☀ Key Stats

◆ U.S. high-tech firms adopting open-source product development through GitHub recorded an estimated 4%–5% increase in gross margin relative to non-adopting firms.

◆ The study covers 977 U.S. high-tech companies from 2001 through 2025, including 231 open-source adopters and 746 non-adopters.

◆ The researchers used as many as 11,854 firm-year observations, depending on the statistical model.

◆ Their GitHub analysis included 323,634 project-level events across 44,147 repositories maintained by the adopting firms.

◆ The profitability pathway through higher labor productivity became statistically discernible when outside volunteers supplied approximately 35% of observed development activity.

◆ External volunteers provided an average of 51% of contributions among the adopting firms. The median was 54%, while the middle half of observations ranged from 41% to 64%.

◆ A 10% increase in the number of active open-source projects was associated with an estimated 0.1% increase in gross margin.

◆ Research and development intensity was present in all five combinations of company resources associated with high profitability.

◆ R&D intensity was absent in two of the four combinations associated with low profitability.

Note: The figures are statistical estimates for U.S.-based high-tech firms rather than guaranteed returns from opening software projects. The study examines gross margin, not net income, and its approximately 35% volunteer threshold is specific to the sample.

Continue reading ↓

U.S. technology firms adopting open-source product development recorded an average 4%–5% increase in gross margin, according to a new study.

The researchers compared companies that opened product-development work to outside contributors through GitHub with firms that did not adopt the approach.

The analysis covers 977 U.S. high-tech firms between 2001 and 2025.

Of those companies, 231 adopted open-source product development during the study period and 746 remained in the comparison group.

Gross margin measures how much revenue remains after the direct costs of producing and delivering a company’s products are deducted.

It is different from net profit because it does not include every operating expense, interest payment or tax.

External contributors changed the payoff

Opening a project on GitHub was not enough by itself to produce the labor-productivity pathway identified by the researchers.

The indirect effect on profitability became statistically discernible when external volunteers supplied approximately 35% of observed development activity.

The authors describe that figure as a sample-specific threshold rather than a universal target for software companies.

In a separate analysis limited to the period before Microsoft acquired GitHub, the estimated threshold was above 50%, demonstrating that the cutoff changed with the sample and period examined.

External participation was already substantial among the firms studied.

Volunteers accounted for an average of 51% of observed contributions, with a median share of 54%.

The middle half of firm-year observations had volunteer contribution shares between 41% and 64%.

The researchers argue that outside developers can bring knowledge from different technology systems, customer environments and software uses that a company’s employees may not encounter internally.

Their analysis suggests that integrating this knowledge can raise the amount of revenue produced per employee, which then contributes to higher gross margins.

Labor productivity explained only part of the estimated profitability effect.

The paper says other possible channels could include product innovation, community governance, licensing decisions and the signaling value associated with maintaining active open-source projects.

More active projects were linked to higher margins

The researchers also examined whether the intensity of open-source development mattered.

A 10% increase in the number of active open-source projects was associated with an estimated 0.1% increase in gross margin.

An active project was defined as one recording software-development activity during the year, including commits, pull requests, code reviews or resolved issues.

The result suggests that the estimated payoff increased with deeper involvement rather than coming only from a company’s initial decision to use GitHub.

The relationship remained positive across several statistical approaches, including comparisons around different adoption dates, synthetic control models and analyses designed to account for persistent differences between firms.

The baseline model estimated an average effect of approximately 4.4%.

Alternative models also produced positive estimates, although the size varied depending on the sample and method.

R&D remained central to open-source profitability

Companies still needed their own technical capacity to make use of outside contributions.

R&D intensity appeared in all five high-profitability configurations identified by the researchers.

It was absent in two of the four low-profitability configurations.

The result suggests that open-source development complemented internal research rather than replacing it.

Companies need employees who can evaluate outside code, maintain a coherent product design and convert community contributions into commercially useful features.

High external participation was present in two of the five high-profitability configurations.

In the other three, companies combined open-source development and R&D with a stronger internal workforce, suggesting that firms could reach similar outcomes through different mixes of internal and external resources.

Capital intensity, by contrast, was absent from four of the five high-profitability configurations and present in all four low-profitability combinations.

Industry concentration was also absent from every high-profitability configuration and present in every low-profitability one.

Those patterns do not establish that capital investment or industry concentration reduces profitability.

They show how those characteristics appeared alongside other factors in the companies included in the study.

The analysis covers more than 44,000 repositories

The researchers combined company financial information from Compustat and annual filings with activity collected from GitHub.

The final dataset contained up to 11,854 annual company observations, depending on the model.

The GitHub portion covered 323,634 events across 44,147 repositories belonging to the adopting companies.

Those events included code pushes, pull requests, issues, forks and review comments.

Contributors were classified as internal employees or external volunteers using corporate email domains and behavioral signals, including when activity occurred and whether an account participated in project governance.

The researchers then calculated the share of activity attributed to outside contributors for each company and year.

✦ Why it matters ✦

Open-source development is often discussed as a way to distribute software freely or reduce the cost of writing code.

The study suggests its commercial value may depend more on whether companies can integrate outside knowledge into their internal development process.

That distinction matters because simply making code public may not attract a community that contributes meaningful expertise.

Companies may also face costs from reviewing outside work, resolving disagreements, securing projects and maintaining a consistent product architecture.

The findings indicate that internal R&D remains important even when a large share of development activity comes from volunteers.

Open-source communities therefore appear to function as an additional source of specialized knowledge rather than a direct substitute for paid employees.

For software workers and contributors, the results also highlight the economic value companies may obtain from community participation.

The study does not calculate how those gains are divided among firms, employees and outside developers.

ⓘ How to read the findings

The study focuses only on U.S.-based high-tech firms with sufficient financial information for the analysis.

Its findings may not apply to private companies, firms in other countries or businesses outside technology-intensive industries.

The analysis covers projects initiated and hosted by the companies studied.

It does not fully capture firms contributing to projects controlled by other organizations or incorporating externally developed components into proprietary products.

The employee and volunteer categories were estimated through a multi-stage classification method.

Accounts with missing, unusual or ambiguous identity information could have been placed in the wrong group.

The study’s main outcome is a transformed measure of gross margin.

The estimated 4%–5% gain should not be interpreted as an equal rise in net profit, revenue, cash flow or stock-market value.

The paper uses several methods intended to separate the effect of open-source adoption from differences that already existed between companies.

Observational company data cannot reproduce the level of control available in a randomized experiment, and the results remain dependent on the study’s definitions, comparison groups and statistical assumptions.

The approximately 35% volunteer threshold is also not a management rule that should be applied to every project.

The authors describe it as the point at which the labor-productivity pathway became statistically discernible in their full sample.

The August 2026 manuscript identifies itself as forthcoming in Information Systems Research.