According to news on December 9, a new study from Stanford University found that many companies have so-called "ghost engineers": these programmers receive high salaries but produce almost no code.The following is the translation:
Yegor Denisov-Blanch, a researcher at the Stanford Graduate School of Business, is not a psychic, but in recent weeks he has unexpectedly established contact with many "ghosts", many of whom are extremely angry.
After the 32-year-old graduate student posted an article online sharing his analysis of programmer productivity data from hundreds of companies around the world, his inbox was immediately flooded with messages.
Denisov Branch wrote on social media He calls these people "ghost engineers."
Denisov-Blanche's post quickly received nearly 4 million views, and some people calling themselves "ghost engineers" even contacted him. Judging from the emails, some people tried to defend themselves and some were angry and abusive, but these employees generally admitted that they took advantage of loopholes in the corporate culture and insisted that the problem did not lie with them.
Although Denisov-Blanche's research has not yet been peer-reviewed, his conclusions reignite the long-standing debate on the topic of "ghost engineers."
“Many people think this is an exaggeration, butMany software engineers I know only change code twice a month, send almost no emails, and have very few meetings; they work remotely for less than 5 hours a week, but their annual salary is as high as 200,000 to 300,000 US dollars.," investor Deedy Das wrote on X last month.
Das did not respond to a request for an interview. He named 13 companies, including Cisco and cloud computing giant Salesforce, and said this behavior was very common in these companies.He also detailed common tactics used by ghost engineers, such as frequently setting online status to "in a meeting" or using cheap "mouse emulators" to disguise ongoing activity.
Aaron Levie, CEO of the file storage company Box, noticed that Das' post mentioned his company. That night, he responded on X: "Today's discussion was particularly constructive."
Levy said his company is already working on solving similar problems. In a phone interview, he noted that while no one was immediately fired as a result, the online debate prompted a reexamination within the company.
Over the past four years, as remote work has become more popular in the technology industry, Box has begun to focus on measuring the productivity of all employees, including engineers. The company has reduced the size of its teams, avoided overlapping responsibilities, reduced the number of meetings and become more prudent in its selection of research and development projects.
Denisov-Blanche began teaching himself programming as a teenager. He mentioned in a phone interview that he did not initially intend to study the "ghost engineer" phenomenon.
He said he developed a machine learning algorithm in collaboration with Michael Kosinski, associate professor of organizational psychology at Stanford University, and Simon Obstbaum, former chief technology officer of Crunchyroll. The algorithm measures programmer productivity by analyzing corporate code bases, and it was in the process that they discovered the phenomenon of "ghost engineers."
Research data shows, Large companies are more likely to breed "ghost engineers", but small companies cannot completely avoid this problem.
Kosinski noted in a phone interview that due to the complexity of the work of programmers at large technology companies and the cumbersome internal structures of the companies,Evaluating engineer productivity is a challenge. However, he emphasized thatIf inefficient employees are not identified and eliminated in a timely manner, it will not only encourage the phenomenon of "filling the gap" but also undermine the rewards that efficient employees deserve.
While Stanford University was conducting this research,Some large technology companies are gradually rolling back their previously implemented remote working policies.After companies such as Google, Amazon, Meta and Microsoft successively laid off large-scale layoffs, the remote working model has also been re-examined. Amazon, for example, plans to require employees to work in the office at least five days a week starting next January. Companies such as SAP, AT&T, Dell and Zoom are also gradually tightening flexible office policies.
Denisov-Blanche pointed out that research from Stanford University found thatA higher proportion of top programmers work remotely, but "ghost engineers" are also more likely to choose to work remotely.Ghost engineers account for up to 14% of engineers who work remotely, compared with 9% of engineers who work in the office at least part of the week, and only 6% of engineers who work in the office every day.
Denisov-Blanche believes that,Rather than saying that the "Ghost Engineer" is deliberately slacking off, it is better to say that it stems from frustration.“The problem is almost always that they are disillusioned with their work and cannot see a clear connection between effort and reward or recognition,” he concluded after in-depth conversations with dozens of “ghost engineers.” “Over time, they lose motivation and their performance continues to decline.”
Denisov-Blanche said that over time, the behavior of "forging" changed from passive to active. Employees may adopt strategies such asMake up dates on your calendar or exaggerate your workload."Sometimes it's difficult for managers to discern the truth," he added.
Krunal Patel recalled that when he first entered the industry, he tried "ghost work" not because of laziness, but to give his manager a heads-up. For a week, he and a colleague detailed their work tasks to their manager twice a day, but nothing actually got done. It wasn't until they confessed that the manager realized they had been slacking off.
“He was in shock,” Patel recalled in a phone interview. “We sat him down and explained our frustrations.” Patel and colleagues expressed a desire to be less over-managed on day-to-day tasks and more involved in the real problems that needed to be solved.
Patel's managers adopted their suggestions, and the engineers became more efficient. Today, Patel is a technology executive and entrepreneur with 20 years of experience in the software industry. “We are working more efficiently and enjoying the process more,” he said.
Sudheer Bandaru encountered a similar situation when he was managing a software engineering team at a mid-sized company.
Bandaru mentioned in a phone interview that during the annual performance review, an engineer who was considered the "smartest" talked a lot in team meetings, but actually produced almost no code. "It shocked me," he said.
After communicating with this employee, Vandalieu found that he did not intend to slack off, but that the position was not suitable for him. “He was more of a researcher and didn’t like sitting in one place writing code at all,” Bandaru explained. After adjusting his position, this employee's performance became very good.
Bandaru mentioned that similar situations are not uncommon, which prompted him to create the Hivel platform, hoping to help companies develop software and perform analysis more efficiently.
If used correctly, software for monitoring technician productivity has the potential to create more gratifying management moments. However, tech industry writer and former software engineer Patrick McKenzie warns,Evaluating an engineer's work solely based on their lines of code can easily lead to misjudgment.
McKenzie said it's "understandable" that some senior engineers don't write code at all, for example because they might be designing software architecture or mentoring new employees. "This kind of person is neither a 'ghost' nor a perfunctory person," he added.
To avoid this misjudgment, Denisov-Blanche said he and his research collaborators developed a productivity measurement algorithm that not only monitors the work of teams and individuals, but also assesses their impact on the company's overall code base.
As his findings attract the attention of investors, Denisov-Blanche is considering how to commercialize them. This may help companies better deal with the problem of "ghost engineers."
However, Denisov-Blanche said his goal was to help the "ghost engineers" rather than expose them. "My job is not to track down these people," he said. “My mission is to understand the causes of this phenomenon and work to eliminate it.”