Meaning and context - the secrets to a successful virtual banking agent

June 11, 2019

Conversational AI is designed get to the heart of what customers really want - it could mean the difference between a great experience and them walking away with a less-than-favorable impression of your brand.

Understanding context is essential for building and maintaining relationships. Even a single word out of context can often be all it takes for a conversation to turn into an awkward mess of misunderstanding and erroneous responses.

Imagine visiting a bank for the first time to speak with a financial advisor about your savings. You lay out your hopes and dreams for a future of financial security, only to be baffled by a completely unrelated answer about pouring your savings into high-risk investments. You realize that the advisor has based their answer on a single word or phrase of what you said, and disregarded the rest of the information. Not exactly reassuring behavior from someone who would like you to trust them with your money.

In essence, this is what happens all too often when customers try to engage with a chatbot that is built on sub-par technology. By attempting to match only a piece of the information with a probable answer, the resulting response can seem almost completely random. Without fully understanding the situation and considering the context, responding appropriately is almost impossible.

Your virtual agent is a customer’s first impression

Gartner had previously predicted that, by 2020, chatbots and virtual agents would power 85% of all customer service interactions. A bold claim that, while unlikely to be met in the next six months, is still bolstered by an incredibly bullish trajectory that the industry has been riding. It’s not a question of if we’ll hit this number, only one of when. If you are a bank or financial institution, chances are high that you will very soon (if not already) have a virtual agent as first-line support for your customers.

This means that most of the time, a virtual agent will be the first impression customers have when they engage with your company. These customers are faced with more options to choose from than ever and, as a result, their expectations are at an all-time high. You only get one chance to impress, and if they feel that your brand doesn’t get them, they have been shown to have a strong aversion to contacting a company multiple times or sharing the same information more than once.

As a bank or financial institution looking to succeed in the digital age, the stage needs to be set for meaningful interactions with your brand. If consumers have to engage with a chatbot solution that doesn’t understand them, chances are that many a first impression will be ruined. By disappointing prospective customers, you run the risk of them taking their money elsewhere, especially if competitors provide a superior experience.

Recent advances in natural language understanding (NLU) have paved the way for conversational AI that serves as a remedy to poor first impressions. As the underlying technology for a new breed of virtual agents, customers are no longer met with a program that is limited to trying to match keywords in queries with the most-likely scripted answers. Instead, thanks to proprietary technology like Automatic Semantic Understanding, they engage with a sophisticated solution that asks questions and seeks to fully understand the situation before providing relevant information or assistance.

Banks in the tech-savvy Nordic countries are already using conversational AI-powered virtual agents at enterprise-level scale to transform customer experience. A noteworthy example is the largest bank in the Nordics (by Market Cap as of June 2019), DNB, who used’s solution to automate 51% of their online chat traffic. Within six months, DNB’s virtual agent was automating 17% of customer interactions across all channels, including phone and e-mail, and doing the work of nearly 1,000 full-time employees - all without customers ever needing to interact with a human.