Microsoft Analyzed Data on Its Newly Remote Workforce

https://hbr.org/2020/07/microsoft-analyzed-data-on-its-newly-remote-workforce?ab=seriesnav-bigidea

After a year of Zoom meetings, people are yearning for in-person connection at work. Interactions once considered foreign (like muting yourself after you finished speaking or turning off your camera to subtly suggest that you have left the room) now seem normal.

Microsoft did research early into the pandemic to explore how their employees have adjusted to the virtual office. Their analytics team observed employee activity on Microsoft 365 (which probably includes emails, calendar, instant message, video conferences, documents) and collected additional information through sentiment surveys. The team used their own product, Microsoft Workplace Analytics, to analyze the data (and the article felt a little like an advertisement for this product). Though some employees may find it concerning how much of their data Microsoft can collect and analyze, the activities described in the article are an excellent use of workforce data.

First, the company had a good reason for observing workforce patterns. It is important to understand how employees are faring with a completely new way of working and using that information to create change. In other words, Microsoft did not analyze data for the sake of analyzing data. Second, this research led to meaningful insights. The team learned that workdays had lengthened and people were increasingly on their computers during lunch and dinner time. From this, Microsoft inferred that the work-life balance lines had become blurred during the pandemic, and they would need to figure out how to manage this to ensure their employees’ lives were sustainable. They also learned that managers’ work time had increased above average. Finally, Microsoft observed an increase in 1:1s across the company and virtual social meetings, suggesting that their employees craved social connection.

Of course, employees may still be uncomfortable with knowing that their behaviors are being observed. Even though Microsoft disaggregated and de-identified the data, it still feels as if the company has the ability to check on the data of any individual employee. But the purpose of using this workforce data is to help employees cope with this new way of working, which makes this research worth sacrificing a little less privacy. Setting privacy concerns aside, employees should not be doing anything at work that they would not want their employer to see.

So what should Microsoft’s next steps be, given their observations? The company’s first priority is to make sure the managers’ work life is sustainable. The data suggested that managers were key to the successful transition to virtual, and yet they seemed to be the ones who were affected most. The answer is unclear what they should do. One small solution is to have a workshop led by successful managers who share with their peers their tips on managing during a pandemic.

Secondly, Microsoft should set a guideline for managers to increase their 1:1s with their direct reports. This seems to go against the previous recommendation; it seems silly to ask managers to cut back on work, but still increase the time spent with each employee. But Microsoft found a negative correlation between the number of hours worked and the time spent in manager 1:1s (i.e. the more time you spend in 1:1s with your manager, the less hours you work). The theory is that the 1:1s help align managers and employees, saving the employees from spinning their wheels or prioritizing the wrong projects. If this is true, managers might actually be able to save time for themselves in the long run by spending time upfront to make sure their employees are on the same page as them.

Finally, Microsoft should set guidelines that encourage certain times be treated as sacred (e.g., lunchtime, dinnertime, weekends). A more radical approach would be to give managers a heads up if their direct reports are working during lunchtime or weekends. This could prompt the manager to talk to the employee and find out whether the employee has too much work on their plate or if this is just how the employee personally likes to work. The drawback here is that people might be uncomfortable with this level of monitoring.

Previous:

Ford: Brick & Mortar Icon Transforms Workforce Home Office Capabilities

Next:

Why police using facial recognition technology is wrong

Student comments on Microsoft Analyzed Data on Its Newly Remote Workforce

  1. Great post, Carla! Microsoft’s example perfectly illustrates how data monitoring by employers is a complex issue and there is both pros and cons. In that case I see more pros, because the intent is 1. to learn how employees’ working habits adapt in this completely unprecedented situation and 2. ultimately improve their well being (which might lead to greater productivity and retention).
    I personally can more easily draw a line when the data we are looking at becomes identifiable/personalized. It feels morally debatable to send managers information about the working habits of their reporters, even if the intention is positive. Moreover, it can be irrelevant (What if some employees prefer to work late because of personal obligations) or deteriorate trust (thus producing the opposite effect on productivity/retention). But of course, we can still consider testing this and quantify the impact.

  2. This is a really interesting post, Carla. The blurring to nothing of boundaries between work and non-work during the pandemic is something I’ve been hearing a lot about anecdotally, so it’s really interesting to read a more analytical take. I like your suggestion of treating certain times as sacred, and I think your concern about an enforcement mechanism is well founded. I think this points to a more general question of how to best deliver data-driven behavioral nudges. When they’re delivered through technology directly, they can be seen as annoying. However, when delivered by a manager or otherwise filtered through a human intermediary, they risk being seen as intrusive. I wonder which delivery method is actually more effective and engenders the higher level of both employee well-being and organizational effectiveness.

Leave a comment