Source: Hays Journal Issue 17 - click here for more information and to get your free copy
Voice search is fast gaining momentum, with some experts going so far as to say it could become the dominant method of search within years. While these claims are facing scrutiny, its impact is undeniable. Should HR leaders and organisations be preparing for this technology to enter the workforce?
While having conversations with smartphone and smart speaker virtual assistants such as Siri and Alexa is now an everyday occurrence for many, it’s fair to say their impact on the workforce has been limited so far. But as use of the technology grows, is that about to change?
Digital giant Google says that the proportion of queries on its mobile app and on Android devices that are made by voice has already passed the 20 per cent mark in the US1, while market researchers have found that 24 per cent of US households now own a smart speaker2 (which of course only responds to voice activation).
These figures are likely to go on rising and are indicative of a worldwide trend. It’s hardly surprising – asking for information and giving instructions by talking is usually easier and quicker than typing. And besides, the power of speech is something humans are uniquely good at.
Meanwhile, advances in machine learning and speech recognition driven by artificial intelligence (AI) mean that voice-activated and conversational technologies are getting better and better at communicating in a human-like way.
Matthew Cain of global research firm Gartner reports that “consumers and workers increasingly interact with applications without touching a keyboard”. According to Gartner’s 2018 CIO Survey, while less than four per cent of organisations have already deployed conversational interfaces (including chatbots), 38 per cent are planning to implement them or are actively experimenting with the technology3.
So, is it possible that voice technology could be used by HR to support workplace wellbeing and engagement? And what are the opportunities and challenges around its use?
Playing by the rules
Conversational technology’s breakthrough was with text-only chatbots – software that could mimic human conversation in writing. It was typically a customer service tool that could search a database to answer straightforward ‘frequently asked questions’ (FAQs). Next-generation chatbots are now voice-activated, but it is the FAQ function that many see as the most obvious HR use, freeing up HR professionals for conversations with staff that voice technology isn’t yet up to.
Professor Chris Ivory, Director of the Innovative Management Practice Research Centre at Anglia Ruskin University, says: “The AI behind this technology works best in HR areas that are rule-bound, ensuring compliance with law.” He adds that, freed from having to deal with these areas, “I see HR professionals increasingly moving towards more strategic activity – planning and implementing cultural change, implementing reward systems and so forth.”
Jonny Gifford, Senior Advisor for Organisational Behaviour at the Chartered Institute of Personnel and Development in the UK, has this to say about the “doom-mongers” who predict that automation such as voice technology will mean job losses: “AI replaces tasks, not jobs. So, typically, they cut some of the heavy lifting within jobs. Research that we released this April shows that, overall, people’s jobs are upskilled by AI.”
Gifford points out that a clear advantage of an automated approach to FAQs for employees is they can get answers “as and when they need them”. He says: “You don’t have to wait until Monday when the HR department have time. It would be the same with training delivered in this way. And you wouldn’t have to sit through a part of a course that you already know. A virtual trainer doesn’t mind if you skip that bit.”
Jeff Adams, founder and CEO of US-based Cobalt Speech, has worked on speech and language technology research for more than 20 years: “Voice technology is simply a more natural interface for people to interact with systems, databases and machines,” he says. “I can imagine that you would pick up a telephone and dial a number that connects you with an ‘assistant’ that can talk to you about arranging annual leave or logging your hours.”
Adams, whose career includes time spent developing the Amazon Echo smart speaker, continues: “You could ask about benefits, where to find something, or about rules and regulations. It would be easy to get information in a friendly way.”
Meanwhile, market research and insights business Kantar Analytics has been looking at how voice technology and analysis can help HR teams to gauge employee morale and sentiment. The company recently used this approach at a large UK bank to assess the management team’s feelings about the company’s new corporate vision.
“Hundreds of face-to-face interviews were organised,” explains Tom Evans, Director for Data Science at Kantar. “Historically, analysis of the emotions of those interviewees would mean listening to all the conversations and writing lots of notes about what was said and how things were said. That’s difficult to do on a large scale, and expensive too.
“We applied the technology and analysis tools we’ve developed with US company Affectiva to recordings of those meetings and were able to retrieve the same insight as the traditional method, but more quickly and cheaply at that scale.”
Other potential voice technology applications Adams can envisage include a system that would not only take notes in a meeting, but also automatically red-flag proposals generated by that meeting that don’t comply with company policy or employment law. “Some people might find that a little Big Brotherish,” he acknowledges, “but most of us are more afraid of doing something inappropriate out of ignorance and would appreciate the protection.” Big Brother or not, this compliance application would mean another task made easier for HR professionals, who could spend more time creating good company policy and less time enforcing it.
He adds: “A number of companies are already using systems that monitor employees. For example, call centres have automated systems listening in on calls to ensure that agents are being polite and are giving the right answers. You can imagine similar applications for sales people, to make sure they aren’t making inappropriate claims to potential customers. That might be useful for a company, but employees might find that intrusive – so a balance has to be struck.”
How voice technology is introduced to a workforce could be just as important as its application. Alan Price is Group Operations Director at HR and employment law consultancy Peninsula, which operates in the UK, Ireland, Australia, New Zealand and Canada. He says: “Consider the impact on your existing processes and staff. Voice recognition software may engage better with a tech-savvy younger workforce than an older group who aren’t used to this level of technology. It is therefore a good idea to implement a test period first in order to see just how useful the software will be. During this time, staff should regularly be consulted to outline if they are finding the technology useful or if it is having a negative effect on their daily activities.”
Adams agrees that the technology should be phased in with transparency. And, where possible, it should be available as an option, “but with the HR department’s door still open to people who want to approach HR person to person”. He adds: “It’s also worth remembering that what works for a large company might not work for a small company, and what works for a law firm might not work for an accountancy firm.”
Talk about risks
It can’t be overlooked that there could be disadvantages to voice technology for both HR teams and employees.
Evans says: “Anything that is perceived as throwing up a barrier to HR could have negative impacts on employees. Whatever you use the technology for, it needs to be of benefit to employees and thought about up front. Otherwise, the results could do some damage to an organisation.
“HR should ensure that any data gathered by the technology is anonymous and that consent mechanisms are in place so that people can easily say no to it. And ensure that data is secure – but that is a risk with any information you hold about your employees.”
Perhaps the biggest risk with voice technology is over-estimating what it is capable of. Adams explains: “We’re taking great strides in developing a conversational interface with machines, but it is composed of a lot of parts and I don’t see it being able to replicate everything a human can do in the foreseeable future.
“In the area of speech recognition, for example, there’s quite a way to go to match a human. In ideal circumstances it might feel like it is, but as soon as there’s a lot of noise, or an accent, or what someone is saying is a little bit nonsensical, the technology isn’t there yet.
“It simply can’t yet carry on a long, meaningful conversation and mimic humans. Ten years ago I would have said it would be 10 years before it could, but I’m still saying that now. That achievement feels like it’s a bit elusive at the moment.”
Voice technologies may still have shortcomings, but someone who is convinced they have a big future is Rishi Kudale, spokesperson for Bengaluru-based Indian start-up Reverie Technologies, which has created the Gopal voice assistant. In a country of 1.3 billion people, their product is aimed at accommodating those Indians who speak one of the local languages rather than English.
He says: “There’s no doubt in my mind that voice technology has a lot of potential. In an HR context, it can help ease the load on HR resources by powering automated channels like helplines that can receive and answer queries. In addition, voice-based systems are a great platform for truly anonymous feedback collection, with the technology being able to record voice, transcribe it to text and store it securely.”
Andrew Spence, HR Transformation Director for Glass Bead Consulting, also believes it can help HR professionals: “I agree that, in theory, using this kind of automation could get rid of some administration work and add some value to HR’s efforts by pushing more resources into people performance.” But he adds: “My experience is that HR needs new people to help achieve that, such as data scientists, technologists, behavioural scientists – and they are beginning to enter HR.”
One voice technology experiment that intrigues him is at Boston-based Humanyze, where employees wear ID badges around their necks that record location, voice intonation during conversation and who they’re speaking to, but not the content of their conversation. “The idea is that the data collected can be used to understand how people use office space and how teams are integrating after things like company mergers. Pattern-recognising AI can then work out the best layouts for office space.
“It’s controversial, but it does rely on employees opting in – and while the technology certainly analyses speech, they are anonymised. It just produces results on a team level; for example, team A had better social interaction in the canteen when they sat at bigger tables.”
Whatever the possible applications for voice technology as an aid to HR functions, Spence shares the view of many when he concludes: “The best way to ensure that the technology is going to work is that both an organisation and its employees share the benefits.”