Pairing is something that has a long history dated back from XP programming. We embraced pairing during the digital transformation. The biggest advantage the teams experienced was that the culture of the team naturally changed when people rotates to work on different things together.
When people work by themselves, silos are created within the team. Even when people occasionally get together to pair or mob, people tend to pick the persons they want to work with. After a while, you may see segregations within a team. Spread of knowledge is not even but linear.
We talk about encouraging people to move around to work on different area of the application or system, it is actually not that easy for anyone to start with a new area. An half hour knowledge transfer may not work well depends on how well that is done. Very often, the person working on the new area may have question or being blocked by something and needs the one with knowledge to help. It would be an interruption to the other person’s work. Or the other person could be busy and won’t be available. Then it will end up pretty frustrating for the one working on the new area. People may later just avoid getting into new area due to previous bad experiences.
When those 2 people pair then the above problems will go away. There is great benefit that multiple people in the team have knowledge or experience on different area of application. When there are some issues, we can easily find someone to look into it, instead of having to rely on that one single person as how teams used to operate. People can go on vacation without being called because they are not the only ones that know a particular area.
With people moving around or leaving the companies, team will not be left at loss due to the knowledge held by a single person.