Conversational Finance: The Impact of Chatbots, AI and Machine Learning

"A decade ago we had the first big leap, and that was web to mobile,[…] Now the next one is mobile to conversational” said Edrizio de la Cruz, co-founder and CEO of Regalii, a startup whose application programming interfaces are used by dozens of financial services providers to build their chatbots.
​
The pressure today to innovate and embrace new technology and practices is significant across a range of industries with financial services at the epicenter of the pressure but research by Econsultancy and Adobe shows that 9% of FS businesses claim to be digital first, compared to 11% across all sectors.

Alongside the drive to embrace digital, there is additional pressure from the market to:

• provide affordable advice to consumers;
• improve and increase access to advice;
• address industry concerns relating to future liabilities and redress without watering down levels of consumer protection.

Which, in context of legacy systems and brick and mortar infrastructure, is a challenge but in reviewing Fintech options in moving forward, AI, chatbots tick a range of boxes, particularly under the affordability and accessibility criteria.

 So, chatbots?

Chatbots are essentially pieces of software that simulate human, natural language conversations and can respond to and act upon queries and commands from users. The advantage these systems have over a real conversation with a human is that they are able to extract and analyse a user’s needs and intent and ultimately return the information a user has requested or perform actions for them faster, at any time of day or night and at significantly lower cost than a human counterpart. 

The benefits of this type of technology are clear with many people choosing to apply and research investments or loans through these types of systems rather than spending the extra time and potentially cash on a human broker that may not necessarily have the best deals available. These systems could potentially pave the way to a fully automated digital process, further removing the potential for human error and bias from financial services.

An apt example would be Capital One’s Eno. Eno is able to interpret text based conversational queries and commands alongside emojis. This includes the ability to check balances and pay off credit cards, while cash transfers are also in the works. Additionally for customers with Amazon Echo, Capital One has also built out a skill that allows for voice commands. At the business side, Capital One stand to make significant savings in terms of time and manpower as users transition from face-to-face and telephone queries to simply asking Eno.

This is only one form that chatbots have taken on so far with the sky being the limit on the functions a chatbot might serve. Other options include:

• Advisory: This could be in the form of banking or investment advice but can be extended to any business vertical. ClearScore have developed a credit checking service known as “Coaching” intended to help millions of people improve their credit score. The bot suggests practical small changes in order to improve your financial future, whether you’re a young person with little credit history, someone with a damaged credit score or just someone wanting to improve their financial situation.
• Admin Guru: Sage launched their chatbot “Pegg” which is used to track spending, record receipts and access accounts with ease. You simply message in with your expenses and it automatically logs it for you. You can ask the bot questions such as “Who owes me money? And when did I last remind them?
• KYC, AML, Fraud Detection: The AI behind the chatbot would be well positioned to evaluate risk, close loopholes and improve fraud detection. Wells Fargo leveraged machine learning originally to prevent fraud, but AI now permeates other areas including compliance, customer experience, underwriting and authentication. The San Francisco based financial institution is also exploring the use of consumer-facing virtual assistants for information updates, improvements in transaction capabilities and business insights.

Furthermore, chatbots do allow for a smoother digital experience overall with the ability to sync with popular applications such as Whatsapp and Facebook messenger, removing the need for an additional download that may alter existing mobile usage patterns. This interface could be instrumental in data analytics as well in improving operational efficiency, promoting innovative practices and improving the customer experience. Banks usually have a mountain of customer data at their fingertips, but often struggle to understand and get value out of it. Customer data is the key to establishing meaningful, personal relationships with banking customers and offering customized products and experiences. Banks need to effectively analyze this data to better know and serve their customers. Conversational banking enables banks to acquire more nuanced customer data. By engaging customers in small talk through conversational interfaces, banks get insight into customer intents, desires, and concerns that are not apparent in banking app and website interactions.

Conversational AI can also help banks better understand this data. Through deep data analytics, pattern recognition, and predictive algorithms, chatbots can communicate intelligent insights about banking customers’ present and future banking needs. These insights can be used to offer more personalized banking products and services to build lifelong customers.

There may be concern that chatbots and other AI supported technology will put significant pressure on a number of customer facing services across a range of industries due to the attractive operating margins and this is certainly true but this is analogous to the widespread worry at the time when motorized options replaced horses, a slew of new roles develop and grew out of the adoption of a new technology.

An August 2016 report by Forrester suggested that banks should focus on developing the AI technology to build better bots in the future, rather than launch bots on messaging platforms now and provide poor experience to their users. The main criticism against chatbots is that they lack the empathy and the emotional response that a human can provide, which makes them less capable of dealing with complex situations involving financial decisions. Generally speaking, humans working in customer services will know how to respond to frustrated customers and not aggravate the situation; they can listen, reason, empathise and read between the lines.

​A lack of emotional intelligence is a serious limitation that existing bots have. There is definitely a whole new world out there to explore in considering chatbots as an investment for your own business or in a third party capacity through a promising Fintech firm but there is a degree of risk in this relatively nascent space with regards to meeting customer requirements and security. There is potential for significant secure data being shared with these interfaces as clients address their needs. In addition, Forrester (2016) found that although chatbots are developing rapidly, customer experience is not. Many fail to effectively meet users’ needs due to poor infrastructure and lack of fundamental understanding of AI. Hence it is imperative to work with institutions that have the infrastructure to support your venture effectively on the front end as well as on the back office.

This post originally appeared on Crowd Valley Blog.

Source: http://www.businessinsider.com/chatbots-are-coming-to-financial-services-2016-8?IR=T