What to Do When Faced With Knowledge Silos On Your Git Team

While knowledge silos may exist between departments, they become an issue when they begin to crop up amongst team members. Having knowledge silos on your team means information is not passing freely among team members, which could be a sign of trouble.

As a manager, you can identify knowledge silos pretty quickly by looking at the code reviews. A dead giveaway, for example, is when a group of engineers begin only to review each other’s work. For example, if you have two or three individuals who are constantly reviewing each other’s pull requests and no one else’s, you might have a knowledge silo. While they may be learning from each other’s work and techniques, they may be focusing on specific areas of code without letting in other members of the team.

Why Knowledge Silos Happen

There’s a variety of causes for engineers to begin to do their own thing and inadvertently create a knowledge silo. They might have just gotten into the habit of reviewing each other’s work and enjoying the team within a team dynamic they have made. They might enjoy working together as they have similar cadence or pacing that isn’t matched by other team members and fall into the belief that if they handle a section together, they can ultimately get more done quicker.

What Happens When Knowledge Cycles Are Allowed to Grow

Once these patterns have existed for an extended period of time, you might notice that these team members fall into the habit of rubber-stamping pull requests. Left long enough, and those in a knowledge silo will likely spend less and less time doing substantial reviews since they have so much trust in the work output that they’ve so far been able to deliver. However, when work is starting to get pushed out without adequate evaluation, mistakes can be made, and code can suffer.

Worse still, if specific individuals are allowed to take over complete sections of code, and no knowledge transfer is happening, if the knowledge silo suddenly goes away (for example, if the huddle of workers leaves the company), suddenly there’s a gaping hole in knowledge that suddenly has to be filled. That’s why it’s crucial to find knowledge silos early and deal with them quickly.

Leverage Git, GitLab, Beanstalk and all Git services right inside Jira

Check out One of Bitband’s Jira Integration apps, Including Jira integration for GITHUB, BITBUCKET, or GITLAB available in the Atlassian marketplace.

Recognizing the Signs

If you happen to have your team working in the office, it might be easy to spot where silos can crop up simply by watching social bonds at work. However, many teams are now much more remote, so, visually, it may be harder to spot. You can also use Knowledge Sharing reports to get a sense of how knowledge is shared across your team. Spotting knowledge silos becomes easier to see how information is being shared and by whom.

If you note that two or three people continuously review each other’s code, the Knowledge Sharing Index will clarify that (by showing as 0). In contrast, if most of your team shares knowledge, the index will trend closer to 1. Check multiple sprints. If a pattern emerges, you’ve found your silo.

Dealing with a Knowledge Silo Without Damaging Social Dynamics

Breaking up a knowledge silo will require bringing in outsiders. The easiest way to do that is to look for individual engineers that may be outliers and bring them into the review process. Look for people who could be cross-trained for the specific area of code to broaden the base of knowledge. Begin to assign other engineers to review the knowledge silo’s pull requests to remove the temptation for them to return to reviewing their own pull requests amongst themselves. Further pull them out of their silo pattern by having them review the work of others in other areas of code as well.

The idea is to break up the pattern of having a few individuals take over an area of code and not to destroy social dynamics. Therefore, while it’s healthy to have knowledge silos effectively broken up, there’s no need to, for example, shake up seating assignments or social bonds, as they are healthy in a team – and expected.

Conclusion

Letting teams form their own social dynamics, but don’t hesitate to mix things up if those social dynamics start causing knowledge silos. Make sure you’re spreading team members around the code base and helping each other collaborate and learn together – without forming closed-off sub-groups.

At Bitband, we offer a variety of Jira integration apps, available now on the Atlassian Marketplace. Let us help your team integrate Jira with apps like BitbucketGitHubGitlab (and more). Learn more here and try us out today!


Want more Bitband insights? Check out:

    Contact Us