There’s seldom a day goes by when I don’t see news about artificial intelligence (AI). And rightly so—the maturity of AI is improving rapidly, and the potential for transformation is unprecedented.
Chatbots, especially, are getting a lot of attention. Some chatbots are relatively simple and limited to answering questions with pre-loaded answers, while others are intelligent and possess the ability to learn and improve their responses. The latter, when purpose-built to provide customer support functionality, are often referred to as virtual agents.
Virtual agents will significantly alter the way employees interact with support desks. At present, however, there is no standard method for evaluating virtual agents and several people that I’ve spoken with in recent months have told me that they are unsure how to compare one virtual agent to another.
Organizations need to look for virtual agents that have a conversational interface that enables users to present their issues in their own words, not the language of IT. Solutions should guide users to provide the information required to create fully formed tickets. With the right solution, issues such as password resets can be resolved without involving human agents.
To help to evaluate the right virtual agent solution, here is a short list of some of the functionality that I believe organizations should deem as mandatory in a virtual agent:
- Offers a wide range of channels for end users to report issues—A virtual agent must enable end users to report issues conveniently by communicating using messaging apps and other tools with which they are already familiar.
- Supports the design of conversational responses that collect all the requisite information—Since users are not typically capable of describing issues using the technical terminology required by support teams, a virtual agent must be capable of filling in the gaps to enable productive conversations with end users.
- Understands requests accurately and precisely—An essential requisite of any productive interaction between two people is the ability for the listener to fully understand the information with which they are being presented. The same is true of a conversation between someone reporting a technical issue and the virtual agent attempting to assist in resolving the issue. Organizations should look for a solution that supports these features and enables conversations with users to be completely understood.
- Reduces the workload on support teams—Once information has been collected, understood and digested, virtual agents must be capable of executing actions to address the issues reported by users, and do so in ways that minimize involvement by agents.
- Integrates seamlessly with a wide variety of support desk software—Virtual agents should be adept at exchanging information with most applications and tools used by support organizations.
- Includes an automation engine that implements actions in external systems—It’s important that a virtual agent is capable of autonomously implementing corrective actions in systems used by different departments across organizations.
- Increases agent productivity by following up on tickets automatically—Last, but definitely not least, organizations should ensure that the virtual agent they choose can mitigate the burden of chasing down end users, and freeing agents from other similarly mundane tasks.
Skimping on functionality is going to impede your virtual agent from transforming the support experience of users and enhancing the productivity and satisfaction of your help desk team. Be sure, instead, to choose a virtual agent that provides you with everything on this list.