Whether they find them helpful or pesky, almost all consumers have grown accustomed to the presence of chatbots, the conversational artificial intelligence programs that use voice or text to convincingly simulate how a human would answer questions and provide advice or assistance. But soon, employees and job applicants will grow equally familiar with the presence of the technology, which has many applications in the world of HR.
Companies have already begun using the technology to enhance HR service, create better and more personalized experiences for job candidates and employees and increase productivity.
Talent acquisition is one area where the use of AI has become particularly widespread in recent years, according to Forbes. Because recruiters have a limited amount of time to spend monitoring the careers site and responding to applicants' queries, chatbots have been used to handle initial screenings, schedule interviews and answer any frequently asked questions that candidates have. All of these applications allow recruiters to devote their time to more sophisticated, high-value work, such as sourcing more important roles or negotiating an offer with a candidate who has not yet committed.
In addition to guiding employees through the onboarding process, chatbots are also being used in HR self-service centers to automate certain high-volume tasks, such as changing an address or updating benefits information.
Many more potential uses for the technology are visible on the horizon, according to HR experts.
"In the past, an employee with a question about how to get something done would ask a knowledgeable colleague for an answer," Arthur Franke, a data scientist and director of data and analytics at KPMG, told SHRM. "In the future, they'll ask a conversational agent, and artificial intelligence will answer their question."
How to effectively implement chatbots in your own HR practices
A good chatbot makes users feel as if they are not speaking with a chatbot. When conversational agents are functioning properly, the discussion feels like any exchange the user might have with a human, rather than a mechanical program. To achieve this result, the chatbots must be trained with algorithms and data that is continuously monitored and audited to guard against bias, and the bots themselves must be programmed to continuously learn about individual users to enhance personalization.
Some of the less sophisticated bots available still have to route the more challenging questions they receive to a human "co-pilot" who can read and respond to the query. Today's bots are, in many cases, simply functioning as a conversational interface for a database, amounting to little more than a human face slapped on a Google search.
But as chatbots improve their ability to personalize conversations, users will grow more comfortable with them. When people feel as if they are having a real conversation, they are more likely to see chatbots as conveniences rather than annoyances.
There are some other barriers to adoption that employers looking to leverage conversational AI technology should be careful to avoid.
For starters, users will feel like chatbots are only complicating the process if they need to log in to a platform to use them. For this reason, bots should be enabled for mobile devices and popular workflow tools like text, email and Slack. Organizations should use inclusive design practices to make sure that the bot interface is accessible to all potential users, regardless of their working situation, capability or language proficiency.
Employers should also be clear and upfront about their implementation of conversational agent technology. Employees are likely to become frustrated if they think they are talking to a real person and then encounter an error message in the middle of their discussion. If a chatbot's personalization abilities are sufficiently advanced, even users who know they are speaking with a bot may say "thank you" to it once their problem is solved.
In these early stages, experts are recommending that experts start small when it comes to bot implementation, so that they can learn from the experiment before scaling up.