: while this is happening, banks had to figure out how to get their entire call center population to work completely remotely. Large financial institutions have as many as 10,000 call center agents.
We saw fraudsters step up their automated attacks. It was an easy way for them to figure out who to target. For companies that made the transition quickly to work from home, agents were dealing with a 150% increase in fraud. The organizations that didn't make the transition quickly, fraud went down. That's because fraudsters got equally frustrated waiting for hours on the phone, just as customers did, and they gave up. Call wait times went up to 55 minutes.