How to organize Amazon

The failed campaign in Bessemer, Alabama, does not prove that workers don't want a union. It shows that workers see no logic in the labor movement's one-shop-at-a-time strategy.

Amazon defeated a high-stakes union campaign last week in Bessemer Alabama. Celebrities and politicians all the way up to President Biden encouraged the workers to say yes to the Retail Wholesale and Department Store Workers Union (RWDSU). They said no. Progressives and socialists who were hopeful about the outcome blame the anti-union campaign of propaganda, threats, and promises. But that’s like blaming gravity for a plane crash. A vicious anti-union campaign is a given. Today’s unions were originally forged against anti-union campaigns based on not only propaganda but also armed violence. So what went wrong, and how can workers at big, distributed companies like Amazon and Walmart get organized?  

Before I crossed over into politics, I was a union organizer throughout the 1990s all over the country, with many unions, including once with the RWDSU. I worked as a staff organizer and as a worker inside plants during union campaigns. I learned that workers will support a union when the union shows a path to real power—power great enough to win significant gains in wages, benefits, and on-the-job respect and autonomy. I learned that every workplace contained many incredible leaders who had all the skills and experience necessary to organize their coworkers and defeat the anti-union campaign. But they would only engage if they believed the union had a winning strategy and trustworthy staff and leadership. I learned that the most respected workplace leaders joined campaigns with smart strategies and would slam the doors in the faces of organizers who came with doomed strategies. 

Organizing one Amazon warehouse out of more than 100 in America is a doomed strategy. If the workers had voted yes, Amazon would have dragged their feet for years before signing a substandard contract, making life a living hell for the workers through it all. They might have closed the facility eventually. They might have excluded the organized workers from raises and perks given to the rest of the workforce. Many Bessemer workers could see all this and knew that’s what they were in for. And they knew that such an outcome would only make it more difficult to organize the rest of Amazon by turning Bessemer into a lesson in the futility of organizing.

On top of that, Amazon showed the workers other retail union contracts that have far lower wages and benefits than theirs. Why should the Bessemer workers have believed that the union would fight for significant increases for them? Are there any examples today of unions fighting for significant gains? Such examples must be out there, but they are hard to find these days. Much easier to find are examples of unions signing contracts with little or no improvements. 

As a union organizer, I was trained to tell workers that we couldn’t promise significant gains—but just that having a union was better than not having one on principle. That outlook led unions through the 80s and 90s to follow a “hot shop” organizing strategy. We looked for workplaces where workers were so angry with their employer that they would join a union regardless of whether it would give them power over their employer or not. To win these campaigns, professional staff organizers were brought in at a ratio of one for every one or two hundred workers. Our job was to develop a personal relationship with workers that we would use to remind them of the anger they had felt when they first called the union and hold them to it all the way up to the union election. That approach doesn’t scale to entire industries or even to large workplaces.

For a few years, at the end of my short union organizing career, I practiced a different kind of organizing with a small team of organizers at a statewide healthcare union. After a deep dive into labor history, specifically looking at how millions of workers organized the world’s most powerful corporations in the 1930s and ‘40s. Instead of going after hot shops, we talked about a strategy of organizing a whole state’s healthcare sector, promising a 10 or 20-year plan to gain the power to influence state funding of healthcare in a place where much of healthcare depended on state funding. We went after some of the highest-paying and largest facilities first, on the logic they had more to give. 

Instead of using personal relationships to guilt workers into sticking with the union after a moment of anger, we focused on finding the most respected leaders in every workplace and partnering with them as they organized their coworkers with little intervention from us. The hard part was identifying who those leaders were, but once we did, if we could get them into a room together, and offer them a sound strategy to win, they always went for it and always succeeded. We won a long string of campaigns with unheard-of 3-to-1 or even 4-to-1 margins of victory that dumbfounded our national bosses—unfortunately not in a way that made them want to try our approach elsewhere. 

There is a way to organize workers at Amazon, Walmart, and other big, distributed employers. To see it, Amazon workers will have to look back to the stories of how today’s big unions were originally formed. This is the story of how half of America made it up into the middle class when workers organized whole industries at once and used their power to win massive gains in wages, benefits, and on-the-job power. 

The campaigns that achieved those victories used new technologies to overcome the challenges posed by the new massive and distributed industrial workforces of the 20th century. Moreover, they were run by new, breakaway, worker-led unions of the Congress of Industrial Organizations (CIO)—unencumbered by the baggage of the old, failing unions of the American Federation of Labor (AFL). Importantly, though, the CIO unions began with the blessing and support of the old labor movement as the AFL’s Committee on Industrial Organizing. 

The RWDSU itself, a CIO union, became a major national union not by nibbling at the edges of large retail chains but by running nationwide, coordinated strikes against whole chains in many cities at the same time. That was only possible thanks to technologies that were still coming into their own such as the telephone and air travel.

CIO unions, such as the United Auto Workers and the Steel Workers Organizing Committee, organized America’s largest corporations by going after a majority of their facilities more or less all at once. They promised workers that huge gains were possible because of their incredible market power and profitability—and delivering on those promises by negotiating hard and being willing to run long strikes when necessary. 

In manufacturing, facilities sometimes employed more than one hundred thousand workers under one roof. The CIO unions used relatively new technology, such as sound trucks, to communicate with workers in that environment. Workers printed their own newspapers to be distributed inside these city-sized factories. Union organizers formed large democratic committees of workers inside the factories who were responsible for organizing their coworkers. In these campaigns, the ratio of staff organizer to workers was not one staffer for 100 workers but one for tens of thousands, if not hundreds of thousands. 

Enough about the past. How can companies like Amazon be organized today? 

First of all, Amazon workers need to organize themselves, not be organized by professional union organizers or activists. Several months ago, I got in touch with an Amazon worker who received some media attention for trying to organize Amazon workers against Covid-related unsafe working conditions. He invited me to a Zoom call on which dozens of organizations and activists were already overwhelming the effort in highly counterproductive ways—and yet none had even sprung for a professional Zoom account that would let a call last for more than 40 minutes. Amazon workers will trust and follow other Amazon workers. They will not trust or follow activists who have all kinds of random and often selfish agendas. They are wise to trust their coworkers more than activists because their coworkers will have better judgment than people motivated by various ulterior motives. Nor will they trust staffers from unions that have presided over the collapse of union representation and wages for decades. Why would they? 

Amazon workers can begin to organize when a handful of respected leaders among their ranks join together to start their own union. That idea sounds preposterous to most professional union organizers—but it is exactly how almost all of their unions were originally formed. 

To form their own union, that first handful of workers need only to announce their intention to do so and have a website where other workers can get in touch with them, and lays out the basics of their strategy. They should ask the AFL-CIO for support, though it’s unlikely they will receive any. Today, however, they can turn to the public via sites like and social media in general. One Amazon organizer has already raised at least $40,000 on The union doesn’t need a lot of money to get started. In the beginning, all that’s needed is some money for incidentals such as a Zoom subscription and a website. Larger expenses would include a labor lawyer and a PR flack.

Once the new worker-run Amazon union has launched, its only job for a long time will be to work consistently to gather the names of all the respected leaders in every Amazon facility. This can be done by simply having a website that allows Amazon workers to confidentially nominate coworkers they respect and follow in their facilities. How would workers hear about the site? The union could spread the word in the mass media, run social media ads targeting Amazon workers, or soliciting non-worker volunteers to leaflet outside facilities. 

It is important that workers don’t start standing up to organize openly until enough respected leaders are identified—then all those leaders will go public at the same time. That way, they will have safety in numbers. More importantly, this will guarantee that the union’s public identity begins with the most respected workers. I worked on many campaigns where a few workers who were not respected and who might be seen as mean, lazy, or having personal axes to grind with management, were the first to stand up. That spelled instant death for the union effort every time. 

As leaders are nominated, they need to be contacted by others who are already on the organizing committee and felt out. All the conversations should be tentative: “If we find enough leaders, and if the leaders want to form a union, then here is what we could do. How do you feel?” 

Most respected leaders will all say the same thing in response, which will be something to the effect of, “I don’t want to be involved in a suicide mission, but if you do get all the leaders, and they do want a union, then I’ll back the effort.” It’s a game of chicken and egg. This is how you put together the egg. 

When I started organizing in that way, I was amazed at how differently workers responded than in the campaigns where our strategy was, as one mentor put it, “To push workers out on a ledge and then start sawing.” When our strategy was essentially to use pressure tactics to trick workers into stepping out for a union inside of a doomed strategy, the smartest, most strategic workers—who also tended to be the most respected—wouldn’t even talk to us. On the other hand, when we approached workers with an offer to work with them as consultants and advisors on a union campaign that they would run and made it clear that we would only stay if nearly all of the workplace leaders wanted us there, then leaders were almost unanimously open. They were excited that we were doing the work of assembling their peers and looked forward to the discussion and their decision on whether to organize that would happen there. When leaders decided to undertake a union campaign in that context, we found that the anti-union campaign had almost no impact. It even backfired sometimes. Workers trusted their leaders, and leaders who committed to each other, not to a union organizer, stood strong until the end. One very helpful tool was a worker-written newspaper that included any kind of writing that any worker wanted to contribute. We even published anti-union pieces, demonstrating democracy and openness in stark contrast to the boss’s anti-union campaign. 

In the case of Amazon, because the numbers are so large, it might take a few years for enough conversations with leaders to take place before it’s time to go public. That event might be a big public Zoom call, hyped in the national media, attended by thousands of Amazon workers, customers, shareholders, and reporters, in which a large handful of the most representative and moving leaders from several facilities speak from the heart about why they want to organize and lay out their strategy to win a union. 

In that moment, a public petition to Amazon would be released, asking Amazon not to hire any union-busting consultants and not to fight the unionization effort, signed by the leaders from all Amazon facilities. This petition would then be posted on a website, with pictures and quotes or full articles written by each member of this newly-formed organizing committee. 

Because we’re talking about Amazon, the best way to communicate to the mass of Amazon’s workers and customers would be through the national media. Of course, workers will also be encouraged to join Zoom calls and subscribe to an email list. An Amazon union campaign with compelling worker spokespeople will be big news. 

There is no reason for the kind of worker organizing committee at Amazon to use the National Labor Relations Board election process. The leads should begin negotiating, through the media, with Amazon’s leadership directly. The primary leverage they hold is over Amazon’s stock price. Suppose Amazon refuses to deal with the union and make improvements the union is asking for. In that case, the union can announce that they are pursuing long term plans for industrial actions—such as strikes or more creative tactics such as “work-to-rule” slowdowns—as well as plans to call for a boycott of Amazon by customers. 

The kind of campaign I just laid out would succeed at Amazon if the right group of workers got it started. They will have to be committed to improving all Amazon worker’s lives, and driven primarily by a desire to succeed in building a union with integrity, not by any selfish or egotistical impulses. My experience organizing workers taught me that the vast majority of respected leaders in any kind of facility are the best people you’ll ever meet, driven by astoundingly pure motives when it comes to matters that affect their coworkers. 

Not only would that strategy work with Amazon, but with other large and famous employers such as Walmart, Uber, DoorDash, fast food restaurant chains, and others. There’s no reason these efforts should be limited to one country. Amazon has 180 warehouses around the world. To truly hold power, workers will need to organize one big union without borders. If you’re a worker who wants to build a new kind of national union movement, and would like to brainstorm about it please send me an email at