workplace1

image via wsj.com

In the battle against “quiet quitting” and other obstacles to productivity in the workplace, companies are increasingly turning to an array of sophisticated tools to watch and analyze how employees do their jobs. The sobering news for America’s bosses: These technologies can fall short of their promises, and even be counterproductive.

Patchy evidence for the effectiveness of workplace monitoring tech hasn’t stopped it from sweeping through U.S. companies over the past 2½ years. Since the start of the pandemic, one in three medium-to-large U.S. companies has adopted some kind of worker surveillance system, and the total fraction using such systems is now two in three, says Brian Kropp, vice president of HR research at Gartner. While there is a broad spectrum of how these systems work and what data they gather, many of them include constant monitoring of nearly everything workers do on their devices.

That is a pace of adoption that is rare in the history of technology—even at the steepest part of the curve of its embrace by Americans, not even the smartphone spread as quickly. This technological shift is particularly jarring for white collar workers who have tended to have greater leeway in their work practices than blue-collar workers who have to punch time clocks.

In changing the very nature of work, how it is perceived by those doing it, and what companies can expect of workers, this shift has the potential to represent a profound unbalancing of the power between employee and employer, say those who study it and even some within the industry who create it.

Intra-office surveillance capitalism

By law, what data employers can gather on employees is extraordinarily broad. There is software that can take a screenshot of a worker’s computer every 10 minutes, while also recording what apps and websites that worker visited, and how long she stayed. Critics of this kind of monitoring have nicknamed these systems “bossware” and deride them as a new form of morale-eroding micromanagement, like having your manager look over your shoulder at every moment of the day.

Gathering this much data about what a worker does all day generally requires what are called “agent-based” systems, in which a piece of software is installed on a company-issued device. Such systems can have complete access to everything that happens on a computer, tablet or phone. (This is, not incidentally, how the most sophisticated forms of spyware used by nation-states capture data from the devices of targets.) Examples of this kind of software include ActivTrak and Teramind.

“Realistically, the vast majority of customers don’t find the need to enable full monitoring on all users all the time,” says Isaac Kohen, vice president of research and development at Teramind. The company’s software is designed to automatically record or flag user behavior if an employee does something untoward, such as emailing company secrets to a competitor, or spending all day on Facebook. But it is entirely up to employers how to use the system, and what rules to establish. “Unfortunately, due to the nature of Teramind, the system can be abused if placed in the wrong hands,” he adds.

ActivTrak says on its website that it has “evolved from an employee monitoring tool to a privacy-conscious, powerful productivity platform.” This means its software now includes privacy controls and anonymized, aggregated data, says a company spokeswoman. Unlike similar tools, ActivTrak does not support keystroke logging, camera access, video recording or email reading or counting, she adds.

At the other end of the spectrum are systems that are more familiar to office workers, such as Google Workspace and Microsoft 365. These gather data, but deliberately limit how much and what kind. Both allow a senior administrator in an organization with appropriate privileges to see which applications within these systems an employee uses, and how often, but may obscure the employee’s identity and offer data only over a period of weeks, not days or hours. For some services, more detailed data, like how many messages or emails are sent and when, can be extracted. Notably, however, these systems aren’t capable of the all-encompassing tracking that agent-based ones do. They can’t track every keystroke, or take screenshots of a worker’s device, for example.

Microsoft has taken pains to make employee-level data and activity within its apps difficult even for IT administrators to access, says a company spokesman. Microsoft does not believe that activity is the same thing as productivity, and organizations should be careful about tracking the wrong kinds of things in order to assess employee performance, he adds.

Companies like Prodoscore take the data common business applications gather and parse it in ways that the makers of those applications generally don’t. David Powell, president of Prodoscore, says the line that marks where “creepy” employee monitoring begins and ends is this: All of the data Prodoscore uses is already gathered by the cloud-based software suites companies use, and is available through APIs, which are connection points for software.

Even with this limited data, Prodoscore can create for bosses an incredibly detailed breakdown of precisely what Prodoscore-connected services an employee is using, when, and whether from the office or outside it. The company’s software can even chart that worker’s productivity across time, digesting it all into a single score, which Mr. Powell compares to a credit report, but for work. While the service is intended to help bosses recognize exceptional effort in some employees, and to coach others whose productivity may be falling below their historical average, this level of data could be misused, if a manager were to be overly reliant on it.

“We think of it like a gun,” says Mr. Powell. “It can be used for good or bad.”

Fierce debate about usefulness

For a relatively complicated and potentially intrusive technology to be so heartily embraced by so many company leaders, you would think there would be overwhelming evidence that it delivers on its promises. But there is no independent, peer-reviewed research showing that it does, say Antonio Aloisi and Valerio De Stefano, professors who recently combed the literature on this subject for their book, “Your Boss Is an Algorithm,” about the growing phenomenon of management by software.

“There is definitely no study pointing out that this increases productivity in any meaningful way,” says Dr. Aloisi, who is based in Madrid.

Ben Waber is president of Humanyze, a company that uses data to assess organizational health over long periods of time, without allowing managers to track individual employees. He says that systems focused on quantifying things like when an individual employee is sitting at their desk, and how many emails they send every day, should be viewed with extreme skepticism. “There is no research showing that those kinds of things have any correlation with any outcome that employers care about,” adds Mr. Waber.

Adrian Reece, principal data scientist at Prodoscore, says that the academic literature on productivity monitoring tools is divided into two camps. One finds that such tools increase accountability at work and can help people set and track their progress toward goals. The other camp argues these tools are ethically dubious, inequitable, and lead to a feeling of low personal agency, which can decrease productivity.

Systems that simply track whether workers are logged in and at their desks, and ding them for not being connected, can cause workers increased stress, according to a number of studies, says Dr. De Stefano.

For example, one 2020 survey of 2,100 call center workers across seven companies found that more intensive monitoring of workers—including recording keystrokes and tracking online activity—was associated with increased stress, lower job satisfaction, higher absenteeism, and more desire to quit.

In a 2020 survey of more than 2,000 of its members, the UK’s Trade Union Council found that 56% said the introduction of new worker monitoring systems had damaged trust between workers and employers.

There is also evidence, adds Dr. De Stefano, that introducing these systems can encourage users to game the systems themselves, rather than doing their actual jobs.

The price of working from home

In some ways, what’s going on here is that companies are conducting a gigantic research experiment on their employees without necessarily being equipped to understand the data their worker-surveillance systems produce. Only about one in three medium-to-large companies has an analytics and data science team capable of parsing the kind of data these systems spit out, says Mr. Kropp of Gartner.

However employees feel about increased monitoring of how they do their work, they may not have much choice about it, as more companies make working from home contingent on employee acceptance of monitoring. One Prodoscore client that recently shifted to remote work specified that employees who wanted to work from home had to use Prodoscore, says Mr. Powell. In the first month, 80% of the company’s employees, or 3,200 of them, opted in, he adds.

Source: WSJ.com September 17, 2022 | By Christopher Mims