Do a Google news search for chatbots and chances are you'll see some doomsday predictions that "the robots are coming!" This messaging might encourage site visits and clicks, but the truth is that automation in the customer service and technical support industries is nothing new. For decades, service desks and customer support organizations have tapped technology to address common questions and problems with minimal human intervention. The latest generation of artificial intelligence, like IBM's Watson and chatbots within help desks, is taking this to a new level. Because these chatbots and AI-based reps are trained to be more effective, responsive, and anticipatory in every interaction, they have the potential to address a much wider array of issues than their automation counterparts.
The concept of chatbots is essentially the same as that of early-generation support technology&mdashdeflecting questions before they get to humans, whom companies must pay. Knowledge bases offer many of the same efficiency benefits, but when an article doesn't address the specific problem a customer is experiencing, frustration quickly escalates. Chatbots, on the other hand, can understand the nuances in customer support requests and intelligently respond to what is being asked.
Another attractive chatbot feature is that companies and call centers can run analytics on interactions to figure out which responses are leaving customers satisfied and which leave a little to be desired. This type of analysis and training is difficult to replicate in human sessions for numerous reasons. And then there is the high turnover in the customer support industry, which requires a significant training investment. As AI technology matures, the continuous improvement element of chatbots could truly change how support is delivered while simultaneously driving down costs.
These changes will also impact the job expectations and qualifications for human customer service agents. It's not unlike what's happening in the manufacturing sector, where numerous straight-forward tasks are now automated by machinery. With AI-based tools able to resolve many of the incidents traditionally reserved for tier-1 support staff, the issues requiring human intervention will be more complicated or complex, with support reps requiring more specialized skills to resolve them. Not only does this change the job requirements for support personnel, it also impacts how companies and call centers recruit for and staff their service departments.
In its most basic form, effective support lies in connecting the individual experiencing the problem with the expert who has the solution. As previously mentioned, when chatbots fulfill the simple issues, the problems necessitating human involvement become more complex. As such, matching the problem with the solution will not be as simple as it sounds.
The valuable employees capable of resolving these complicated problems are not likely to be sitting around waiting for requests to come in. They might be seasoned engineers whose primary job responsibilities are not technical resolution. That's why companies need a way to broadcast for help from experts, allowing available people to resolve incidents without diverting resources away from larger projects and priorities.
Companies also need to wrestle with the perennial demand for better service, faster. As technology grows more specialized, problems can't always be addressed via phone or chat. For example, fixing malfunctioning systems on a ship out at sea might require the tech expert to view a program on the remote computer, inspect the physical equipment via camera, perform tests on supporting applications, etc. Time and budget constraints might make flying someone to the physical location an unrealistic option in all but the direst of cases. This is where remote screen-sharing and mobile camera-viewing come in. Many remote support providers offer multiple technology options so support staff can access and see both the software and hardware the end-user is seeing, and I expect these solutions to evolve and adapt in the years ahead.
AI + U
The impending robot revolution makes for good headlines, but we're still a long way off from chatbots fielding all common support requests. AI technology has not been adopted as far as it needs to for chatbots to handle a significant volume and variety of issues. But people have really embraced chat as a support model, which leads me to believe that they will be open to more chatbot interactions.
Companies can't forget the importance of the human element, however. If an end-user hits a dead end with a virtual agent, he will be far less likely to interact with a chatbot for a future issue. As such, these machines must be taught to recognize when someone is growing frustrated with their responses and direct them to humans as soon as possible.
While these and other AI elements are still evolving, the industry is going to change significantly when chatbots become more mainstream. What can your organization do to prepare? A good place to start is modernizing the customer service department and ensuring the support tools and technologies can support cross-platform and multi-device needs today and into the future.
Donald Hasson is director of product management at Bomgar.