Status update: conversational AI in the banking sector in 2019

By Henry Vaage Iversen, CCO & co-founder, Boost.ai

Banks and financial institutions are especially well-suited for the adoption of conversational AI. This is due to the typically high volume of small-to-mid-sized inquiries that they receive on a daily basis. While each and every customer inquiry is important, it is possible for menial, repetitive tasks to eat up valuable time of employees in the contact center, plummeting productivity levels.

Henry Vaage Iversen
Henry Vaage Iversen

A virtual agent offers the perfect solution to automate the bulk of these interactions. Juniper Research predicts that the success rate of customers interacting with artificial intelligence in the banking sector will reach over 90% by 2020. With the right implementation, it is possible to handle everything from balance inquiries to funds transfers, freeing up support staff to assist customers with nuanced topics like mortgages or financial advice.

Banks love this because it means their employees are more efficient, and they ultimately save money. And their customers love it, too, because they get the answers they need without falling prey to any of the usual customer service pain points.

Standing out in a crowded market

Being able to understand every customer is critical if you want to give your customers a fast and frictionless experience – not just the ones who know what keywords they need to use to navigate a chatbot.

My team and I like to say that ‘you can’t help anyone if you don’t first understand the problem’, which is why virtual agents need to be built using the most advanced natural language understanding (NLU) on the market. Combining natural language processing, deep learning and proprietary technology called automatic semantic understanding (ASU). ASU works alongside deep learning models to seek out more complicated connections between sentences, increasing the complexity of what inputs it can understand and reducing false positives to a minimum.

Any fit for purpose solution needs to make it possible to talk with a virtual agent, rather than being talked at by a chatbot. This technological advantage gives banking users the confidence to implement a ‘chat-first’ strategy, where customers first have to interact with a virtual agent before a human operator can be brought into the loop.

Delivering results with a ‘chat-first’ strategy

We’ve seen the use of conversational AI grow exponentially in the banking sector. Norway’s biggest bank by market value, DNB, began using conversational AI in 2018 to help handle the thousands of daily customer inquiries they received via their website. Using a ‘chat-first’ implementation strategy where their virtual agent acts as first-line support, they are seeing even better results than expected.

Since launch, DNB have seen a 55% drop in traffic to human chat. This translates to their virtual agent completely automating 20% of all their customer service traffic – that’s across all channels including email, phone, etc.

Ultimately, what this proves is that customers are less concerned whether their answers are coming from a human or a machine, just as long as they are correct and delivered in a timely manner.

Looking ahead. Bringing conversational AI stateside

In the Nordics, banks need to be forward-thinking in order to keep up with the expectations of their tech-savvy customers. The perfect proving ground for preparing to tackle the US banking sector.

There is a clear gap in the market for artificially-intelligent virtual agents that could increase customer experience and satisfaction for US banks and not make it worse through impersonal automation from poor quality chatbots.

Conversational AI presents a real opportunity to deliver a personalized customer experience on par with the good old days when businesses could maintain meaningful relationships with customers. Only now they can do so on a much larger scale than ever before.

Many banks in the Nordics are already taking their virtual agents beyond a basic info-bot implementation and branching out into the transactional capabilities that really deliver on the potential of conversational AI.

SR-Bank in Norway, is in the process of having its virtual agent officially certified as a financial advisor. This will allow it to give banking customers recommendations on retirement plans and mortgages, and could be a true game-changer for the industry.

The recent launch of the world’s first virtual agent network also opens up exciting new possibilities for the use of conversational AI technology. Allowing connected virtual agents to talk to each other, determining which one can provide the best assistance to the customer – all in the same chat window for a truly seamless experience. It’s a fantastic way to bridge the gap between business units that can often be siloed away in large enterprises like banks.

The future looks bright and we’ve only just begun to scratch the surface of what is possible for conversational AI in the banking sector.