Why do large companies keep abandoning remote work?

Reading Time: 9 minutes

“There’s nothing bad about remote work…to be googled.”

— CTO of a growing tech company

That sounds great on the surface, but it’s an unfortunate realization to reach when you’re trying to weigh the pros and cons of remote work for your company. Because even though the first thing you find upon Googling is a list called The Most Common (Bad) Arguments Against Remote Work, you know that there have to be drawbacks in there somewhere.

Faces and screens link to one another with dotted lines, forming a network on top of a map of the world.

You know this for two reasons.

  1. There’s a documented history of large companies recalling their remote teams. These companies tried to embrace remote work, but they called off the experiment. If it didn’t work for them, how do you know it won’t fail you?
  2. You have noticed, when people work remotely, that you have some frustrations with the workflow that you did not experience prior to the remote addition. Are those frustrations worth continuing the remote experiment? If you expand your remote policy, will they get worse?

We’ll talk about both of these reasons, but there’s a lot to unpack.

So we’re going to start with that troubling history of recalls (this post).

Then we’re going to talk about that those troubling issues that you yourself have noticed with remote work (future post).

Why do large companies keep abandoning remote work?

In 2013, Yahoo recalled their remote employees. In 2014, Reddit recalled their remote employees. In 2015, Bank of America recalled some remote employees. In 2016, Aetna recalled some remote employees. In 2017, IBM recalled their remote employees. Why? What wasn’t working? And how do we know it will work for us if it didn’t work for them?

When we dig into these historical cases a little further, we start to notice some common threads.

Let’s start with Yahoo. When CEO Marissa Mayer discussed the reason for the recall decision, she explained that “People are more productive when they’re alone…but they’re more collaborative and innovative when they’re together. Some of the best ideas come from pulling two different ideas together.” She exemplified the Yahoo Weather app, which crawls people’s flickr pictures and cross-references their location to ‘get a more accurate picture of weather in your area.’ The app, evidently, was conceived by two developers who worked in the same office (neither of whom, one must conclude, have any training in data privacy. But that’s a separate thing). That defense, it turns out, pops up in several of these cases.

Take IBM. After 20 quarters of straight losses, the company decided that they needed to create more opportunities for serendipitous ideas. They would bring their employees into the same building to increase the probability of random collisions that would result in new business ideas that could turn the company around. (It’s worth noting on IBM specifically that, due to the long-running losses before the recall, employees speculated that the order to commute to an office or leave was an attempt by the company to conduct a mass layoff without the press coverage and severance expense, rather than a strategic decision).

Aetna, similarly, made the change because it “believes its generous work-at-home policy is harming collaboration.” Spokesman Matt Clyburn assigned the recall “the goal of increasing collaboration and driving innovation.”

Reddit closed their New York and Salt Lake City offices in 2014, and they insisted that employees who wished to stay move to the San Francisco flagship office. The stated reason was that “There were too many times when we just needed to be able to walk over and tap someone on the shoulder and discuss a complex issue in-depth, right away.”

Bank of America did not disclose much about their reasons. Spokesman Mark Pipitone said only that “the decision is part of an ongoing evaluation of the bank’s business needs as it tries to best serve customers and clients.” So I won’t speculate on what their needs were. It’s also notable that Bank of America’s decision only affected a few of their distributed offices, so presumably they used some kind of metric to determine where to make the change and where to leave it alone.

That doesn’t change the fact that, in three of these five cases, the stated reason for the recall chiefly dealt with having employees run into one another in person more often. In a fourth case, the reason chiefly dealt with some employees wanting to specifically seek out other employees in person more often. Maybe these companies simply could not find a distributed solution for creating the collaboration and innovation they needed.

Or maybe that’s not quite what happened.

To determine what might have happened, we need to talk about some remote work success stories: Basecamp (formerly 37Signals), Trello, and the U.S. Military.

Basecamp writes team collaboration software, but that’s not the most interesting thing about the company. What’s interesting is that Basecamp has very little trouble filling their engineering positions, including the senior ones. If you’ve ever been a hiring manager at a tech company, you understand that this is a big deal.

So what is it about Basecamp? Among software engineers they’re famous for a few things: policies that explicitly encourage employees to have a life outside of work, salaries that match San Francisco rates regardless of employee location, and letting employees work from wherever they want. It’s the third of these that interests us today. Basecamp has made this remote workforce a mainstay of their business: in fact, 70% of their employees work remotely. Remote work has not stopped the company from releasing Campfire chat, Highrise, and Know Your Company. It has not stopped them from switching from a services company to a product company and changing their name and branding. It has not stopped them from regularly making the headlines of tech news for innovative policies (like the aforementioned lifestyle benefits and payment policies, among other things), nor has it stopped them from making the Forbes Small Giants list.

Trello writes team project management software, and about 60% of their employees take advantage of the remote work policy. But this team does not seam to suffer from lack of creative collaboration either: their product possesses the flexibility to augment the work of software teamsawards committeesbloggersbook authors, and more. Trello also has a helpful, well-written blog and a transparent feature request process. Shouldn’t coordinating these kinds of projects require the team members bumping together in the hallway?

A number of other companies have embraced remote policies and seen success with them: Evernote, Litmus, Zapier, Buffer, inVision, Automattic, and StackOverflow, among others.

So how come the Basecamps and Trellos of the world don’t seem to be suffering from a dearth of collaboration and innovation like the companies recalling their remote teams?

You might notice one glaring difference between Basecamp or Trello and Yahoo, Bank of America, Aetna, or IBM. Basecamp has 54 employees and Trello has around 100. Evernote, Litmus, Zapier, Buffer, inVision, Automattic, and StackOverflow all have fewer than 1,000—most fewer than 200. By contrast, the remote recalls come from much larger companies (except for Reddit, which matches the approximate sizes of Litmus, Zapier, Buffer, or StackOverflow). Yahoo had 12,000 employees at the time of the recall, which is about 6 times the roll sheet of StackOverflow. It’s 24 times the roll sheet of Basecamp. And it’s still just a fraction of the 50,000 employees at Aetna, the 208,000 at Bank of America, or the 380,000 at IBM. Sure, it’s easy to make it work when there’s only a few people! Right?


But before we reach that conclusion, let’s look at another successful example: The United States Regional Commands.

The U.S. Regional Commands govern the U.S. Military’s activities in each theater of the globe. Each morning, they conduct a remote meeting of about 8,000 people to establish what resources are available that day for which units to do what. The meeting has proven more effective for them than a meeting of a few representatives, followed by dissemination of information across the chain of command, because there is no game of telephone in the reporting scheme. But 8,000 people is a lot: it’s almost as many remote people as Yahoo has total people on the payroll today. So they’re breaking the size rule that we thought we’d identified in comparing Basecamp to the larger companies that recalled their remote teams.

There’s something else that separates the remote work success stories and the recall cases. In the cases of companies recalling remote employees, remote employees made up a minority of the affected employee group. In the case of Yahoo, the recall affected only 200 of their 12,000 employees. At bank of America, the number of employees affected was in the dozens. At Aetna, 43% of employees reported telecommuting some of the time or all of the time, and a fraction of those were affected by the recall—so, 20% of the company affected? Tops? At IBM the recall affected around 7,600 employees, or about 2% of the total employee base. For Reddit I could not find a number, but based on the size of their San Francisco office at the time of the announcement and the 100% pay bump offered to folks who were moving from cheaper areas, all signs point to it was just a few: maybe 20% of the company, maybe less. Compare that to Basecamp and Trello, where 70% and 60% of the workforce is remote, respectively. This minority-majority thing is important: we’ll come back to it later.

So far, these companies’ plans to turn things around by recalling their remote employees do not seem to have worked. IBM has not turned it around. Neither has Aetna. Yahoo has shed another 3500 employees and gained a reputation for irrelevance since their recall. Reddit missed its revenue goals by more than forty percent on the heels its latest large infusion of capital. Bank of America continues to struggle (it didn’t help that much of a generation divested from the bank upon finding out that they financed the Dakota Access Pipeline).

Look. Bringing the remote teams in-house probably won’t save these companies. There is, however, a logical reason that they tried it: the leaders themselves experienced frustrations with remote work that they did not experience with colocated work. You might have some experience with the same on your team.

If so, let’s talk about where this comes from and what it means for companies like yours.

If you liked this post, you might also like:

Advanced Pair Programming: Pairing Remotely

More best practices for remote pair programming

One comment

  1. Basecamp has three weeks holidays (compared with “no vacation policy” at Netflix) and has very restricted policy for people who want to do some activity outside of work, like attending conferences or contribute to other projects, defending this as it would damage their performance at work. Whilst the latter can be justified as “have life outside work”, three weeks of holidays per year and “generous” permission to take one month leave on own expense every third year, doesn’t sound very much like it.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.