Nearly every developer (92%!) is either using or experimenting with AI tools for coding, highlighting just how much AI is changing the way we develop software. But even with rapid technological advancements in tools and foundational models, one thing hasn’t changed: the need to understand code. No matter if a developer or AI writes the code, someone else will need to understand and interpret that code.
What is code knowledge sharing?
Code knowledge sharing is the process of distributing engineering expertise and understanding of codebases within a group. This activity encompasses various methods such as documentation, engaging in pair programming, conducting code reviews, real-time questions on Slack/Teams, email threads, and the list goes on.
However, more often than not R&D organizations find these processes to be lacking – especially when it comes to documentation. Quantifying the impact of poor knowledge sharing used to be something we could only do anecdotally. Until today.
We surveyed over 200 US-based full-time software engineers, senior managers, and executives up to the VP level in engineering departments at companies with more than 75 developers.
Here are some of the highlights:
Code ships 50% faster when devs understand your codebase
73% of developers surveyed emphasize that code comprehension is essential for their productivity, noting that the exchange of detailed and comprehensive code insights can, on average, increase project delivery speed by 50% or more. This underlines the crucial importance of technical knowledge sharing in accelerating development, refining processes, and enhancing the developer experience.
Your knowledge sharing methods (probably) suck, says your team
Only 1% of developers view their company as outstanding in code knowledge sharing, while fewer than half (46%) are confident in their organization’s ability to share code knowledge effectively. This is a considerable gap in confidence, and highlights the urgent need for better code knowledge exchange, especially as an increasing amount of code is AI-generated.
What devs want
Developers highlighted key functionalities they seek in code knowledge sharing tools: the capability to automatically update sources of technical knowledge, such as internal documentation (45%), provide dashboards showcasing usage metrics (42%), and support the creation and easy discovery of codebase-specific technical knowledge directly within the IDE (41%).
Manager vs developer
An interesting trend presents itself throughout the entire report: the divergence in viewpoints between developers and managers. For example, 81% of developers surveyed acknowledge the importance of sharing knowledge when it comes to increasing productivity. A smaller proportion of senior staff and managers (65%) recognize its importance.
