HR and Voice Recognition Technology
HR industry vendors have started moving their voice-activated applications into real-world use.
Hey Alexa, what’s the head count in our finance department?
Siri, can you submit a time-off request for me?
Hey Google, who’s available to trade shifts with me next Friday?
The ability to make this type of query has long been a staple of our personal lives, but the prospect of bringing similar technology to the workplace has largely gone unrealized. That’s beginning to change as voice recognition technology has evolved and as HR industry vendors have started moving their voice--activated applications out of the demo stage and into real-world use.
Vendors, including Ceridian and IBM, now have clients using voice response as an alternative to typing on keyboards or swiping on mobile devices when conducting common HR transactions or accessing data around key workforce metrics. These vendors and their customers are betting that the time for voice has come and that growing interest in interacting with technology via voice commands at home will finally translate to the workplace.
Experts say advances in voice recognition technology have created momentum for bringing the ability to talk with smart devices into the office. The proportion of search queries made by voice on Google has passed 20 percent, and almost half of adults in the U.S. now report using voice-controlled assistants such as Amazon’s Alexa or Apple’s Siri in a recent Pew Research Center survey.
Some experts see the use of voice response in HR to be at a similar stage as adoption of mobile technology around 2010. “We’re still in the infancy of voice recognition, but companies are starting to see where the technology can have real impact in time savings and in creating new efficiencies,” says Chris Havrilla, vice president of HR technology and solution provider strategy at Bersin, Deloitte Consulting in Atlanta.
Helen Poitevin, an analyst specializing in HCM technologies with research and advisory firm Gartner, says adoption of voice-activated technology for business purposes remains higher in some Asia Pacific countries, India and China than it is in North America, because those regions are more accustomed to using voice to interact with technology and because it can be easier to speak the queries rather than type them out in the languages used in many of those areas.
But Poitevin says her research also shows growing use of the technology in the U.S. for such tasks as time-tracking to record hours dedicated to business projects or to different client accounts.
The Time-Savings Advantage
Minneapolis-based Ceridian is at the forefront of the voice-response movement, having evolved voice capability within its Dayforce Assistant tool from the demo stage to real-world use by select customers. The voice app is designed to help employees conduct common HR transactions faster and more efficiently, says Andrew Shopsowitz, director of product management for Ceridian. Workers can use the technology to check schedules and swap shifts with co-workers, for example, as well as to review paid-time-off (PTO) balances or to request time off.
Using voice response can save employees significant time over previous ways of conducting these transactions, Shopsowitz says. For example, in the past, employees would have to log into the tracking system, find a page dedicated to PTO or vacation time, list a reason for a time-off request and then submit the query. If a request was approved, they’d need to take such additional steps as updating an e-mail out-of-office notification and declining or canceling any planned meetings for those days.
Now, employees can simply say to a Google assistant device, “Google, open Dayforce” and then ask to submit a time-off request. Once that’s completed, the voice app will follow up with “Would you like me to set up your out-of-office and cancel your meetings?,” and the Dayforce Assistant could execute those tasks automatically through the device.
“We believe it won’t be long before it’s an expectation from employees that they have an option to interact with work systems via voice rather than a keyboard or computer screen,” Shopsowitz says.
SAP SuccessFactors also has voice-activated assistants in development, says David Ludlow, group vice president of product management with the vendor. The voice tools, dubbed “conversational AI,” are currently in beta testing with early-adopter customers and are designed to respond to employee questions around HR benefits, policies and more, he says.
“We believe there’s a significant amount of HR content and services that can be queried using voice commands that can quickly present end-users with answers to their questions,” Ludlow adds.
Ben Eubanks, a principal analyst at Lighthouse Research, an HR advisory and research firm in Huntsville, Ala., and author of Artificial Intelligence for HR (Kogan Page, 2018), says he has a client currently working to build voice recognition into a core HR system to make it easier to track key HR metrics. Eubanks compares the capability to Alexa’s “flash briefing” function, which gives users a daily overview of news headlines and other content customized to their preferences.
“The technology is designed to allow HR leaders to ask the system things like ‘What’s our headcount in XYZ division?’ or ‘How many people are currently on maternity leave?’ and instantly get results back,” Eubanks says. The capability saves them from having to log into the system to check software packages or spreadsheets to extract that data, he says, and also allows HR to run a report based on the initial verbal query to send to key stakeholders.
“Rather than having to click through screens that might take minutes,” he says, “HR leaders could get data around the most common metrics they track within 10 seconds using this technology.”
What the Future Holds for Voice Response in HR Some experts believe HR lags significantly behind other business functions when using voice or video in ways that can better engage employees or create new process efficiencies. John Sullivan, Ph.D., an HR thought leader and professor at San Francisco State University, says if HR intends to capture the attention of workers in the future, it will need to offer more content in alternative formats such as voice, video or podcasts. Sullivan foresees a future for voice response not just in answering FAQs from employees involving topics like benefits or pay, but also for policy questions or to provide decision support for line managers around common but challenging scenarios such as hiring, performance reviews, sexual harassment and layoffs. Other industry experts see potential applications for voice response in recruiting—for example, intelligent systems that could record interviews and give hiring managers feedback on improving their technique—as well as for onboarding, gauging employee morale through voice recordings that are automatically transcribed to text and for automated search functions. Ceridian sees potential in applying voice response to workforce analytics. “We don’t think you should be tied to having to access key reports or spreadsheets on mobile devices or desktop screens,” says Andrew Shopsowitz, director of product management for the vendor. —D.Z. |
Limitations of Voice Response
Cyberspace is rife with cartoons and anecdotes about how Alexa or Siri has misinterpreted user questions and returned with off-base or even hilarious responses. But the reality is that the natural language processing (NLP) technology that undergirds voice recognition—a process that transcribes speech to text—has evolved considerably in recent years and features an improved ability to understand the intent of language as well as variations in phrases with similar meanings.
Christopher Phan, a technical expert and product owner of the Dayforce Assistant voice response tool, says Ceridian built its own custom NLP capability on top of leveraging Google’s NLP technology to help interpret language and phrases common to the HR domain.
For example, an employee might ask the Dayforce Assistant “Can you help me trade my shift?,” but the assistant could mistakenly hear “Can you help me train my ship?” Ceridian’s own HR domain-specific NLP logic would be able to determine that the word “train” might have something to do with “training” but that the word “ship” likely has little to do with HR and would interpret it to mean “shift” instead in this context. “When the technology compares ‘train my ship’ versus ‘trade my shift’ contextually, it knows it’s likely the latter and can provide the appropriate response to the employee,” Shopsowitz says.
Voice response also must be equipped to handle global and regional differences in language, Phan says. For example, vacation or sick days are often called different things around the world, and other common HR terms can have various meanings. “People often call a work schedule a roster in Australia, so we trained our technology to understand what a roster means in the context of work-scheduling queries,” he says.
Poitevin says that while the accuracy of voice recognition has improved, many vendors still face obstacles once a user’s voice commands have been translated to text by NLP. “Once you have that text, you still have to pass it on to a back-end system to interpret it and respond correctly,” she says. “I still see a lack of maturity in the ‘intent generation’ and integration capabilities in many vendors’ tools that are needed to deliver accurate responses to queries.”
Despite such challenges, many believe voice will continue to make inroads in HR. Ludlow of SAP SuccessFactors says the way that employees access HR information and services in the future will inevitably evolve.
“Employees will no longer have to search for something like a policy document, download it and then look for the relevant information within it,” Ludlow says. “Instead, they’ll be able to ask for and be delivered the relevant information within the document with voice commands without having to hunt around. It requires a mindset change in moving from a typing channel to a voice channel, but there’s momentum in that direction.”
Dave Zielinski is a freelance business writer and editor in Minneapolis.
Illustration by Mike Austin.
An organization run by AI is not a futuristic concept. Such technology is already a part of many workplaces and will continue to shape the labor market and HR. Here's how employers and employees can successfully manage generative AI and other AI-powered systems.