As more and more companies start implementing some type of artificial intelligence, machine learning or automation, it can sometimes seem like these technologies are being used just because everyone else is doing it.
The potential for AI and automation are huge; McKinsey forecasts that the potential economic impact of automating knowledge work could be between $5.2 trillion and $6.7 trillion by 2025. However, there’s still a lot of room for growth. And as we move into the new year, AI will continue to evolve beyond chatbots to something more akin to a digital colleague.
The Evolution of the Chatbot
While Alexa, Siri and their ilk might seem like modern-day inventions, the concept of the chatbot has actually been around since the 1960s with the creation of ELIZA by MIT professor Joseph Weizenbaum. Of course, ELIZA was fairly simple by today’s standards. And in light of what is possible today, Star Trek’s talking computer revealed the stunningly prescience of the series’ creator.
Chatbots and voice assistants as a means of engaging users are proliferating for a good reason. For instance, almost all of human knowledge is now available with a simple voice command. “Where’s the closest Chinese restaurant?” “How much money is in my checking account?” “Who discovered the vaccine for polio?” The ease and convenience of on-the-go assistance and information caught on quickly and continue to grow. For example, U.S. smart speaker ownership rose 40% in 2018; with an adult population of 253 million people, the U.S. employs 133 million smart speakers.
As Chatbot Technology Adoption Continues, A New Approach Is Needed
Gartner predicts that artificial intelligence (AI) will be a mainstream customer experience investment in the next couple of years. In fact, 47% of organizations will use chatbots for customer care and 40% will deploy virtual assistants.
This is great news for overworked help desk and customer service staff, but chatbots can only go so far. That’s because not all chatbots are created equal. Some chatbots are touted as possessing artificial intelligence, but it turns out that they are actually scripted chatbots with pre-programmed logic and flows. They can still perform important functions, from offering technical help and scheduling appointments to assisting with sales and marketing efforts by signing users up for webinars or newsletters, for instance.
However, these handy assistants are experiencing a change in popular sentiment. When chatbots are ineffective, only slightly better than a pre-recorded phone tree, customers will revolt and demand to speak with human agents. Forrester was predicting this last year, suggesting that organizations would begin to reverse course and start making more efforts and investment in humans as a result of the backlash.
At the time, Forrester analysts said, “Customers will lead a community-based revolt against corporate chatbots. Human resistance against ineffective chatbots is on the way, and a groundswell of jaded customers will crowdsource tips for end runs around chatty chatbots. A movement similar to the GetHuman movement from 2005 will start.”
We’re starting to see that backlash, but organizations will still need to find a solution to the reason they deployed chatbots in the first place: the talent gap is real. In a Gartner survey, 63% of senior executives indicated that a talent shortage was a key concern for their organizations, particularly for IT teams. Gartner, in fact, characterized the talent gap as the top emerging risk for organizations. Enter the concept of the digital colleague.
Meet Your New Digital Colleague
Imagine you’re currently working in a NOC (Network Operations Center), where people have to monitor multiple dashboards. They have to often escalate it to L2/L3 workers to troubleshoot incidents, make war room calls regarding root cause and then actually take the manual steps to remedy them. What if you could take some of that manual work off the NOC workers and give them a digital colleague to assist? The digital colleague can reduce the number of errors, find root cause, and file service tickets and wherever possible automate the incident resolution. Not only would it reduce operations costs, but it also frees up these workers to focus on more meaningful endeavors. That’s the concept of the digital colleague.
This holds true for ITSM, ESM or any function in need of assistance. A digital colleague for Service Desk will converse with end users in natural language through multiple channels, understand the user intent, map it into service catalog, ask clarifying questions and automate resolution of up to 50% of service requests, thus taking load off operations staff.
The digital colleague would need to be AI-based to prevent backlash from end users, since ineffectiveness was the primary complaint against chatbots. Digital colleagues are specifically designed for and trained in specific applications. They are often coupled with automation to resolve end-user issues instantaneously. This better focus leads to higher accuracy.
A Well-Rounded Solution
AI, machine learning and automation are transforming our world, although some of their applications are clunkier than others. Voice assistants are wildly popular, and consumers find chatbots acceptable – but only when they are effective in resolving issues. Faced with a potential chatbot backlash and a concurrent talent shortage, companies need to up their game. Digital colleagues are the more effective alternative to the typical scripted chatbot that lacks true artificial intelligence. They can be trained for specific tasks and combined with automation for a higher “first time resolution” rate. This serves the company, its employees and its stakeholders.
About the Author
As chief product officer, Dr. Akhil Sahai heads product management and marketing at Symphony SummitAI. He has more than 20 years of experience in product strategy and management marketing and business development. He is a product leader with extensive experience managing and scaling high-growth businesses with multi-hundred million-dollar product portfolios and SaaS-based early-stage startups.
Featured image: ©sdecoret