The turnover rate of jobs in the tech sector is notoriously high.
The average job turnover rate across all industries is estimated at 10.9%. But according to an 2017 analysis of LinkedIn user data, the average job turnover rate for tech is 13.2%, a full 2 points higher than the national average. Within the technology sector, the occupations with the highest turnover rates were:
- UX/Designers: 23.3%
- Data Analyst: 21.7%
- Embedded Software Engineer 21.7%
We have found that recruitability can be modeled, with a level of reliability and accuracy, based on an amalgamation of several factors, including the past tenure trend of the individual, the tenure trends of the cohort (e.g. Programmers, UX/Designer, Data Scientist, etc.), and tenure trends at the current company.
That is, for example, if a UX worker has been at their job at Microsoft for 2.5 years, and they’ve never stayed at any job for longer than 3 years, and the average UX job at Microsoft is 2 years, then that worker is probably highly recruitable.
Other models for turnover also exist. But gone are the days when we just chalk it up to a simple lack of loyalty. There are so many opportunities for tech folks, and the tech market is demanding talent.
But the talent demands that tech companies rise to meet them.
In this blog post, we will examine some of the tech-job tenure dynamics and phenomenology that we see at various tech companies and in various tech roles.
Do Perks Really Work?
Tech companies pay a heavy price for turnover. Each departing employee costs employers an average of $144,000 in lost productivity, salary, and recruiting costs. And there are potential hidden costs of loss of morale, reputation, and momentum, within the company.
The total cost of turnover for the tech industry as a whole adds up to nearly $50 billion.
Many tech companies invest time and money to make the work environment as attractive as possible for their employees. Common perks at top tech firms include everything from flexible time-off, to free on-site gyms, to daily catered meals. These perks are used to attract new hires, but it’s unclear if they have an appreciable impact on average retention of tech employees.
But with everyone doing a razzle-dazzle show to boost morale, the novelty wears off quickly, and it has become an expected benefit.
The differentiator will probably always remain the underlying belief in the people and the product.
We used the humanpredictions database to measure the average tenure for humans in common tech roles at selected well-known tech companies to see if we could spot any trends.
Google, for instance, is synonymous with beautiful state-of-the-art office spaces, games, and recreational activities for its developers. Google headquarters offers a enormous array of food for its employees, all available for free, 24 hours a day. The options range from high-end gourmet meals to trail mix, fruits, and nuts.
Google seems to invest a lot in employee attraction and retention — gaming them and feeding them — and yet the average tenure at Google in 2017 was 1.1 years.
Why would you leave all that cheddar, so to speak?
At the other end of the spectrum, there are major tech firms that seem to be doing a much better job at retaining their employees. Employees at Apple stay an average of 5 years, while at Cisco, they stay for a whopping 7.8 years.
It is very tempting for the human brain to begin pattern recognition procedures — even in the absence of data — in order to make conclusions about the causes for the differences in the average tenures of tech workers at Google, Apple, and Cisco.
So let me add some points to the pattern so that your brain can start to connect some dots.
As promised, we selected common tech roles at well-known companies and measured the average tenures in months.
Average tenures vary across companies, as expected, but they also vary across role across company. For example, in our data sample, tech folks in User Experience (UX) can expect to build the biggest dust bunnies in their cubicles at Sun Microsystems, but won’t even last a year, on average, at Twitter.
If you call yourself a “Developer” while working at Facebook, you’ll stick around about 10 months longer (on average) than those uptight “Software Engineers” that you see at the Facebook cafeteria. Those nerds!
But that naming convention — “Developer” beats “Software Engineer” — doesn’t seem to work that way at Google, IBM, Intel, Cisco, Sun, Salesforce, Twitter, or even Yahoo! (go figure!).
There appears to be a tech-role-at-company dependence on job tenure (which we have modeled at humanpredictions).
Time’s Up, Tech! Turnover and Fairness
For a second there, it seemed as if the high rate of turnover might have signified a tech doomsday, and that tech was going to see an exodus, according to some commentators. But the overall rate of turnover has been decreasing since 2015:
But according to one a report from the Kapor Center, a significant factor in early turnover is still lurking in the background, underappreciated by employers: unfairness in the work environment.
There are three factors most associated with turnover:
- Lack of career advancement opportunities
- Heavy-handed recruitment by competitors
- Perceived unfairness within company culture
But, in the Kapor Tech Leavers Study, “unfairness” was cited more than anything else by employees as a reason for leaving an organization.
Very recently, Google employees walked out of work in November 2018 to tell management to change their ways about sexual harassment. They walked out of work across the world; they walked out in London, Zurich, Singapore, and New York. Whether or not they walk out permanently remains to be seen — how will Google address unfairness and sexual harassment?
According to the Kapor Study, “unfairness” arose in four categories:
- unfair performance management
- sexual harassment
In this study ‘unfairness’ was cited more than anything else by employees as a reason for leaving an organization. This came in the form of four categories: unfair performance management, stereotyping, sexual harassment, and bullying/hostility.
Also, “unfairness” could be experienced indirectly. Observing unfairness in the workplace increased the likelihood of job turnover. Upon exit interview, 37% of leavers cited the observation of unfairness towards others as a factor in their decision to leave. In comparison, 22% of leavers said they were recruited.
The largest group reporting unfairness as the cause of turnover were underrepresented men of color (40%), followed very closely by White and Asian men (39%). Women also endorse unfair treatment much more often than men.
The Kapor data suggests that unfairness is exacerbated within the tech sector. Technical employees were significantly more likely to cite unfairness as a factor for their departure than non-technical employees (42% versus 32%). Tech workers have so many options for work. And culturally speaking, the overwhelming majority of tech workers have post-modern pluralistic values, which means they are frequently taking swift action for social justice and walking out of unfair situations.
It’s estimated that turnover caused by unfairness alone costs the tech industry $16 billion per year to replace employees. Beyond the direct costs, 35% of respondents reported that experiences of unfairness would make them less likely to refer others to work for that company, which indirectly increases the cost of recruiting. And 25% of respondents reported being less likely to recommend the products/services of that employer, which indirectly impacts the bottom line of the company.
Tech companies have missed their opportunity: 62% of the leaving tech employees said they would have remained with the employer if visible steps were taken by the company to improve fairness with a more positive and respectful culture, and 57% said the same if the company took steps to increase inclusivity. And companies with as few as 5 diversity and inclusion initiatives were significantly less likely to be perceived as unfair by their employees.
There is a debate over the causes of the lack of diversity and inclusion within tech.
Some think that tech culture itself has unfairness baked right into the fabric of its very existence. And in the presence of unfairness, the underrepresented and their friends flee.
Some think the pipeline is unfair. From the very beginning, the populations that are underrepresented in tech are less likely to pursue the type of education and types of experiences that are required by tech companies.
When it comes to inclusion, some hard metrics are available. Women comprise 25% of tech employees. Black and Latinx comprise 30% of the U.S. population but comprise 15% of tech employes. And at top companies — Apple, Google, Facebook — black and Latinx employees comprise only around 3-5% of employees.
Like a fish needs a bicycle
We analyzed gender diversity using the humanpredictions database. We looked for correlations between gender diversity and tenures at tech companies. In order to expose any potential correlations, we looked at tech companies that had at least 100 employees. Then we created a scatter plot of the ratio of the number males-to-females versus the average tech-job tenures at those companies.
The scatter plot shows a clustering of average tenures towards the middle of the time scale with no apparent differentiation by the ratio of males-to-females. The scatter plot also shows a clustering of average male-to-female ratios around 10-to-1.
The plot also indicates that there are a cluster of big tech companies (100+ employees) out there where there are 25 men for every woman.
And at these big tech companies with 25-to-1 ratios, these tech folks are keeping their tech jobs for 2 and 3 years.
And every day at that tech company, for 3 exemplary years, there are at least 25 men doing their tech jobs that may or may not notice that there is 1 woman doing a tech job. And, at that same tech company, there is at least 1 woman doing a tech job that may or may not notice that there are at least 25 men doing their tech jobs.
That seems like a very strange company, with fish on bicycles. But, hey! There are even stranger companies.
Those “strange” companies exist. And tech workers have choices. If tech workers get bullied, or their co-workers are harassed, the righteously indignant declare: “Time’s up.”
If you’re interested in strategies to reduce unfairness, and in reducing turnover related to unfairness, read the Kapor Study.
In general, the evidence suggests that comprehensive diversity and inclusion initiatives should be developed. According to the report, it is absolutely critical that the initiatives be endorsed by the CEO and the Executive team. The effectiveness of initiatives should be measured and reviewed regularly.
Other interventions can also ease the tech turnover rate. Obvious interventions such as greater compensation (73%), promotions & increased responsibility (67%), and improved managerial efficacy and leadership (69%) were cited in the Kapor Study
It was the best of times, it was the worst of times
So with these many caveats and nuances kept in mind, we can take a look a selection of the best and worst companies in terms of average job tenure of technical employees. We selected tech companies with at least one-hundred current employees from the humanpredictions database.
The top 50 employee-retaining tech companies are:
The bottom 50 employee-retaining tech companies are:
The Matrix Has You
Tech-job turnover is a giant, multi-dimensional matrix of humanity at the crossroads of culture, technology, and the need to feed ones face.
It even exposes some of the nastier side of our humanity. Even in culturally forward, post-modern companies such as Google, we still find unfairness in forms such as sexual harassment. There are tech companies with 25x, 50x, 100x as many men as there are women.
Maybe the tech sector is a little too close to the world described by the character, Morpheus, in the movie “The Matrix”:
“You have to understand, most of these people are not ready to be unplugged. And many of them are so inured, so hopelessly dependent on the system, that they will fight to protect it.”
And yet here I sit on my computer. Typing away. A data scientist. Male. Working at a tech company with a 9:4 male-to-female ratio. It’s a not 1:1 ratio at humanpredictions.
Fairness and Foresight
Unfairness is highly correlated with job turnover, especially in the tech sector.
And job tenure is also correlated with job title. And the correlation of tenure to job title varies company to company.
So, it’s complicated. Matrix-level complicated.
And it’s expensive to lose employees. Like $50 billion expensive.
But with strong foresight, deep analysis, and profound humanity, we can create a long-term stable pattern of fair ebbing and flowing of humans though the tech industry.