Analysis / Research / /

TNS Research: How Many Company Developers Should Work on Open Source?

1 Feb 2016 9:25am, by

2356663850_3b2b473686_z

How many employees in your organization contribute to open source projects? Earlier this year, The New Stack asked this question to companies in the container ecosystem.

Among the 36 responses we received, the median response was ten employees, which is a lot, but even more significant if we look at the size of the companies involved. Taking this into account, we found that the median company actually said 47 percent of their employees were contributing to an open source project.

We found that the larger the company, the fewer employees are involved with open source software (OSS). In addition, we identified a relationship between a company’s maturity and the number of developers it employs. Overall, this research confirms that container companies are indeed are developer and OSS-centric.

36 Surveyed Companies Came in Many Different Sizes

Minimum Median Mean Maximum
Contributors to an Open Source Project 0 52 10 600
Total Employees 1 731 35 21,928

Almost half of all employees contributing to open source is a shocking figure so we decided to investigate further. We found that company size and the survey’s semantics impact the results.

The smaller companies tended to have a larger percentage of their employees contributing to open source projects. This makes intuitive sense because many of these companies’ founding employees contribute to the open source project for which their product is based on.

The companies we surveyed with fewer than 10 employees had on average 67 percent of the workforce contributing to an open source project. The larger companies (more than 100 employees), of which many have a broad range of non-software products, averaged 30 percent of employees contributing to open source projects. It is important to note that we did not weight the percentage figures based on the company’s number of employees. If we did that, the largest company’s thousands of employees would skew the full sample figure to just 7 percent of employees.

Survey data was combined with LinkedIn stats to get a different perspective on how many employees contribute to open source projects.

Survey data was combined with LinkedIn stats to get a perspective on how many employees contribute to open source projects.

There are varying viewpoints on what it means to be a contributor to an open source project. Even GitHub’s definition of a contributor differs based on what dashboard you use. In their most narrow definition, a contributor is someone who makes a commit to a project. Looking at individual responses, we believe not everyone used this definition. In fact, we found that two executives from midsize companies had given responses that were identical to the size of their entire company. Instead, these executives were probably thinking that since their core product is based on open source then everyone that works for them is supporting open source.

The New Stack evaluated the complete list of contributors of many of these company’s open source projects and can attest to a very high level of participation in open source projects. Thus, although we are skeptical about the accuracy of every response, we believe the survey accurately depicts the importance of OSS to these companies.

If your company says it is dedicated to open source, how many employees should be contributing to open source projects? And should those projects be directly related to your commercial activities? How you answer these questions may influence how “open source” friendly your company actually is.

Looking at Developers, Not Just OSS Developers

These types of questions make us think about how companies should be allocating their developer “human resources.” So, we looked at the number of developers and software engineers employed by various companies. Using these figures, we hope to help companies determine how many of its employees should be developers and inform how much of their time should be dedicated to open source projects.

We focused on 95 container-related companies that were covered in the first of a series of eBooks about Docker and the Container Ecosystem. Data about the number of employees were collected using LinkedIn by selecting “Current Company” via a checkbox and then searching for employees based on their current employer. The specific search criteria in the Title field was “’software engineer’ OR developer.” Median and average figures are first calculated per company. A full list of the companies investigated, along with non-survey related data, can be found here.

The broader group of 95 companies tended to be larger than those that completed the survey, with a median of 41 employees. However, even as the number of developers employed changes, the percentage of employees that are developers remained relatively constant.

Workforce Data on 95 Container-related Companies

Year Company Founded # of Software Engineers or Developers # of Employees
Median Mean Median Mean
Before 2010 11 3,383 798 35,480
Between 2010 – 2013 9 18 35 122
After 2013 2 2 7 9

So, we dug deeper into these companies and found more trends by looking at company maturity. It wasn’t surprising to find that younger companies have fewer employees. It is important to realize that in many startups the CEO or founders are also developers but do not identify themselves as such on LinkedIn. If we considered them as developers, many companies would likely have half of their employees involved with coding.

The middle-aged companies, those founded between 2010 and 2013 have the largest percentage of employees with “software engineer” or “developer” in their title. Many of these companies are focused on growing their development teams, to find companies with more than 30 percent of their employees being developers. Among this group are The New Stack sponsors Codenvy and SignalFx. The older companies have fewer developers overall, but this is partially due to IBM (eight percent of 418,168) and HPE (five percent of 219,111) bringing down the numbers.

Although the number of both employees and developers goes up over time, the percentage plateaus.

Although the number of both employees and developers goes up over time, the percentage plateaus.

Using LinkedIn to determine the number of developers can be problematic because it relies first on users self-identifying as actual employees of a company and then as a person with “developer” in their title. That being said, it is true that as companies grow, they by their very nature will increase their headcount.

After a few years of existence, many of the companies we reviewed are in a growth stage, which means they are focusing on hiring developers and perhaps building their intellectual property with VC funding. The older companies we reviewed are not by their very nature focused only on containers, but they are still tech companies with significant resources dedicated software development.

Codenvy, HPE, IBM, SignalFx  are sponsors of The New Stack.

Feature Image: 1917 organization chart of the Tabulating Machine Co., forerunner of IBM, via Flickr.


A digest of the week’s most important stories & analyses.

View / Add Comments