What is Coder Score and how is it calculated?

SeekOut analyzes code contributions from GitHub users to determine how well-accepted or influential their code is. Using that analysis, we assign each member a Coder Score of one to five stars based on their demonstrated contributions.

This feature is available for users with Professional and Enterprise licenses.

SeekOut analyzes code contributions from GitHub users to determine how well-accepted or influential their code is. Using that analysis, we assign each member a Coder Score of one to five stars based on their demonstrated contributions.

Use the Coder Score filter while searching in the GitHub talent pool to narrow down candidates by the quality and impact of their coding contributions.
  • Five stars corresponds to the top 2% of all GitHub members.

  • Four stars is for the top 10%.

  • Three stars is the top 30%.

  • Two stars is the top 60%.

  • The rest are given one star.

How does SeekOut calculate Coder Score?

SeekOut looks at all public code contributions on GitHub to assess a candidate’s SeekOut Coder Score. We compute a score by looking at where the code goes in order to measure its impact. SeekOut calculates this score per language as well, so candidates are rated on their expertise with Java vs. Python for instance.

The two biggest factors on determining a score are how many other people are working in a GitHub project and how often the code is copied.

Number of People Contributing to a GitHub Project

Some GitHub projects have only a single person working on them; other GitHub projects have thousands of contributors. The number of contributors is related to code quality because there is much more scrutiny of any new code in a project with more contributors.

  • If a candidate adds code to a project where they are the only contributor, they get very few points toward their Coder Score.

  • If they contribute to a project with 5000 contributors, they get a lot of points toward their Coder Score.

How Often Code is Copied

GitHub supports a culture of code copying called “forking.” Candidates get points toward their Coder Score based on how often their code is forked by others. Candidates who seldom have their code forked get very few points toward their Coder Score, whereas candidates who have their code forked often get a lot of points toward their Coder Score.

Additional Notes

We link to the candidate’s actual code in GitHub. You can click through to their code and see what the candidate wrote before starting outreach with them.

Click on a repository or contribution to view it on Github, or click View all repos in profile to see a list of all code contributions a candidate has made.

Because we measure demonstrated ability based on actual GitHub contributions, you can be confident a developer with a 3, 4, or 5 star rating is very strong. Developers with lower scores, however, may still be very good or even great, but they don't do a lot of Open Source development.