By Alex Bray and Sachin Pai, Genpact
In Star Trek, Majel Barrett-Roddenberry was the voice of the computer for nearly 45 years. Across multiple episodes and movies, she embodied that ubiquitous sci-fi trope – the talking computer. Google even called its first speaking assistant prototype,Google Majel. Now, as with so many other aspects of the Star Trek universe, the talking computer is becoming a reality.
Artificial intelligence (AI)-powered voice interfaces are beginning to gain traction in the market.
Siri launched in 2011 leading the way for mobile devices.
Siri now has more than 41.4 million monthly active users in the United States.Amazon launched Alexa in 2014 for the home device market. Fewer than 4.5 million smart speakers were sold in Q4 2016. By Q4 2017, this had jumped to over 17.8million units.
So why is voice user interface (UI) proving to be attractive to retail banking customers? Voice offers a new way to simplify complexity for customers – a crucial benefit for a complex business like financial services. Fundamentally, voice UIs can remove the need for a customer to navigate screens and a bank defined information architecture to find the answers to their questions. Customers will be able to ask questions with natural language and have answers spoken back to them. Not only is this easier to use, it also improves accessibility for the elderly or the visually impaired. Now customer access to smart phones and home devices which support voice UI, there is a real reason for them to use it.Equally, as customers acquire more and more connected devices, the Internet of Things (IoT) will become more prevalent. Voice UI will become a crucial factor in building interoperability between devices.
So what does this mean for banks? It is tempting to see voice UI as an evolution of digital banking channels. That would be a mistake. Ten years ago, many banks approached mobile banking as an extension of online banking. The lesson that banks quickly learned is that each new channel needs to be designed from the ground up, based on extensive research into customer behaviours. It is still early days for voice UI in banking, but already we can glean some insights and forecast some trends.
We can already see some banks trialing voice UI. Since November 2017, Ally Bank customers in the United States can ask Alexa, in their own words, to move money,what their balance is, and what the price of something costs in hours worked. Capital One and USAA have also launched Alexa skills.In March 2018, JPMorgan Chase even launched a new Alexa skill that will allow institutional clients to access its research through the voice assistant.
What is the opportunity for banks? According to Genpact research of more than 6,000 consumers worldwide,27% of customers would be happy to set up new accounts using voice assistants like Alexa. While still only early days yet, this already points to a potential visual-free UI future for new product sales. Banks can now explain their products and services simply and in natural language, without needing to go to a customer. Banks can also use voice UIs to seamlessly weave their services into their customers’ everyday lives. Why should a bank not offer price comparisons on a customer’s purchases made through voice UI?
Despite these opportunities, there are risks which banks need to consider. First and foremost, banks need to move quickly. There are many competitors in the market – and first mover advantage is significant. The greatest threat is likely to come from other players outside of financial services– aggregators, social networks, or financial platforms. For these companies, the provision of a highly differentiated user experience and interface is core to their business models, so they are likely to go further and faster than many banks.
Banks will also benefit from addressing a common problem in mobile banking – how will they differentiate their brand relative to others? It is easy to make a branch look luxurious or contemporary – in keeping with your brand. How will this brand differentiation be achieved in a voice UI world? Also as with mobile banking, how will customers be able to transfer from voice UI to online to contact centre?
Once a bank encourages customer to use a voice UI, there are several further challenges to address. How will the design processes to ensure customer security? Can the bank use biometrics to recognize voice patterns? How will processes will the financial institution put in place to stop customers being overheard, revealing their passwords or financial details?
Finally, banks need to work out how to market, upsell, and cross sell effectively through a voice UI.
The big question for financial institutions is what should their next steps be? First and foremost, banks need to start to build and test voice UIs. Rapid prototyping based on ethnographic research and customer journey design, backed up by experience analytics,is crucial. This is a tough journey to start on. In July, Nordnet in Sweden terminated their contract with IPSoft for their AI-powered assistant, citing a lack of progress. Yet, in the same market, SEB has been able to increase customer engagement with the same technology. Banks should seek out expert partners that can leverage experience from across multiple industries to help them deliver an experience that is differentiated from their competitors.
The success of voice as a user interface ultimately rests on the degree to which customers find transacting with their banks easier and more frictionless. This depends largely on the level of artificial intelligence (AI) that underpins the voice user interface. AI is at its infancy and algorithmic theory, let alone technology, is far from enabling general purpose conversations systems.
As a result, domain depth in customer engagement in retail consumer banking will be a critical success factor in coding, building knowledge repositories, and determining how the AI makes its decisions. Machines are only as good as they are programmed, so the human component in the training is key, bringing in experts in all areas of customer service. These knowledge assets that will enable the underlying AI platform to successfully navigate the specific use cases that customers care about while transacting with their banks. Coding for and building such repositories and decision graphs is no menial task. At this stage in the cycle, financial institutions may benefit from working with an experience partner with a conversational AI technology stack and a consultative approach to innovate and orchestrate a purpose-built system to solve a bank’s specific business challenges.
To bring us back to our starting point, it is time for banks ‘to boldly go’ into the future of voice UI.