Artificial intelligence could improve our ability to manage people when they work from anywhere – but there are realistic barriers to achieving this.
Managers have been able to allow their employees to work remotely for some time, yet it took a pandemic to accept that working from home, or working from anywhere (WFA) was not just a useful resource, but potentially a productivity booster.
Today we are all spending more time on Zoom than we would want, and probably missing some analogue interactions with colleagues. It has also been a taxing year for most managers as they had to reinvent their approach to organizing, monitoring, supervising, and rewarding people’s output without so much consideration of their input. Again, this should have happened before, as in any area of life there has long been a gap between style and substance and the essence of meritocracy and value is to focus less on the former and more on the latter. Ask not what people look like, how they appear, and if they are in the right place at the right time, but what they actually deliver and contribute to their groups, teams, and organizations.
When Susan Cain wrote her bestselling book Quiet it appeared counterintuitive to some that introverts are often high-performing people, and that it is OK to do your job and deliver if you are not spending all the time self-promoting or blowing your own trumpet. Apparently there are some who do and others who pretend to do. As a client noted when he was sent to work from home during the early stages of the pandemic: “but without the office, how will I pretend to work?”.
Now that we have experienced several phases of crisis and endured many months of partial and total lockdown, turning our offices into empty spaces and homes into our offices (yes, we used to work from home and we are living at work), the question is whether technology and humans have been able to upgrade their partnership to provide better ways of managing talent. With all these months of practice, are managers changing their fundamental approach to understanding and motivating people, and has tech been of any help beyond WiFi, e-mail, Zoom and related (not precisely novel) apps? Although progress appears rather limited to date, we can envisage a not so distant future in which technology – more precisely, our ability to extract useful human insights out of the data – can support human managers who are interested in improving their effectiveness while people work remotely. For example, consider these simple 3 ideas:
(1) Applying Natural Language Processing to real-time decoding of people’s mood, motivation, and emotions. NLP is well-established and has been used in call centers and automated customer services bots for some time. When you sound angry on a customer service call, your emotions will be decoded (not just the words you use but certain physical attributes of your voice) that will enable the bot to fast-track you or direct you to a kinder human. You can imagine this same methodology helping managers understand that some of their team members need empathy, kindness, or attention, because they detect patterns in their voice that suggest they are anxious, stressed, or needy.
(2) Leveling participation and boosting fairness. We have all experienced the dreadful in-person meetings where some people (usually narcissistic men) dominate the conversation, not least because their desire to show off trumps any intention to leverage productive team behavior, or because they feel much more interesting than they actually are. Clearly, things like mansplaining are a lot harder on Zoom or Teams, but we could stretch the tech application of tools slightly more in order to track how much people speak during meetings, provide live charts to the manager and team members on whether people spoke too much or too little, and even real-time AI feedback on how people’s comments is received or perceived by others (if we combine 2 with elements of 1).
(3) Giving feedback to employees. Although one of the most important things managers should do is to provide employees with feedback (on their potential, performance, behaviors, etc.), most managers are not very good at this. This is where tech could help: imagine a chatbot that gives employees individual and personal real-time feedback on what they do, during meetings or outside of them, by translating their digital records (what they say, write, do, their meta-data and social networking patterns, and each of the dimensions that makes their output different from others). Employees are especially deprived of feedback when their managers are not with them, and even if managers are good at evaluating what employees do, they have limited data on their input, and tend to see mostly the final produce rather than the in-between actions that should be the target of developmental or effective feedback. It is preferable to tell people “stop doing X” or “start doing Y” than “you should have done Z”.
Of course, any technology should be deployed in an ethical, fair, and legal way. This would involve asking employees to opt in, ensuring their is informed consent, and protecting their data, as well as ensuring there is a benefit for them (and not punishing them if they don’t opt in). This would also apply to managers.
And yet, it is unlikely that any of these or similar tools become mainstream any time soon. The reason is not that the technology isn’t helpful, or even concerns about fairness perceptions or reactions – if people are not happy with human managers as they are, they would and should certainly welcome tools that make them and their managers better – but the fact that managers may not be willing to be helped. There are some very well-known reasons for this.
The first is that managers tend to overrate their own potential and performance, particularly when they are not very good. This means that precisely those managers (bosses) who would most benefit from using technology are actually less likely to use it, and vice-versa. In the realm of health analytics and fitness apps, this is equivalent to health tracking and feedback tools being used only by those who are already healthier and more conscientious to begin with – that is, used and enjoyed mostly by those who need it the least.
The second is that organizational cultures need to be mature and ready to embrace these tech and data-driven tools. Most are quite far from that stage, though some of the tools and processes they have in place are obviously more basic or rudimentary versions of these tech innovations. So, for instance, when you go from bi-annual to yearly to monthly or pulse-like engagement surveys, there is only a small step between that and having passive data scraping or real-time sentiment analysis to monitor people’s present emotional states, engagement, or motivation. And if you have in place NLP tools to detect corrupt, unethical, or inappropriate behaviors, then a “bright side” version of that could help managers and people understand when they behave like a high potential employee or do something that is valuable to the company (gamifying real-time feedback like when the Uber app congratulates drivers for unlocking a high surge trip or picking up people in profitable locations, etc.).
The third is that none of these tools will likely suffice to make bad managers good, or substantially boost the performance of managers. In fact, it is still the human and humane acts that cannot be replicated or automated – empathy, attention, consideration, and inspiration – that are likely to account for competent managerial performance, even if tech and data can provide some of the raw ingredients, at least at the diagnostic level, to direct and focus such human behaviors in the most strategic way. Alexa or Siri can tell you if someone is sad, but their ability to cheer them up will always be marginal compared to what the right human can do with the right choice of words, acts or emotions. In short, it is still people enhanced by technology, and technology enhanced by people, that can have the best effect on people (more so than one without the other).