|
| 1 | +[⬑ back to the overall graph](../use_gqm.md) |
| 2 | + |
| 3 | +# **Metric:** Contribution Distance |
| 4 | + |
| 5 | +**Synopsis**: Number of organizational hops from owning team to contributing team |
| 6 | +It can be useful to know not only when contributions to a repository are made, but the distance between the |
| 7 | +owning team and contributing individual in terms of organizational distance. |
| 8 | + |
| 9 | +**Unit of Measurement**: manager levels |
| 10 | +This is measured in terms of number of managers you have to move up in the hierarchy before you can go down to |
| 11 | +the contribution team. For example, if the manager in common of both parties is two jumps up the chain, |
| 12 | +then the distance is 2. Most contributions will be in-team and therefore are 0 distance. For a contribution |
| 13 | +to be "InnerSource", it has to be at least 1 distance. |
| 14 | + |
| 15 | +**Interpretation**: Contribution distance is a measurable proxy for collaboration |
| 16 | +A repository with a pull request that has a high contribution distance is likely to be reflective of a |
| 17 | +repository with more impact and more effective set up for collaboration. Repositories that have pull requests |
| 18 | +with a high contribution distance are on average probably more likely to be useful to others than repositories |
| 19 | +with only a single short distance collaboration. |
| 20 | + |
| 21 | +Collaboration distance and counts combined can also tell you something about the type of value a repository provides. |
| 22 | +A repository with many pull requests of 1-2 distance is likely to be reflective of partner teams working closely |
| 23 | +together to solve a shared problem. A repository with many pull requests of 3-4 distance but only a single |
| 24 | +pull request from each contributor is likely to be reflective of a repository that is being used for information |
| 25 | +sharing or as a reference. Contribution distance is a metric that can be combined with others to build up |
| 26 | +a picture of the community around each repository where InnerSource contribution behaviors are occurring. |
| 27 | + |
| 28 | +**Measuring** |
| 29 | + |
| 30 | +Measuring this metrics requires a mapping of your organization's hierarchy and a way to define whether the person |
| 31 | +that submits each issue or pull request is a member of the owning team or not. There are likely multiple |
| 32 | +ways to do each of these tasks and the best way will depend on your organization. |
| 33 | + |
| 34 | +As an example, Microsoft determines if an individual is on the "owning team" by finding the individual that approved |
| 35 | +the majority of pull requests and then finding the manager that is accountable for that individual. |
| 36 | + |
| 37 | +One possible pitfall is that manager and team relationships change over time. For this reason, it can be preferred |
| 38 | +to measure contribution distance soon after the pull request is merged and then store the distance as |
| 39 | +manager relationships will change over time degrading the data quality. |
0 commit comments