Gone are the days when virtual agents where just a concept on drawing board. Today it is a reality that is quickly augmenting and replacing the need for human support personnel. And everyone, including small and large companies, have embraced this trend by employing virtual agents in large numbers. For instance consider SFR, a subsidiary of mobile communications giant Vodafone, which has employed virtual agents to carry out nearly 750,000 conversations a month. Or consider the online auctioneer eBay, which is facilitating approximately 200,000 customer conversations a day with virtual agents across six countries. And the reason behind this growing popularity is not the fact that virtual agent’s cost only 1/10th of what live human assistance does. But rather, it is because customers prefer virtual agents and the self-reliance that it brings in.
In fact as recent survey by Forrester, a meager 28 percent of consumers in the US prefer human support. And experts believe that these numbers will further dry up, as todays tech-savvy younger generation takes the place of baby boomers as the largest segment of the market.
But is artificial intelligence-based virtual agent a better customer service option? And can it completely replace human assistance? The answer is no. Yes, if utilized effectively, automated customer support can provide faster and easier resolution to customer issues. And it can drastically reduce the amount of time an agent spends on resolving common issues.
However, customers don’t necessarily call in to seek answers to general issues. A lot of the time, they will have real concerns that need special attention and a certain level of understanding, which is clearly lacking in the case of AI-based virtual agents.
Furthermore, virtual agents have no real intelligence other than trying to make out what was said by matching it with the available knowledge base. Even if the automated agent matures to such an extent that it can understand everything that you say, it doesn’t really have what it needs to take action and eliminate common service problems, such as billing errors.
The road ahead
Yes, thanks to the latest advances in technology and use of natural language processing (NLP), virtual agents can now interact with customers in a natural conversational style, interpret requests, and then respond appropriately. But still there are limits to this technology and there is still a lot to come in order to reach the optimal level. And experts are of the opinion that the best way to make the most of the available technology is by using it in conjunction with human agents. This way, you can perfectly compliment human agents and make them focus on the tier-two and tier-three support issues, which are more complex and higher value tasks.