Skip to Main Content

In the News

Evolution of Voice Assistants in Banking

From Simple Q&A to Personalized Advice

Mobile banking was one of the first apps for smartphones. Financial brands led the charge on the first mainstream wearable, the Apple Watch. Now, conversational interfaces including voice-first interfaces are the next frontiers to explore as consumers increasingly look for speed, convenience and security in their banking tasks. Why should banks invest in voice?

Because, it is the most natural means of communication and can lead to more personalized interactions. With the growing popularity of voice devices, voice experiences will soon become table-stakes. Moreover voice makes banking more inclusive especially for the visually impaired, giving them more independence in day to day banking. In order to become truly customer centric, banks need to stay abreast of this cutting edge innovation and define their conversational strategies now.

The growing voice-first landscape 

In October 2017, India became the fourth market to launch Echo, Amazon's voice-controlled device. Amazon Echo and the associated Alexa application is already popular in the US, UK and Germany and it remains to be seen how the Indian version with a localized accent resonates with the audience. Amazon is not alone in the voice assistant race - Google Assistant, Apple's Siri, Microsoft's Cortana, Samsung's Bixby and host of other platforms from the likes of Alibaba, Tencent and Xiaomi are competing for the share of voice.

Statistics around voice-first interactions are already encouraging - In 2016, Google estimated that 20% of searches from mobile devices were done using voice functionality. Amazon has sold millions of Echo products globally and there are over 25,000 Alexa skills (voice functionalities) from a wide range of companies like Starbucks, Uber, and Capital One. According to a forecast from eMarketer, 35.6 million Americans will use a voice-activated assistant device at least once a month in 2017 and Amazon's Echo speaker will have 70.6% of those users.

Currently these skills are primarily functional, the experience rather mundane and thus maintaining sustained user engagement is a challenge. Unsurprisingly only 3% of Amazon Alexa and Google Assistant voice applications that users enable remain active by the second week, according to a survey by Voice Labs. But, usage is likely to grow as more people get familiarized with voice assistants and the voice ecosystem evolves. Moreover as with all AI driven experiences, voice assistants are likely to get smarter over time benefiting from continuous improvements in voice recognition, analytics and machine learning, leading to greater engagement.

Voice-assistants in banking today 

Voice assistants in banking today support basic queries such as checking account balances, making payments, paying bills and getting account related information. A few banks have integrated voice functionalities into their mobile banking apps. Now several others are looking to grow their presence on voice-first devices including Echo (Alexa assistant) and Google Home (Google assistant).

Beyond mainstream usage, voice guided interfaces have tremendous benefits for those with visual and cognitive impairments. USAA, for example, introduced a voice-guided remote check deposit feature in its mobile app which helps its visually impaired members with voice guidance while trying to scan a check. Another intelligence voice capable assistant is NADIA, which is being developed to communicate with the thousands of national disability insurance scheme participants in Australia.

Developing a voice strategy for tomorrow 

We are still far from realizing Iron Man's version of JARVIS! For banks it's vital that they start preparing to get a foothold in voice interactions. But they would need to proceed with caution especially given the high expectations consumers have and the relatively mundane experience offered by today's voice assistants. Banks must identify relevant use-cases, invest in designing the experiences, test, evolve and alongside set realistic expectations on the experience to avoid disappointing their consumers. 

Identify relevant use-cases: Banks need to first identify which customer processes and micro-moments can benefit from cognitive technologies and conversational voice platform, both in the near term and in future. As noted earlier most voice use cases today are very basic. However with rapid advances in Natural Language Processing (NLP) and Machine Learning, voice experiences will only become more personalized and meaningful. As banks adopt an open API approach, aggregate data from multiple sources and develop a partner ecosystem, voice interfaces can be made to deliver far more personalized advice in future. For example a user could ask her bank's voice assistant if she is saving enough for retirement or to provide suggestions on rebalancing her savings and fixed deposit accounts. Of course such advice can also be offered through other channels but it is the highly intimate nature of voice that can make the user experience more rewarding.

Voice bot's functionalities - authenticate, inform, transact and advise

Give your brand a voice personality: A key consideration is to define the voice assistant's mannerisms, characteristics and attitude. Further, a bank needs to decide the relevant - whether the voice character is to sound more functional or playful or efficient. This will help the bank keep interactions consistent over time. Moreover, as voice search begins to proliferate banks would need to be recognizable based on their own voice persona too, perhaps with audio mnemonics or an audio logo to complement the visual logo. Visa for instance came out with a signature sound post completion of a payment which it feels is optimistic, energetic and not overtly intrusive.

Social skills must also be considered while designing voice applications. Banks would like their voice bots to be likeable, trustworthy, and empathetic and adapt according to the tone of the interacting user. New Zealand start-up Soul Machines' bot Nadia is programed to read human emotions and depending on the mood perceived, Nadia can adjust her answers. Mattersight, a provider of SaaS based personality software solutions helps decipher people's personalities from their voices. Already bots like Softbank Robotics's Pepper and use such technologies to create empathy. 

Choice of platforms is also vital. Currently most banks are going after Alexa, given the popularity of Echo devices. In order to make the adoption widespread, banks will need to make the voice experiences available regardless of the platform, wherever the customer needs them, be it at home, on mobile or while driving in a car. 

Let your voice assistant learn and evolve: Providing relevant answers to specific questions posed by a user will always be important for a voice assistant. In the long run though, personalization and contextual advice will be vital for sustained engagement. Machine learning capability will thus play a critical role here enabling the voice assistant to learn from every customer interaction, relate it to the current context and provide relevant advice. All the more reason for banks to start deploying their voice assistants! Additionally, robust analytics capabilities need to be in place to help banks assess the impact of their voice applications. Over time analytics can help banks identify which voice skills are used more often than others and accordingly lesser used skills can be pruned or made more relevant.

Voice assistants are unlikely to replace current communication channels anytime soon. People will likely be less keen to use them in public places because of privacy concerns. But their adoption is definitely growing. Given that speech is one of the first skills we master, expectations with voice experiences will always be high. Investing in voice solutions today will help banks develop a new communication channel with their customers and better prepare for future conversational engagements. Banks will benefit from being seen as more innovative, offering smart, contextual solutions to their customers. Moreover, voice can lead to better inclusion, offering more independence to customers who are mobility impaired, visually impaired or with cognitive impairment. Overall voice would help banks widen reach, improve engagement and ultimately improve their bottom lines.


Economic Time

Publish date

January 20, 2018


Varun Mathur


A group of people in conversation