With the emergence of virtual assistants in retail banking, the key to personalized user experience is a simple and intuitive interface that anybody can use.
On the latest episode of the If I Ran the Bank podcast, Jake Tyler, Co-Founder and CEO of Finn AI, sat down with FISPAN’s Co-Founder and Chief Strategy Officer Clayton Weir, to discuss the emergence of virtual assistants in retail banking. Jake explains how a personalized virtual banker can benefit the consumer, through a simpler, more intuitive user experience.
Don’t miss this episode with Jake Tyler, listen here.
Based on your experience working with banks globally, where do you think banks are the most aggressive in regards to innovation?
When we started, we were definitely at the innovation stage of the market. I had always thought about building a big global company out of the box. We essentially took deals wherever we could get them, like in Central America, South Africa, across Asia, Europe, North America, Australia. In hindsight, we probably should have just spent more time in North America, but I think the challenge is that North America is sometimes a bit of a laggard on the innovation side, and to get the fire started sometimes you just need to go where it is.
I think in general, Asia does a much better job of digital than in North America and Europe. Europe has a bit of a different flavour than North America. Australia is a much more advanced ecosystem in the banking technology world than Canada or the US, so there’s a lot we can learn there.
In our world, in the virtual assistant chatbot space, we saw adoption happening much faster in Asia to start with and then at more innovative banks, but mostly not in North America.
If you ran the bank, what would you go all-in on?
I’d go all in on virtual assistants as the default digital banking experience for retail banking. I think it makes sense that we have a virtual personal banker for everybody. If you want to get something done, you just talk to your virtual personal banker, ask what you want to get done and you get it done.
The experience is personalized to you, so it’s YOUR virtual personal banker. It’s monitoring your transactions, your balances and nudging and coaching you to make better financial decisions. Maybe it sends you a nudge to pay your bill on time, to pay down your credit card, move money to savings. These are the sort of things that I think are possible today, that add a significant amount of value above what I currently get when I log in to my banking portal.
Banking and finance is complicated. There are lots of products and services and it’s not always clear how they work. Banks have so much, with a lot of terms and conditions and all sorts of strange, wonderful jargon. In the past, you could get past some of this complexity by walking into the branch, as we moved into the digital world, we tried to move these things digital and I think we lost the benefit of that personalized, simplified, intuitive user experience. I think a virtual assistant brings a lot of that back, but still provides all the benefits of digital.
Do you see virtual assistants becoming the primary servicing channel for retail banking product consumption?
Yes, and I don’t think it’s that controversial of an idea. If you look at Bank of America’s Erica, for example, that virtual assistant is the primary banking experience in mobile. I think they have around 15M active users of Erica at the moment. You can also see Capital One with Eno moving in that direction. It wouldn’t be a top 15 US retail bank without a major virtual assistant chatbot project. If it’s not happening already, it’s likely to happen in the next year or two at most major banks.
Chatbots and virtual assistants started out as being basically an FAQ, that’s not what I’m talking about. I’m talking about a virtual personal banker that lives in the authenticated environment, where you can get all your transactional stuff done, plus it’s nudging, coaching and guiding you to make better financial decisions.
What do you think the value driver, or business case for chat bots were a couple years ago versus now?
I think in 2016-2017, the primary value driver was innovation. These were all innovation experiments back then. I think now when banks are getting into it, it’s a product that can work and can deliver value.
It is very hard to get things done with online and mobile banking still and it would be much easier if you could just ask to do it. About 80% of people who call-in, tried to self-serve first. It’s much better for the user to be able to ask online/mobile banking using simple, natural language. That core use-case of great self-service, 24/7, for simple routine things that are well suited for a bot, if I can deliver end-to-end journeys within the assistant, there is a very significant amount of value there. I think that’s the foundational layer.
In addition, what you’re seeing from Bank of America with Erica and Capital One with Eno, and others in the market, is add-in nudges and insights on top of that. They’re starting to build a deeper, more personal relationship with the consumer by adding a layer of value on top of the basic banking.
Bank of America is one of the oldest and largest banks, should we be surprised that we’re not seeing newer neo-banks and digital banks adopt this technology?
I think the reality is, the amount of money you need to invest to get something like this working well is significant. Bank of America’s investment has been in the tens of billions, a very significant investment. If you’re a digital bank, you don’t have that budget at all. I think that’s one of the reasons why Bank of America is a leading example of a product like this, is because they went in very hard, very early and now they are reaping the benefits. Certainly through this COVID-19 pandemic, you can see that the use of their product has gone up significantly, it has been a major asset for them.
In the digital banking space, and the same for regional banks and credit unions, the challenge has been how do I train the language model to understand banking queries? Because they are starting from scratch, they don’t have 5-10M dollars to spend on data scientists training it up, so that’s part of the problem.
Now we have a wave of vendors – including Finn.ai – who help to solve that problem with a product that works. In the same token, you have a wave of vendors like Personetics and MX, who can solve the problem of personalized insights as well.
I think if you’d tried a couple years ago, it would have been very hard and expensive which is why someone like Bank of America was early in, but now I think it’s much more accessible because of a much more mature vendor ecosystem.