Conversational Bots – The Third User Interface?

Natural language has become an increasingly popular means of communicating with computers. In 2019, rather than entering individual keywords into Google's search bar, we are asking questions in a conversational manner, such as "What should I have for dinner?" or "How can I open a bank account?". The expectation today is that the computer will handle the rest and find recipes, directions, facts, and recommendations for us. This shift is also evident in website browsing and the use of mobile applications - People are no longer willing to waste their time searching for information on portals but expect to effortlessly express their concerns and ask questions using their own words. The paradigm has shifted: Information should now find its seeker, not the other way around. How did we get here?

There have been several stages in human-computer communication. The initial user interfaces were command-based, like the DOS prompt, which utilized command lines for program execution and navigation. These CLI interfaces, however, were swiftly replaced by graphical, visual interfaces built on icons and controlled by a mouse. In the graphical user interface, users found it easier to execute commands without knowing the system in detail or memorizing commands and syntax. Fastforward twenty years, mobile devices and touchscreens enabled mobility and navigation through a graphical interface with a simple touch. Although mobile was clearly going to be the future, the mobile app market quickly became overheated – millions of apps became available in an instant, and within a decade, people grew tired of constantly downloading and uninstalling software from their smartphones. Eventually, it was the simple messaging apps like WhatsApp which emerged as the most popular, even surpassing traditional social media usage years ago. We are entering a new era of user interfaces where machines are expected to understand humans in their natural environment, learn human skills, and thus appear intelligent. We are now seeing videos uploaded to YouTube featuring babies uttering their first words to devices like Alexa. What expectations might this new generation have regarding user experience when they buy their first smart devices in ten years time?

Conversational assistants, such as Siri and Alexa, are designed to interpret text and speech using artificial intelligence and machine learning. These media, which can be either text or voice-based, are more commonly known as chatbots. I prefer to call them conversational bots. Conversational bots are digital assistants that serve both internal and external customers of an organization across multiple channels simultaneously, are available 24/7, capable of retrieving information from backend systems, and possibly performing tasks like resetting passwords or generating emails. Development is rapid. Gartner predicts that within the next two years, 80% of businesses will be leveraging chatbot technology.

While initial use cases have been seen in customer service and the automation of simple FAQ answers, the real potential may lie in internal business processes: HR, IT support, legal departments, marketing, and so on. In the future, we may also use personal bots that converse with other bots, exchanging information and handling routine tasks on our behalf. We likely cannot even imagine the future use cases and their impacts yet. The shift will be towards conversational organizations where all user interfaces will be replaced by intelligent conversational bots. Instead of “mobile first”, many companies have already chosen a “bot-first strategy”, preparing for the era of conversational user interfaces.

In the customer experience context, the chatbot often facilitates a crucial interaction between the brand and the customer. As conversations with the bot convey personality and emotion, many companies have found it important to invest in conversation design and tone of voice from the beginning. How does the brand want to sound? Is the chatbot female or male, or neither? How does the chatbot respond when a customer asks about the company's stance on climate change? These are questions that need to be addressed early in the design of conversational user interfaces.

As conversational bots become more prevalent, some job tasks within companies will likely decrease, but on the other hand, new roles have already emerged: AI trainer, bot engineer, bot insight analyst and conversation designer, to name a few. It is also conceivable that bots will free us from many routine tasks, allowing us to focus on more creative and challenging assignments. Answering FAQs and checking simple information can often be dull and exhausting work. A bot for many might be a welcomed colleague, freeing up admin time to handle more challenging customer inquiries.

The adoption of chatbot technology should be seen as a comprehensive and long-term development phase within a company. A recommended approach is to examine internal processes, identify a specific use case, and build a pilot project based on that. It is advisable to start with internal processes first, familiarize oneself with the technology at a steady pace, determine what works and what doesn't, and then expand the bot's use to customer interfaces. An agile approach, experimenting, building prototypes, and monitoring their use while staff develops their skills with the bot, is recommended. After the initial trials, use cases are likely to pop up throughout the company. Training a conversational bot is an ongoing process after the pilot. As the bot becomes smarter, there is usually demand to expand its area of responsibility.

The transition to conversational interfaces may take years, even decades, but it is worthwhile to start preparing now by considering potential use cases and problems in the organization that a conversational AI bot could possibly solve. Early experiments with prototypes help incorporate the technology into daily routines and build internal expertise. The earlier potential failures are experienced, the better. Introducing a chatbot is much like hiring a new employee – investing in onboarding and preparation is crucial at the beginning, but the expertise pays off multiple times in the future. Chatbot is the next user interface.

(Note: In 2024, as we are already living with Copilots and AI enriched applications, it's once again time to wonder what's next. As AI is guiding more and more of our decision making, it seems likely that the fourth user interface would one where humans become if not passive, less proactive. It sounds science fictionish, but there may be a point in time when our machines will know us so well that even words won't be needed. Communication may evolve into something not far from mind reading. Possibly, the next user interface is AI itself).