Why chatbots will rule everyday banking
In our continuing investigation into AI and the banking industry, we interview Jake Tyler, a former Perth-ite who’s now CEO of a Vancouver-based fintech startup called finn.ai.
Developers of the world’s first, full-featured, everyday banking chatbot in Facebook Messenger for ATB Financial, finn.ai are soon to roll out bots for their first Australian bank, amongst others.
How are your earliest clients faring with their chatbots?
It’s still early days in terms of the rollout of bots to banks, but what is very clear is that this is going to be a way that people bank in a year or two. This will happen fast, for three good reasons.
One, this is where people are, it’s where they spend time. You will interact by chatting to your Alexa unit, by chatting to your Google home unit, by chatting on Facebook Messenger, or by chatting in a bank’s existing iOS or Android app.
The second part is, for consumers, there’s a lot of latent demand here. We don’t use our phones for calling anymore, we use them for messaging.
For me, and people my age and younger, I don’t want to jump on the phone and call a 1800 number, it’s a huge bit of friction. So from a customer care perspective it makes sense that we move in almost a wholesale way to messaging.
The third, really big piece here is that banking in two to three years will not be this utility that we currently get; but it will be a personal finance coach that helps you save, budget, and make better financial decisions.
All of those things offer a step-change benefit that will drive adoption in short order. For a bank, they need to get out and start deploying the early versions of these experiences now rather than wait.
What is the danger of waiting?
The reality is, if you don’t get out and learn, in market, you won’t get to market quickly. So the danger is you’re inefficient, you’re spending more on customer care than you have to, you’re not matching consumer expectations.
Banks aren’t competing on digital experience versus other banks, they’re competing against Uber, Facebook, Amazon, so they need to step their game up. So this is part of how they’ll do that.
And from an experience perspective, if my bank helps me save and budget and manage my credit score, and all this other good stuff, I’m more likely to consolidate my wallet around that bank.
What are the challenges for banks?
We’re in the process of rolling out with a number of tier one and tier two bank clients that we can’t name yet. One of those is in Australia.
From a more macro perspective, the impact machine learning and AI are going to have on banking is very significant across all parts of the bank, and what is going to be more and more of a competitive advantage is an open core banking platform, with the ability to partner and collaborate easily, both in the technical architecture / infrastructure and in people and processes.
Fundamentally banks have a lot of setup that they need to do. They have a lot of data at the moment. The way they manage that data needs to change and they need to be a bit more open in the way they work with partners.
If we want to deploy out to every customer in Australia, with this personal financial management assistant in your pocket, that’s not going to happen overnight. It is a learning system and us humans need to learn how to deploy them as well.
What are the major benefits for banks?
One is cost-saving. We’re yet to see good hard data about what that looks like but certainly it should be big. With our early clients, we’re able to take 10 to 30 percent of volume from the call centre. That’s millions to tens of millions of dollars a year in savings for any of the big four [Australian banks].
On the digital engagement side, banks spend a lot of money delivering a digital user experience which is how most people interact with their bank these days so banks need to be where customers are, in these new channels.
And they need to do a good job of customer experience, and that customer experience is changing. If they want to maintain loyalty, improve share of wallet and reduce customer acquisition costs, figuring out the kind of experiences we’re working on are also fundamental to delivering that experience.
Are you aware of what the banks are doing in Australia?
I saw that UBank just launched a product with IBM’s Watson technology for loans.
Every bank, certainly the big four and most of the tier two banks in Australia would either have active pilots or have plans to roll this out in the next year or two.
Your vision for AI scenarios 10 years hence?
The bulk of customer care – the routine, mechanical queries, will be handled by machines. What we should add is the people who do customer care will be dealing with more complex queries - they’ll be there to chat to you, instead of dealing with these very simple queries.
This is about better access to humans when you need it, not frustrating IVR systems.
You’ll also talk to your bank in your car, you’ll be able to pay your bill by talking to your fridge, or your watch, or Siri on your phone; so a lot of these simple transactional things will be done with voice command, but otherwise banking will be wherever it’s convenient for you to interact with it.
Your own private banker
The thing that gets us most excited is that with AI and machine learning we can deliver private banking to people - personal banking to everyone, at close to zero marginal cost.
If you think about how banking was done 50 or 60 years ago, you had someone in a branch who you knew and they helped you manage your money. What happened with digital self serve, all of that thinking and effort got pushed to the customer in an effort to save costs for the bank.
What we get excited about is that we can help people budget and save and pay their bills on time and understand how their behavior impacts their credit score… at the end of the day, we’re here to solve consumer problems.