Work from home is going to change how you get promoted, and why

Now that work from home has become an established facet of the corporate landscape, all the initial hassles — glitchy Zoom meetings, how to squeeze a desk into the bedroom — are being replaced by deeper, more consequential anxieties. If I don't go back to the office, employees are wondering, how can I show the boss how hard I'm working? Will I lose my chance to get ahead? Managers, meanwhile, are struggling to figure out how to monitor and evaluate employees they seldom see in person. Is there a way I can tell if my direct reports are abusing the freedom that comes with remote work? How can I separate the go-getters from the nap-takers?

As a result, the rise of remote work is forcing companies to overhaul the way they measure performance. And the good news is: Research shows that the new way of managing is a whole lot better than the old one. Instead of evaluating employees based on meaningless considerations — how late they stay at the office, say, or whether they're fun to hang out with around the water cooler — companies are starting to focus on the thing they can best measure from afar: the work employees actually do. That, in turn, could lead to fairer raises and promotions for employees, and bigger profits for companies. Thanks to work from home, we could be entering an era of better management. 

"This is an amazing opportunity for companies to rethink the traditional model of evaluating performance," Emilio Castilla, a professor at the MIT Sloan School of Management and a leading expert on performance management, told me. "Measuring outcomes is a much more efficient way of measuring performance that happens to be aligned with the modern professional worker, who wants more freedom and knows what makes them more productive." 

You can think of the way managers evaluate their employees in two broad categories: inputs and outputs. Inputs are all the things that people do to get the work done, like putting in long hours or being collaborative — the how of work. Outputs are the results they produce — the what of work: the big sale, the finished brief, the new product that just shipped. 

Ever since World War I — when the US military devised a "merit rating" system to grade soldiers — managers have primarily looked at inputs. In the age of offices, there were many benefits to this approach. First, it made tracking performance fast and easy. Managers kept track of how early employees showed up and how late they stayed, and relied on in-person interactions to gauge who was smart and who was collaborative and who was mediocre. It gave bosses a lot of discretion: rather than defining what constituted good performance, they relied on their gut. 

But over the decades, the flaws to this approach became evident. Inputs encouraged a certain degree of faking it. As we all know, appearing to work — e.g. typing furiously whenever the boss looks your way, or staying late to get a gold star — is very different from actually doing good work. And research, on the whole, didn't support the assumption that certain traits — like a passion for one's job — actually led to better performance. 

Giving managers so much discretion in performance evaluations also created a good ol' boy system. "Bosses tend to like the person who, when you call them, they say yes, and they do what you ask," Jennifer Glass, a professor of sociology at the University of Texas at Austin who studies gender stratification in the workforce, told me. "That's the kind of biased evaluation that generally gives white men an edge. Why? Because the boss is more likely to be a white guy. So you've got this homogamy: they look like you, they act like you, they come watch football with you." The old system was grading for similarity, not merit.

Well before the pandemic hit, some HR departments had started searching for better ways to measure performance, based on what employees actually produce. That was easy enough to do in lower-paid, repetitive jobs, where success is easy to put a number on and track. Call center work, for example, is probably the most quantified job on the planet: Agents are held to a litany of statistics, like average call time and customer satisfaction ratings, that are extensively measured by software. That's why call center work was one of the few jobs that was widely performed from home before the pandemic. Managers could use precise metrics to track the productivity of their teams, even if they couldn't see them.

But in better-paid, professional jobs that involve high levels of complexity and creativity, it's tougher to reduce productivity to hard, cold numbers. Some employers tried — evaluating software programmers, say, based on the number of lines of code they wrote. But narrow, results-based evaluations had a funny way of warping incentives: Engineers were rewarded for writing more code than was necessary, a recipe for crappy software. Other companies, in a recent fad, have replaced annual performance reviews with "continuous" check-ins — informal, conversational feedback that takes place on a more regular basis. 

The problem with this approach, MIT's Castilla told me, is that organizations still need to decide who gets a raise, who gets a promotion, and who needs to go. Frequent feedback is great, but without some kind of rigorous system to evaluate performance, critical personnel decisions still end up getting made in a vacuum — with little transparency and accountability. "By getting rid of performance evaluations, by completely leaving it up to the manager to decide," Castilla said, "you give a license for discretion and for bias." 

Not the what, but the so what

So how can companies evaluate professionals in a fair and effective way? This is the core of Castilla's research. For years, he has been working with companies to help them design better systems of performance management. And the things that work in an office environment become even more important in hybrid and remote workplaces. 

The solutions, he cautioned, depend on an individual company's goals and culture. "It's not so much that there's one best practice, where if I tell you to do A, B, and C, it's going to work," he explained. But he does have a few rules of thumb. And the biggest one is to establish a clear definition of a high performer.

That's more important — and more complicated — than it sounds. In one study at a Bay Area tech company, Castilla and his co-author Aruna Ranganathan found that managers had wildly different notions of what factors should be considered in an evaluation — notions shaped by their own experiences of being evaluated in their previous jobs. White male managers tended to incorporate intangible and qualitative attributes, such as an employee's likability, because they had benefited from such an approach themselves. Castilla and Ranganathan call this the diffuse approach. Women and managers of color, on the other hand, stuck to narrower criteria of measurable work results — what the researchers call the focused approach — because they had been dinged in the past for lacking certain attributes, like assertiveness, that they felt were irrelevant to performance. 

"It's only when employees, managers, and executives all have a common understanding of what's a great performer," Castilla told me, "that we can start to hope that we're using the right metric to measure it." 

When companies are establishing the criteria for high performance, Castilla added, they should make sure they're prioritizing the results that line up with the organization's broader objectives. One way companies are trying to do that is by shifting from measuring output (the number of lines of code your engineers write) to measuring outcomes (how useful your code is to your customers). That can help ensure employees are working on the right projects — the ones that will make the most difference to the bottom line — instead of wasting time on peripheral things that won't move the needle. It's a shift that's starting to catch on in many workplaces: In a report released by Deloitte this year, more than 65% of executives agreed that performance evaluations should begin focusing more on outcomes and less on output. 

You can't see how many hours someone is sitting at their desk in this environment, but you can see how much impact they're having. Eddie Kim, chief technology officer of Gusto

One company that got a head start on the process is Gusto, a payroll and benefits software provider based in San Francisco. For years, its performance reviews consisted of free-form essays that listed everything an employee did over the course of six months. But as the company expanded its new offices and began allowing for more remote work, it realized that it needed to bring more structure to the process. In 2019, it devised a template to guide its engineering, product, and design teams to focus less on their effort and output and more on the impact of their work. Employees and managers honed in on the big questions: Did Sam's code help customers save time running payroll? Did the new product Alex shepherded increase Gusto's gross margins? The goal, according to Eddie Kim, the company's co-founder and chief technology officer, was to make performance evaluations "not just about the what, but about the so what." 

This year, Gusto also started tracking the "extracurricular" ways that employees make contributions to the company. Now, evaluations reward employees who help others get better at their jobs — by mentoring young coders, say, or by providing detailed documentation on a software system so that others can understand it too. The review process recognizes that employees benefit the company in all sorts of ways. "People are different," Kim said. "Some like to spend most of their time in Project and Team Impact. Some like to spend more time in Org Contributions. The idea is to lean into a diverse team's strengths, because all of it adds value to our customers." 

Gusto's new thinking proved vital last year, when the pandemic suddenly accelerated Gusto's transition to remote work. "You can't see how many hours someone is sitting at their desk in this environment, but you can see how much impact they're having," Kim told me. "It's allowed us to focus on what really matters." 

Another company that has overhauled its performance reviews is mmhmm, a startup that provides video presentation tools for virtual meetings. During the pandemic, when the company got rid of its offices and went fully remote, it not only changed its compensation system, as I reported in August, it also overhauled its approach to evaluating employees. "Before, we were kind of flippant about it," CEO Phil Libin told me. "We were relying too much on what we thought we knew about the in-person stuff. Whereas now, we can't. So we really do have to be more about: OK, what are the objectives? That made it more actionable." 

In hindsight, Libin says he was guilty of the kind of built-in biases that came with managing the old way. In his former job, as CEO of the note-taking app Evernote, he often made decisions by spontaneously taking employees out for coffee on long walks. "The people who got to be important in the company were the people I liked going for walks to get coffee with," he said. "The people who were getting rewarded were the ones that were good at being seen in person."

Lazy management

The transition to outcome-based evaluations isn't going to solve everything, of course. One big concern is whether the widespread option of hybrid work — part remote, part in-person — will create two-tier workplaces that favor the employees who come into the office more. Researchers call this phenomenon proximity bias — the sociological equivalent of "out of sight, out of mind." Study after study before the pandemic demonstrated that face time matters when it comes to promotions — and it's even more worrying now, since surveys show that women with young children prefer to work from home more than their male counterparts. Evaluating employees based on what they actually produce should help — good work is good work, regardless of where it takes place. But if managers end up leaning on the  team members whom they can walk up to in person, office employees will get more chances to prove themselves than their remote colleagues. In July, when the Society for Human Resource Management surveyed supervisors, 42% admitted that they sometimes forget about remote workers when assigning tasks. 

And even if a company wants to start evaluating their employees based on outcomes rather than outputs, the transition will be tricky. Outputs are immediate — you can tell right away how many lines of code a programmer wrote each day. Outcomes are slower — it takes a long time for an engineer's work to make a difference for a customer, and even longer for that difference to show up on the company's balance sheet. Plus, it's not like there's some cold, hard algorithm for measuring impact; it requires making inferences, which are inherently subjective. "There will never be a point where it'll be completely objective," Kim told me. "I would argue that if you make it completely objective, you can incentivize the wrong things. People start aiming for exactly what the objective scorecard is. That's not how you deliver value to the customer." 

I see that dilemma at my own employer, which is probably more quantitative and outcomes-oriented than most newsrooms. My boss sits down with me every quarter, over Google Meet, to come up with three or four concrete goals I'll be expected to achieve over the next three months. The primary goal is both simple and quantifiable: how many new subscribers my stories bring in. That places the emphasis on an outcome that's important to Insider — growing its subscriber base — instead of how many words I write or how quickly I respond to my editor's Slack messages. 

But there are downsides to the metric. At Insider, readers hit our paywall as soon as they click on my stories — which means that the only information they have before they decide whether to pull out their credit card is the story's headline and illustration. That might incentivize me to pump out sensational headline after headline, without much care for the reporting or writing that follows it. That, in turn, would be demoralizing for me — and counterproductive for the company. Subscribers won't keep subscribing if all they get access to is junk. 

Having come from a more traditional news outlet, I was initially worried about Insider's obsession with quantifying performance. But 10 months into my job, I'm starting to come around. I still don't think the metric of new subscriptions is perfect, but I sense that management is constantly looking for ways to get a better read on how subscribers feel about my work — from how long they spend reading my stories, to how widely they share them. I'm glad I work at a company that's at least trying to get it right. 

Because that, ultimately, is what will enable me to keep the extraordinary amount of autonomy that working from home affords me. If companies can't figure out how to track and reward the performance of remote workers, they'll try to force everyone to return to the office — which is exactly what many employers are now doing. And that's a real shame. By seizing on this mass experiment in remote work, employers have an opportunity to set aside old forms of management, replacing squishy judgments about attitude and effort with meaningful measures of success.  

"People advancing based on how good they are at seeming productive at the office — that's always been an act," said Libin, the CEO of mmhmm. "A lot of the way we were doing it sucked. It was lazy management." 

Reed Alexander contributed reporting.