Stop Overengineering People Management

Harvard Business Review, 2020-09

Basic Revelations

  • While employee empowerment is not anything new anymore, the pendulum swings the other way towards command-and-control management principles.

  • This is done under the pretence of “optimization”, i.e. centralizing decision making and control under experts and algorithms.

  • Labour is considered a commodity, and the expensive human judgement is shifted to algorithms, with humans having to comply.

  • At the same time, there’s no body of evidence proving that taking away from empowered employees is consequence-free.

  • This engineering approach to management in the employer’s market can only be countered by employee walkout.


  • The “liquid” workforce. It’s quite hard [or at least – not cheap] to downsize during recessions and quite expensive and long to hire the people back. So the gig economy with powerless contractors instead of benefitted employees has started looking really attractive.

  • “Talent on demand” is a lucrative proposition, as contractors can be managed more brutally with KPIs like 999/1000 on-time deliveries.

  • [MK: my personal take is that gig work is a hidden form of unemployment.]

  • The practice of firing people fast at the first signs of trouble is not linked to positive financial performance and business outcomes for firms. The companies postponing layoffs do better long-term.

  • Using gig workers alongside employees has negative effect on permanent staff, weakening loyalty and relationships with peers, and lowering operational performance.

  • Gig workers are disincentivized to do anything above and beyond the scope of their engagement.

  • In non-COVID-related downsizings contractors don’t leave ship first: both employees and contractors suffer more or less equally.

  • Uneven starting salaries. For some reason it’s the norm to offer employees different starting salaries and have them negotiate them up. Are the savings really worth the employee’s discomfort?

  • AI and optimization. When people are told what to do, they hold off operations improvement suggestions (and often the feedback pathways are closed anyway). Managers don’t get to manage staff (the source of authority) and are equally unhappy. Monitoring white collar workers’ screens and the time they spend in the office (i.e. operational monitoring) is incorrectly considered to be the predictor of business outcomes.

  • Even work from home is not immune from nosy monitoring. Even if the monitoring is made for noble reason (social monitoring), the same data can backfire on employees after the pandemic is over.

  • Monitoring is often lauded as a means to make better offices, but we’ve already seen that it’s easy to kill collaboration in the office under this pretence.


  • It may be a good idea to still focus on optimizations, but engage front-line employees (and not take away the empowerment from managers and staff).

  • When it comes to timing and planning, unfortunately, AI is not as capable of efficiently planning schedules with employee needs in “mind” as teams can do.

  • For managers it’s important to understand that there’s no single way of performing tasks better, so AI is not the ultimate solution. Asking employees may not be as efficient, but the benefit of the human’s ability to make conclusions out of unstructured information is worth the try.

© Peter Capelli, Wharton Business School