It’s All Bullshit

Free food, nap rooms, wellness walks, unlimited vacation days: such are the workaday perks of a job in tech. These perks, along with the six- and sometimes seven-figure salaries that accompany them, are, we’ve long been told, well-deserved. Not only are tech workers portrayed as feverishly hardworking, they are the epitome of innovation and productivity.

However, it’s become apparent that they aren’t as productive as we’ve been told to think. Ex-Meta employee Madelyn Machado recently posted a TikTok video claiming that she was getting paid $190,000 a year to do nothing. Another Meta employee, also on TikTok, posted that “Meta was hiring people so that other companies couldn’t have us, and then they were just kind of like hoarding us like Pokémon cards.” Over at Google, a company known to have pioneered the modern tech workplace, one designer complained of spending 40 percent of their time on “the inefficien[cy] overhead of simply working at Google.” Some report spending all day on tasks as simple as changing the color of a website button. Working the bare minimum while waiting for stock to vest is so common that Googlers call it “resting and vesting.” ​

In an anonymous online poll on how many “focused hours of work” software engineers put in each day, 71 percent of the over four thousand respondents claimed to work six hours a day or less, while 12 percent said they did between one and two hours a day. During the acute phase of the Covid-19 pandemic, it became common for tech workers to capitalize on all this free time by juggling multiple full-time remote jobs. According to the Wall Street Journal, many workers who balance two jobs do not even hit a regular forty-hour workload for both jobs combined. One software engineer reported logging between three and ten hours of actual work per week when working one job, with the rest of his time spent on pointless meetings and pretending to be busy. My own experience supports this trend: toward the end of my five-year tenure as a software engineer for Microsoft, I was working fewer than three hours a day. And of what little code I produced for them, none of it made any real impact on Microsoft’s bottom line—or the world at large.

For much of this century, optimism that technology would make the world a better place fueled the perception that Silicon Valley was the moral alternative to an extractive Wall Street—that it was possible to make money, not at the expense of society but in service of it. In other words, many who joined the industry did so precisely because they thought that their work would be useful. Yet what we’re now seeing is a lot of bullshit. If capitalism is supposed to be efficient and, guided by the invisible hand of the market, eliminate inefficiencies, how is it that the tech industry, the purported cradle of innovation, has become a redoubt of waste and unproductivity?

In tech, bullshit jobs—which the late David Graeber defines as “a form of paid employment that is so completely pointless, unnecessary, or pernicious that even the employee cannot justify its existence” even though they are obliged to pretend otherwise—come mostly from bullshit projects. At Google, such projects are aplenty. According to Killed By Google, an archival project that documents discontinued Google services, products, devices, and apps, the company has discontinued nearly three hundred projects since its founding. These range from software systems to help businesses distribute and manage job applications (Hired by Google) to social media platforms that tried to mimic Facebook’s success (Google+) to manifestly uncool wearable technology (Google Glass). But projects at that scale are far and few between. Many were small and hackathon-sized in ambition. All failed to make a splash on the balance sheet. These “dead” projects are also just the tip of the iceberg. The website only documents projects that have been publicly launched; untold more have been nixed before ever being announced.

Google has always had a founders-first mentality; it’s ingrained into the company’s DNA. According to a former employee, Google tries “to reproduce the circumstances that led to their initial success” by recreating the environment in which it was founded: On top of having the flexibility to decide when to work, how to work, and, in some cases, what to work on, Google allows employees to spend 20 percent of their paid working time on side projects. Senior employees can also pitch project ideas to upper-management and get funding to work on them, while junior employees are encouraged to “take ownership” over their assigned projects and to act as if they were the CTO of it. Google even has a literal startup incubator—called Area 120—for employees to pitch projects and start new companies.

To tech optimists, projects that don’t immediately contribute to a company’s bottom line are seen as sustaining the intrepid spirit of innovation—necessary to achieve breakthrough technologies that could generate profit at some point in the future. As one employee puts it, Google is into “generating luck,” since they are willing to try “a whole bunch of stuff in the hopes that a few efforts will pay off” for the rest of their failed projects. Many will fail but those that succeed are supposed to rake in so much cash that they pay for the failures. But for this model of innovation to work as intended, employees have to believe in it; to the faithless, it has become little more than a means of personal advancement.

The compulsion to launch new projects in order to scale the corporate ladder has become so ubiquitous that employees call it the LPA cycle: launch, promo, abandon. “The [promo] incentive throughout,” as explained by a former employee, “is to create a product, launch it, apply for promotion, and move on to bigger and better things as soon as possible.” For years at Google, promotions weren’t given at the discretion of an employee’s manager; instead, employees initiated a promotion by compiling a “promo packet”—a collection of essays that explain why their contributions merit a promotion, corroborating evidence, and recommendations from teammates. This packet was then evaluated by upper-level engineers and management, who then determined an employee’s “impact.” But because the vast majority of projects do not, in any direct way, contribute to revenue, impact can be difficult to assess. So the number of product launches became a proxy measure of value. As one Google employee posted on Hacker News: “You cannot get promoted beyond a certain level in this place unless you ‘launch’ something big. So what do you get when you add all of these perverse incentives? Nine thousand, eight hundred and eighty-three chat apps.”

Writing promo packets became an artform at Google, one untethered to productivity. Instead of solving legitimate problems, many engineers found themselves gravitating toward tasks that could build their case for a promotion. This frequently led teams to build new products that compete internally, creating confusion for all parties involved, while pressing engineering tasks get neglected for months at a time. Sometimes, projects were simply redundant. One employee told me that “there are some people who launch a project then switch to another org to launch the same project.” The performance of usefulness thus replaced the act of actually being useful.

Even failed launches could land someone a promotion. One employee told me the story of his team’s tech lead, who was “super-gifted [and] one of the smartest engineers” he has worked with, but “going from project to project without a final launch is routine to the way he did things.” Even after eight years at Google where he worked his way up to the title of staff engineer—a highly respected role—none of his code has ever made it to Google’s production servers. Instead, he has swung from one failed or canceled project to the next, collecting promotions along the way. As the employee put it to me: “It’s literally failing upwards with no end in sight.”

Bullshit projects are abundant, in large part, because of the LPA cycle, and its proliferation is only aggravated by Google’s army of increasingly despised middle managers. Employee dissatisfaction with middle management is hardly a new phenomenon. Writing in the 1990s, labor economist David Gordon argued that U.S. businesses are addicted to expanding the ranks of middle management, resulting in corporations that are “Fat and Mean,” which is also the title of his last book. According to Gordon, the waning competitiveness of American industries could be attributed to the bloating of middle managers who are at once expensive (fat) and prone to suppress wages (mean).

Google managers may not be mean, but they have—by Gordon’s definition—become fat. In the early days of the company, founders Larry Page and Sergey Brin thought that middle managers were a layer of bureaucracy that obstructed engineers from doing good work, so they kept the number of managers to a minimum—and even experimented with eliminating them altogether. In 2009, they funded a multiyear research initiative called Project Oxygen to test whether managers mattered (the results indicated that, yes, apparently, they did). But while the project gave Google the data to justify the role of management, the ensuing lack of scrutiny over the role has resulted in bloat. Today, about 15 percent of Google’s workforce is made up of middle managers, roughly one manager for every five to six employees, far surpassing the average manager to employee ratio in the service sector of one to fifteen. Where it was possible for a hundred engineers to report to a single manager in the aughts, most engineers are now placed on teams of no more than a dozen, frequently less.

The goal for managers, though, is to grow their teams as much and as quickly as possible since the number of people who report to them functions as a measure of their own “productivity.” One Googler told me that management is “incentivized to grow their own team blindly, like a cancer cell.” To demonstrate their own managerial prowess, they must sell the illusion that whatever it is their team is doing is good for business and users, even if it clearly isn’t. In the absence of concrete metrics to evaluate a team’s productivity, headcount becomes a key, if wildly inaccurate, metric. As a result, management is forced into a vicious cycle of upselling their team’s importance in order to be allocated a higher headcount, meaning they then have to come up with new projects to justify the new headcount. The more workers there are, the more important the work must be, and the more important the work is, the more people must work on it.

Another major stumbling block to actual innovation is that projects that may not actually be bullshit get nixed all the time, transforming all that work into bullshit. The primary reason projects get cut is because of Google’s pathological addiction to reorganization, or reorgs—when organizational boundaries get shuffled often because of a power struggle among managers. Sometimes reorgs are productive and can result in the elimination of redundant work. However, reorgs aren’t always economically motivated. If a manager can use a reorg to expand the size of their team, it is almost certainly the right career choice even if it doesn’t make the company more efficient. For employees, this can inhibit their ability to perform their jobs, participate in launching projects, and, as a result, build their case for promotions. According to a why-I-quit-Google blog post by former employee Michael Lynch, this was a central concern:

Six weeks before the performance period ended, my project was canceled. Again. Actually, my whole team was canceled. This was a common enough occurrence at Google that there was a euphemism for it: a defrag . . . My teammates and I all had to start over in different areas of the company.

In some cases, workers anticipating a re-org may stop working altogether since they expect their project to get canned. Lynch sums it up: “Google kept telling me that it couldn’t judge my work until it saw me complete a project. Meanwhile, I couldn’t complete any projects because Google kept interrupting them midway through and assigning me new ones.” Between the drive for headcount and the never-ending square dance of reorgs, it’s clear that middle management is just as responsible for waste as their promo-hungry employees. Bullshit, it would appear, permeates every level of the organization.

Innovation at Google is clearly in crisis. After establishing search advertising as a veritable cash cow, Google has spent hundreds of millions trying to reproduce that initial success—with little to show for it. In its early years, the company quickly developed platforms like Gmail, Google Maps, among others that locked users into the company’s ecosystem, driving traffic and ad spending. Today, the platform economy has calcified. The platforms that most people might want already exist; and major ones like Facebook and X, the platform previously known as Twitter, are even past their peak.

“Google got lucky 15 years ago and managed to turn on an absolutely massive firehose of money in ads,” wrote a Google employee on HackerNews. This model of innovation is also the reason Google employs more people than it needs, according to the employee, who added that because the company may never “strike it lucky [again] . . . they have to settle for attempting to starve potential competition of talent.” For Google, however, this stagnation hasn’t produced an urgent existential crisis: advertising revenue has grown to $224 billion from $43 billion in 2012. Today, Google has over 90 percent of the search market in the United States. With virtually no credible competition to threaten the firm’s position, it can sit back and collect its fat paycheck.

Google’s monopoly position in search has long been a problem with critics demanding stronger antitrust regulation; this year, the U.S. government is finally taking Google to trial over illegal partnerships with phone makers (like Apple and Samsung) and browser companies (like Mozilla) that effectively block rival search engines from gaining a foothold. These partnerships mean that millions of people use Google’s search engine by default, which has, according to the Justice Department, allowed the company to secure and maintain its monopoly. If Google loses, it could undermine the firm’s advertising empire—and set a precedent for curbing the power of other platform monopolies like Amazon, Meta, and Microsoft.

But even with stronger regulations, the manifold problems that these tech giants create for society will persist. Unlike the industrial firms that Marx and Smith wrote about, tech firms like Google differ in one key aspect: how they turn a profit. Where industrial firms extract profits from the gap between value produced and what workers get paid in wages, internet platforms rely foremost on rents, which generate profit on the basis of existing property, intellectual or otherwise. No additional labor is needed for each unit of output; profit thus becomes fundamentally extractive, or “unproductive” as Marx, or even classical economists like Smith and Ricardo, would say. For companies like Google, rent comes from owning the infrastructure through which we live our digital lives. The vast majority of Google’s windfall profits—$60 billion in 2022—come from rents.

For workers, this model has grim implications. Once a cash-generating platform is built and attains a monopolistic position, the primary, and far less demanding, task becomes maintaining the platform. Moreover, where older firms might have to stave off competitors and engage in practices like predatory pricing to keep the competition at bay, platform monopolies have a built-in moat. Since a platform’s value comes from its number of users—a phenomenon known as the network effect—competing against large platforms is nigh impossible since its monopoly position makes it inherently more useful. This means that workers are freed up to do other things.

Google’s strategy has been to diversify revenue streams with new products. But this has clearly proven to be a bit of a challenge. Even after over two decades of R&D, advertising remains as Google’s unmatched profit-maker, with their cloud division clocking only a distant second. “Google hoovers up as much talent as they can in hopes that they’ll strike it lucky and turn on a second or third revenue faucet,” according to one Googler, “but spoiler alert: they never will.”

Google itself might finally be beginning to come to terms with this. Last year, the company ended its practice of doing promo packets and has shifted to a more traditional approach where direct managers determine promotions. Earlier this year, the company also announced that it would be culling leadership and management roles. Management has also emphasized the idea of “landing not just launching” in response to the unchecked proliferation of bullshit projects throughout the company. After laying off roughly twelve thousand employees, or 6 percent of its workforce, in January, CEO Sundar Pichai vowed to increase focus and run on fewer resources. Could it be that Google’s era of abundance—and abundant bullshit—is coming to an end?

Tech workers have had an unusual amount of agency over the past decade. In recent years, a historical shortage of labor in a sector awash with capital has emboldened them to demand more from their employers. Pinterest employee Ifeoma Ozoma, for example, exposed her employer for their discriminatory practices. Frances Haugen, a product manager at Facebook, disclosed internal documents to the Wall Street Journal and the Securities and Exchange Commission revealing the detrimental societal impacts of her employer’s platform.

Others have participated in collective action. Kickstarter employees, for example, won their union and successfully bargained a contract in 2022. Amazon workers led multiple walkouts to support the global climate strike. Googlers unionized, fought against increasing corporate hostility, and challenged their company’s unethical partnerships. Even in China, tech workers banded together to campaign against the 996—9 a.m. to 9 p.m., six days a week—schedule that was expected of them. According to the Collective Action in Tech database, there have been hundreds of publicly reported actions since 2019, some involving thousands of people. While these are promising developments, tech workers still have a long way to go before they are organized enough to have enough leverage to counter structural problems in the industry.

But as internet platforms enter a period of slowdown, if not outright decline, bullshit jobs and projects will become a justification for management to trim the fat and exert greater control over employees. Just this past summer, Google revamped its return-to-office policy to include badge tracking, which would allow the company to identify non-compliant employees. Venture capitalist Keith Robias also lambasted Google employees for being “entitled” and happy to “sit at their desks and do nothing.” If left unchallenged, management will—as they have already started to do—shift the balance of power back to their side.

Herein lies the quandary: If the power that workers have over their employers is the ability to withhold their labor, admitting to working a bullshit job robs them precisely of that power. Moreover, the fact that platforms profit from extracting rents rather than productive labor puts workers in an even worse position: withholding their labor put the kibosh on profits. The recent layoffs illustrate this point; Google laid off twelve thousand employees without a single one of their services faltering. Elon Musk cut X’s workforce by 80 percent without crippling the platform’s fundamental usability. In some cases, withholding tech workers’ labor will have clear ramifications on platform services. New updates won’t roll out. Bugs won’t get fixed. The sustainability of the platform could come into question. However, if in aggregate, profits are generated through extractive practices that do not require the labor force these companies have built up, what leverage do these workers have? The obvious answer is that they can withhold future profits by halting the creation of new products and revenue streams. But even this won’t work if the projects they work on are bullshit. 

In the end, the tech workers themselves are not responsible for the proliferation of bullshit. Nevertheless, the work they do (or don’t do) reveals that the problem, at its core, lies in the business model of the platform economy, one that depends not on productive labor but on rents. In our present economic paradigm, rents are undifferentiated from productive work where the creation of surplus value vis-à-vis the exploitation of labor lies at the heart of profitability. Making explicit these differences, however, would reveal that these platforms are not all that different from the banks of Wall Street: they are not making value, they’re taking it. And as long as this remains the underlying form of profit, tech firms’ imaginary of a technological future will be limited to the fast and cheap profits of monopoly rents. This is at the center of tech’s innovation crisis. That tech work is becoming bullshit is not the problem. The problem is that the companies themselves are bullshit.