Coding and Coercion

Björn Westergard

Unions have been trying to organize software engineers for decades, with little success. Here's a look at the organizing campaign that might turn things around.

Christiaan Colen / Flickr

Interview by
Ben Tarnoff

Earlier this year, a tech company named Lanetix fired its entire staff of software engineers for trying to unionize with NewsGuild–Communications Workers of America (CWA). While the outcome of the campaign is still unknown — NewsGuild–CWA has filed a complaint with the National Labor Relations Board alleging illegal retaliation — the Lanetix fight is potentially historic.

Unions have been trying to organize software engineers for decades, with little success. But this time might be different — and a victory for the Lanetix workers may inspire similar efforts elsewhere, especially with the help of leftist tech groups like the Tech Workers Coalition, which recently helped stage a well-attended picket outside of the Lanetix offices in San Francisco.

Ben Tarnoff recently spoke to two of the fired Lanetix engineers, Björn  Westergard and an anonymous engineer called “Will” in this interview. They discussed why they organized, how they did it, and what lessons their experience might hold for the future of tech organizing.



How did the first conversations about workplace organizing get started?


There was one worker — I’ll call her Jane — who really kicked it all off.

Jane had a tight-knit circle of friends at the company. They had all attended a women’s coding bootcamp in San Francisco called Hackbright Academy, and then came to work at Lanetix.

At some point, they created a channel in the company Slack for them to chat in. It became a place for personal conversations, such as those devoted to supporting coworkers who were going through rough times in their personal lives. But at some point, the issue of paid time off came up.

Lanetix offers two weeks of paid vacation. In practice, however, the number of days people could take off was up to their manager. People kept saying, “I’m annoyed with the way my manager is always pressuring me to not take time off.” The feeling was that managerial discretion ruled the day.

Jane brought this issue to her manager, and encouraged other people to do the same. There was a manager in the channel, so management was already aware of these conversations. But Jane began to push the issue — and she eventually won a big concession. The CEO announced that people would be receiving more paid time-off days.


Jane had a history of going to her manager with concerns that were shared by other engineers at the company. She was by no means alone in doing so, but she was definitely one of the more vocal folks.


Soon after, however, another issue began coming to a head. In the beginning of 2017, the CTO had announced a grandiose plan to compete with Slack. By November 2017, it was becoming clear that the engineers, including Jane, had done everything that was asked of them in building the Slack competitor — but the market wasn’t into it. The usage numbers were poor.

By this time, Jane and her team had been working on this project for several months. And then, as it became clear the project wasn’t succeeding, they were asked to totally change the UI (user interface) in just a couple of weeks. Taken unaware and shocked, Jane asked, “I’d like to understand what happened — how did we end up here? And what can we do to ensure that we don’t end up in this situation again?” I believe that this, combined with her history of being vocal about other issues at the company, was what led to her being fired abruptly on November 26, 2017.


By all indications Jane was performing well. She was in good standing. She actually had just been made a team lead and given more responsibility. But it seems like her questions did not go over well.


How did people respond to Jane’s firing?


The engineers were all pretty freaked out. We didn’t understand how she could be fired out of the blue, especially given her stellar performance, and we were worried for our own jobs.

The next day, my “chapter lead” — a kind of small-scale manager with a couple of reports — sat a number of us down. He had received talking notes from management. He literally had them written on index cards. He told us, “Don’t be divisive. If you have any problems, don’t talk about them among yourselves — bring them up to your manager through approved channels.”

Then, in my one-on-one, he told me he had been directed to speak to all of his reports regarding an external Slack. He said that if I was a part of it, I should leave, as anyone that was a part of it was on management’s shit list.


It sounds like Slack played a key role in organizing the workplace.


Initially, we were talking in a channel on the company Slack. I was unaware of an external Slack at this point.

But we did end up creating a new Slack to discuss these issues, as we did not feel that management was in the right in trying to control the flow of information. We believe that workers should be allowed to speak amongst one another however we like. We planned to use our new Slack to discuss how best to raise our concerns and frustrations to management.


Management made a lot of coercive statements about our external Slack. They didn’t want us using it, since no managers were on it. They even shoulder-surfed around the office to see whether engineers were using it.

The open layout of the San Francisco office made it relatively easy for management to see who was using the workers’ Slack. But that’s also why engineers were using it — the open layout increased the chance that managers would overhear their conversations. So in San Francisco, engineers were talking on Slack even when they were sitting right next to each other.

This campaign would not have happened without Slack. Upwards of 90 percent of the activity was happening on Slack. It couldn’t have happened any other way because the company had two different offices: San Francisco and Arlington, Virginia.

So Slack became both an organizing tool and a grievance. People were incensed not only that Jane was fired, but that they couldn’t use their own Slack where management wasn’t allowed. Bosses are the best organizers, as they say.



How did the idea of a union emerge in the aftermath of Jane’s firing in November 2017?


It didn’t happen overnight. Overall, the process took about three months.


One friend of Jane’s had been a journalist before she became an engineer, and had worked at a newspaper in Milwaukee that was a NewsGuild–Communications Workers of America shop. She was the first one who said the word “union” aloud in the week after Jane was fired.

She didn’t say we should unionize — not at first. But she said something to the effect of, “When I was in a union shop, workers couldn’t be arbitrarily fired.” So the idea was out there, but it took awhile before people began discussing it seriously.


What were the specific incidents that made people start discussing it more seriously?


After Jane’s firing, we all came together and decided to write a letter. We asked the company to 1) recognize our right to organize and 2) do right by Jane.

At first, management ignored us. Then, finally, the CEO reached out to try to schedule one-on-one meetings with each of us. We told him we wanted to meet as a group, since we had written the letter collectively. So he scheduled a conference call.

On the call, he was very condescending and very paternal. The gist of it was, “Hey, I’m sorry this hasn’t been handled well — but Daddy’s home now, and I’ll make it all better.”


He repeatedly compared himself to a parent and us to his children. That pissed us off.


The takeaway was that this was going to be a long, dragged-out fight. The CEO said a lot of vague niceties, but it was clear he was not going to address our concerns. Still, the majority of the engineers wanted to wait and see what management did. At this point, talk of unionization had begun, but it didn’t have a lot of support.


Then, on January 5, 2018, we received an email from management that said the company was planning to open an office in Eastern Europe sometime in the near future. It didn’t have any specifics — it was intentionally vague. This same email disinvited nearly all of the engineers from the annual company retreat. Those two things together rubbed a lot of us the wrong way.


That email was the turning point. People had begun coming around to the idea that if we liked our work and our workplace enough to stick around instead of looking for a new job, then maybe we should unionize.

But it was the threat to outsource our jobs that pushed the majority into supporting unionization. To be clear, we didn’t think it was a credible threat. The codebase was exponentially the most complicated I’ve ever worked on — it wouldn’t have been easy to outsource the coding labor.


Management didn’t have a plan. They just wanted to intimidate us. But it backfired, because that’s when the majority decided to unionize.

We had overwhelming support. On January 16, 2018, we sent a letter to management outlining our grievances and announcing our intention to unionize with NewsGuild-CWA. We also filed a petition with the National Labor Relations Board (NLRB).

Ten days later, Lanetix management fired the entire engineering staff. NewsGuild-CWA has since filed a complaint with the NLRB, claiming that the mass firing was illegal retaliation for trying to unionize. We also requested that the NLRB and the courts intervene immediately to force management to give us our jobs back, with back pay.



You mentioned that the letter to management listed your grievances. Some people find the idea of a tech workers union implausible, because they find it hard to believe that tech workers could have grievances in the first place. After all, software engineers are relatively well paid. What’s your perspective on that?


One issue that really motivated us — and an issue that I’ve discussed a lot with friends who work in tech — is the role that managerial caprice plays within workplaces where there are no real defined rules.

The general trend for most working people outside of this very privileged white-collar context is endless micromanagement and highly restrictive work rules — think of Amazon warehouse workers. The discipline becomes ever more fine-grained. In tech, by contrast, the discipline isn’t even acknowledged.

A real turning point in the campaign came when the engineers recognized how much they had in common. People would say things like, “I thought that I was having a bad time because I wasn’t as productive as other people or I didn’t fit into the culture.” But as soon as they started to compare notes, they realized that each manager was just trying to individualize the complaints that everybody had.

For us, that was the major attraction of going the NLRB route. We’d have a contract that would specify exactly how much time we got off, exactly when we would be on call, and so on. There couldn’t be any more psychological gamesmanship between worker and manager.

People didn’t even want that much more time off, or that much more time not on call. They just wanted to know where they stood. They didn’t want to have to engage in highly personalized bargaining all the time. It wasn’t that there were rules that favored management — there were simply no rules, and this ambiguity worked in management’s favor.


Software engineers do command relatively high salaries, but compensation isn’t the only thing people care about. If you have a manager acting capriciously or vindictively, or you have a lack of clarity and transparency around rules, it can deeply damage a company’s culture. Sure, many engineers could simply leave and find another job somewhere else. But if the workers are loyal to one another, and like what they’re working on, they will want to improve their workplace rather than jumping ship.


The working conditions of a software engineer are obviously far better than those of most workers. But software engineers often work long hours. Could that issue offer another way to start organizing conversations in tech?


There is an inherent boom-bust cycle to software engineering. In other types of engineering, like civil or mechanical engineering, you have a static input that leads to a static output. If you want to build a car door, you start with a known set of materials and always apply the same process to get the same result in the same amount of time.

In software engineering, though, you have a variable set of inputs and a variable set of outputs. This leads to a common scenario where projects run behind, as you adapt to changing requirements and changing situations. That’s why software engineers often have to work crazy hours towards the end of projects to push them over the finish line.

It can become a vicious cycle. Anyone who has worked in software will recognize the “hero mode” cycle: you and your team are working nights and weekends to fix a crisis, but you’re under so much pressure that you’re cutting corners and writing bad code. If you manage to get the project done, you’re hailed as heroes, but soon enough, that sloppiness will trigger another crisis — and it’ll be even worse than the last, if no one actively works to break this cycle.

That’s a deteriorating situation that only leads to workers getting more and more exploited. But that situation isn’t inevitable. You don’t have to accept it. You can organize and take collective action to push back on management, if individually pointing out the recurring problem goes nowhere. You can create a better workplace, and unionizing will help you get there.


Another promising issue to organize around is the pervasive anxiety about professional development. Software engineers are expected to be constantly learning new technologies in order to stay employable, and they’re terrified of falling behind.

In fact, this came up a lot at Lanetix. When two of our senior engineers left, people were afraid they would no longer have as many opportunities to keep up on that skills treadmill. Because those two engineers were extremely knowledgeable, and simply by working with them, we were all learning things that materially improved our long-term job security.

You see this dynamic a lot. If management decides to institute a speed-up, or make people work longer hours, the learning opportunities are what get lost. That’s really threatening to someone who’s trying to make a living as a software engineer.

So I think the desire for engineers to sharpen their skills on the job could provide a motive for organizing. But it’s tricky — it’s not a straightforward story about labor solidarity. Because workers might band together at a particular company in order to keep their skills sharp, but only so they can prevail in market competition against other workers. The question is how to take people with these anxieties and draw them into a broad social project that isn’t particularistic.


That raises an important point. How evenly distributed are the opportunities for organizing in tech? I’d imagine it’s easier for certain workers to push back on management than others.


We always have to consider the context. Your capacity to push back on management will depend on what kind of company you work for and your own status as an engineer. An engineering staff composed mostly of recent bootcamp graduates will have a much weaker negotiating position than a bunch of senior engineers.


There’s a growing amount of polarization within software engineering. The day-to-day work of people with the title “software engineer” is becoming increasingly diverse. Some software engineers are doing work that requires extensive theoretical training, like writing compilers. Others are recent bootcamp graduates who know just enough to crank out a JavaScript app.

That presents certain organizing challenges, because the bargaining power of the former is much greater than the latter. In fact, despite the rhetoric about high demand for software engineers, many bootcamp graduates struggle to get a toehold in the industry.

Further, the software industry is highly uneven. It’s usually heralded as a glorious post-Fordist utopia of labor relations, but the reality is that many software companies are incredibly sketchy fly-by-night operations. They’re terribly managed and will produce a useless product, if they produce anything at all, statistically speaking. So, there are companies that are doing extremely well, where collective action is more feasible because it’s objectively in management’s interest to play ball with you — and then there are the majority of companies, where that’s not so clearly the case.


Finally, what role do you think leftist politics can play in these organizing conversations? Over the past couple of years, a small but significant “tech left” has emerged, embodied by organizations like the Tech Workers Coalition and the Tech Action Working Group of the Democratic Socialists of America’s New York City chapter. Did awareness of this political current among your coworkers play any role in the organizing campaign?


If you’d asked me the day before we were fired how aware my coworkers were of the “tech left,” I would say not at all. But then after we were fired, it turned out that a few of my coworkers had been to a Tech Workers Coalition march in San Francisco. And several had friends in the San Francisco chapter of DSA.

Still, as much as I may have wanted it to be the case, we weren’t carrying forward the banner of the tech left. We were motivated by the purest porkchop stuff: defend Jane, defend the Slack, get more paid time off.

But even though politics wasn’t a major motivator for our organizing, management did try to use politics to dissuade us from organizing.

Individual managers, when they wanted to appear “woke” in the face of incipient unionization, would always talk about how bad Trump was. As long as you’re against Trump in your heart, the thinking went, then you can’t be an evil union-busting manager. It was fascinating to me that management tried to use opposition to Trump as a shield to avoid directly confronting the tense relationship between the worker and the boss — the tensions weren’t new, but acknowledging them was.

That’s ultimately what I found so gratifying about the whole experience: the feeling that the veil had been lifted. Everyone said something along these lines: it’s not that everything suddenly got terrible, it’s that we became aware of how rotten everything was the whole time.

Having been your classic internet Marxist, I thought I was prepared for it. But to actually see it happen was really remarkable. People who were gunning for management positions a month earlier were suddenly very clear which side they were on. You very quickly learn whose side you’re on.