Let’s Chat about Chatbots

Introduction

Last year, a rather hilarious (/creepy) story about Facebook went viral across the Internet. The rumor was that company data scientists had frantically shut down a research program that went off-piste, when two A.I.’s involved created their own language and began conversing with each other in it, to everyone’s horror. Like the A.I. operatives in question, the story took on a life of its own yielding some sublimely sensationalist headlines, many going as far as professing the beginning of the end of our species… It turned out that the story had been blown completely out of proportion, as things tend to in our increasingly click-bait driven existence. That said, for many readers, that story marked their first exposure to a technology that is poised to make a huge impact on their lives, whether they realize it or not: Chatbots.

What are Chatbots?

By way of definition, chatbots are computer programs that are capable of maintaining a conversation with a user in natural language, understanding their intent, and replying based on preset rules and data. Since the Facebook story, chatbots (particularly in the voice-activated category) have become more visible, with the rise of Amazon’s Alexaselling in the tens of millions, Google Home’s Assistant, Apple making a late (but characteristically stylish) entrance to the market with its HomePod, and of course the odd amusing anecdote.

As we continue our voyage through how key technologies are being implemented in the FinTech ecosystem, I decided to follow up on my last piece with a deep dive into chatbots, the most ubiquitous (and immediately promising) application of A.I. currently being used in the space today.

Chatbots: A Brief History

As the science of Artificial Intelligence itself, chatbots have been around a surprisingly long time. The first chatbot, named ELIZA, was developed by MIT professor Joseph Weizenbaum in 1966. It worked by passing words users entered into a computer which then paired them to a list of pre-coded, scripted responses. The voice assistants gaining popularity today are also descendants of a technological child of the swinging sixties – IBM’s Shoebox.

So why, you might ask are they only hitting mainstream traction 50-odd years later? The answer is twofold:

1. Technological Breakthroughs: The pace of A.I./chatbot innovation has hit a tipping point, barriers to entry are lowering as underlying technologies become cheaper and easier to use
2. The Rise of Mobile Messaging: Messaging apps have exploded. They now see more monthly engagement than even the most popular social media apps and have become the dominant mode of communication for friends and families. This rise has also coincided with “App Fatigue”, with comScore data pointing to users spending just 85% of their time on 5 or fewer apps.

Understanding the Tech

There are two basic buckets into which chatbots fall: Scripted bots and Natural Language Processing (NLP) bots. Scripted bots, the simplest form, follow predetermined paths with decision-tree answer options which the user can select from. Financial players looking to implement more complex use cases (like customer service or advanced analytics) into their business-models however, NLP bots are the key.
Taking a look under-the-hood, NLP chatbot technology is underpinned by three basic concepts:

1. Intents (what the user is trying to say),
2. Entities (what information did the user give) and
3. Utterances (how users can ask for something).

The first two are used in structuring the bot, and the last facilitates training the application to improve over time. The details of the technology, while fascinating, are beyond the scope of this piece, but for a comprehensive and easy-to-read look at the magic of chatbots, Amir Shevat’s “Designing Bots” is a sensational choice.

The FinTech Chatbot Landscape

Why are chatbots receiving such a warm reception in the Financial Services industry? Well, in a word: Millennials.
As we’ve discussed before, the demographic shift to a generation of consumers who’ve grown up on and had their expectations set by technology poses the biggest challenge to the Financial Services industry in a multitude of ways.

Chatbots in “Traditional” Finance

For the enterprising financial firm Chief Digital Officer, chatbots offer banks a huge range of benefits, including:

1. Slipping seamlessly into the communication ecosystems of users (most people are on messaging apps, anyway)
2. Personalized, 24/7, high-quality user experience at scale
3. Streamlined operations
4. Rapid customer query resolution (done more cheaply than alternative methods)
5. Positioning them as forward-looking and tech-savvy

It’s these reasons (and more) that have led to their swell in popularity. Indeed, chatbots are one of the A.I. technologies that traditional banks have latched on to without much resistance.
Here are five examples that prove exactly that:

1. Erica (Bank of America): BAML has taken an aggressive lead in the race to integrate A.I. chatbots into financial services. The soon-to-be-released Erica, a voice/text-enabled chatbot for customers which acts as an “intelligent digital assistant” to help make smarter banking decisions, is their baby. Erica can send notifications, spot areas where customers can save cash, update them on their FICO score and facilitate bill payments.
2. COIN (J.P. Morgan): J.P. Morgan’s application of chatbot technology is inward-facing. COIN, short for Contract Intelligence, is able to do in seconds what took staff 360,000 hours of work, streamlining back-office tasks and slashing costs.
3. Eno (Capital One): Eno, Capital One’s gender-neutral text-enabled chatbot helps customers manage their money via their smartphone. They can ask questions about balances, transaction history or pay bills. They also launched an Amazon Alexa skill, so customers can get voice-based help.
4. MasterCard: Not wanting to miss out on the conversational finance wave, MasterCard released a Facebook Messenger chatbot which enables customers to access their accounts while they chat with their pals.
5.  American Express: Like MasterCard, AmEx’s bot facilitates account access via the Facebook Messenger platform. With real-time purchase notifications, benefit reminders, contextual recommendations and “add a card” features, the range of applications is expansive.

Other Noteworthy FinTech Chatbots

Of course, the big boys aren’t the only people heavily invested in this innovative technology. Venture-backed start-ups around the globe are in an arms race to create chatbots that solve myriad customer problems in the slickest way possible. Below are some of the front-running examples:

1. Albert – “Financial genius” Albert helps customers with a range of financial needs, ranging from savings to financial planning.
2. BOND – BOND is a bot designed to understand a user’s financial goals and help you hit them in the shortest time possible. The A.I. engine optimizes spending, refinances loans, chooses investments and creates auto-pilot savings!
3. Circle – Five-year-old Circle enables users to share money instantly, securely and for free, regardless of location in the world. Circle’s backed by big money and recently splashed out $400m on a crypto exchange – stay tuned.
4. Cleo – London-based startup Cleo offers a chatbot (or “digital financial assistant”) to help users manage their finances. They have just expanded to the U.S.
5. CogniCor – Barcelona-based CogniCor is an “intelligent version of Apple’s Siri, or Amazon’s Alexa, for the financial industry”. This A.I.-powered chatbot enables Banks and Insurance companies to drive automated pre and post-sales in natural language, with their customers.
6. Freda – HighRadius’ very own chatbot, Freda, was recently released to assist Credit and A/R teams. With the ability to answer over 5,000 new inquiries for A/R customer management and reporting, it has never been easier to get answers, run analytics or predict the future in the credit-to-cash business.
7. PayKey – Tel Aviv-based PayKey took a novel approach to chatbot integrations: by creating a Mobile Banking Keyboard which makes it possible for banks to give their customers the ability to initiate and complete a range of financial services from within any application.
8. Personetics – Cognitive banking bot Personetics is an A.I./predictive analytics tool that helps banks personalize customer relationships. They currently have 45 million+ customers dotted across the globe.
9. Plum – Plum is one of a number of fintech startups reimagining how we manage our finances online. Unlike competitors that exist primarily as standalone apps, the London startup (for now, at least) has decided to bet big on Facebook Messenger.
10. Truebill –Truebill makes tracking, paying for and canceling subscriptions and bills easy, by combining all of them in one place and automating their management.
11. Wally – Wally was built with a mission to create a “seamless, and intuitive tool to manage your personal finances”. They rank #1 Finance App in 22 countries, so it looks like they’re doing a good job.

Opportunities Still Abound…

While the majority of the start-ups mentioned above were in the intelligent money management/personal assistant niche, the range of potential use cases doesn’t end there. It is very easy to imagine a whole host of additional segments of the customer experience where chatbots could be deployed. Off the top of my head:

1. Lead Generation – Chatbots could quite conceivably be programmed to pose questions to potential customers. This would gather valuable data for the institution using them, while automatically filtering out unqualified prospects and passing through high-quality leads directly to sales reps.
2. Cross-Selling Financial Products – As their understanding of the customer increases, bots will be able to tap this data and serve up precision-target content and offers. Personalized promotions delivered at just the right time in the customer’s favorite messaging app are a recipe for excellent conversion rates.
3. Process Automation – a lot of banking tasks/services could benefit from bot-mediated automation. Business Insider estimates that the average cost of enabling a customer transaction via phone is around $2.50, while for a digital transaction (online or on mobile) it comes down to a mere $0.17.
4. Predictive Product Offering – As bot technology gets better, it is a matter of time before they develop the ability to “learn” individual customer preferences, thereby being able to glean where they are in the buying cycle and offer products which suit them perfectly.

The most effective bots in the future will be the ones that operate in at the intersect of A.I., analytics, and chatbots into a single platform. This will enable the “cognification” of the entire user experience. Bots this sophisticated would provide customers with personalized financial guidance, conversational self-service, and automated programs to help them successfully meet their financial goals.

Conclusion

The evolution of A.I. is now alive and kicking in the FinTech world, and chatbots are only a faint splash in a tidal wave of progress. That said, artificially intelligent chatbots are rapidly transforming the financial services industry, and their importance to our sector shouldn’t be underestimated…

Original Source: LinkedIn