A Stanford study reveals 9.5% of engineers contribute almost nothing, costing tech $90B annually, with remote work fueling the rise of "ghost engineer...

Socket

A recent Stanford study has revealed that approximately 9.5% of engineers in tech companies, termed 'ghost engineers,' do minimal work while earning high salaries. This issue is more prevalent among remote workers. The economic impact is significant, with companies potentially saving billions by addressing this inefficiency. Moreover, ghost engineers pose security risks due to unreviewed code and outdated systems. The challenge highlights systemic inefficiencies in talent management and performance evaluation, raising questions about the effectiveness of remote work policies.

Tech's $90B Ghost Engineer Problem: Stanford Study Finds 9.5...

Economic Impact of “Ghost Engineers” #

Security Implications of Employing Ghost Engineers #

The Twitter post referenced in this post exposed that they have very little data points to differentiate who are slacking and who are higher performant and who have more business responsibilities. They also used commit frequency as a metric, which everyone knows is not a metric for how productive or good someone is, as frequency does not mean quality. They also used AI to score contents of commits apparently, how much you can trust the AI output is up to debate.
For example, support handling, reviewing other people’s code, pair programming, talk to stake holders, break down tasks, holding other engineers hands. Individual commit value, more commits does not mean more work done.
If you dive into comments, there are plenty of people mentioning about lack of datapoints and consideration of different responsibilities.
Premature conclusion, the research needs more work. Honestly expect better from Stanford research.

This article is a bunch of crock that is being intellectually dishonest on both the metrics and their reasoning in order to make it look like the fault likes with the employees rather than the truth that is poor/bloated middle management that is under utilizing talent or higher ups that intentionally put their best Dev talent doing menial jobs rather than sitting down with them and discuss how to better utilize resources. Software se is about results and compensating by the hour just leads to workers being punished for being good and, encourages the very behavior they want to put on blast.
The fact they pay 200k+ for 5 hours daily is a show of how bad the management is, not the quality of the work that gets spne

As ever, when you read the words in a headline “study finds” or “Experts say”… you can almost definitively disregard the article as propaganda.

Readers should critically evaluate whether such narratives are backed by robust, unbiased data or selectively framed to serve specific interests.
In conclusion, the article is likely to find favor among commercial office property stakeholders due to its alignment with their goals to revitalize and maintain demand for office spaces.