Recap

This is Part 3 of my series on building effective software engineering teams. In this installment, I delve into a few more tangible aspects of enabling an effective engineering team. It’s remarkable how much of a difference effective tools make in improving outcomes for software engineering teams.

Overview

I define tooling as any element that helps streamline the process of delivering quality software. Like any production process, the goal is to maximize output for every unit of input. Anything that works counter to this objective should be considered wasteful. Similar to lean manufacturing, waste in software development is anything that reduces the quality and speed of our desired output. Since software is mostly intangible, one of the ways to measure waste would be in terms of lost productivity. Unfortunately, while the impact of waste is easy to observe, it is often difficult to quantify. In essence, tooling helps eliminate waste in the software development process.

Hardware

Through my journey as a software engineer, I’ve often encountered situations where engineers are assigned under-powered machines. This is usually due to a disconnect between administrative teams’ expectations and what engineers actually need. It doesn’t help that hardware costs have been rising, and many engineers prefer Apple hardware (because it just works out of the box). For e, the key propositin here is that organizations with software engineers should establish a reasonable baseline hardware specification for their engineering machines.

Integrated Development Environments (IDEs)

IDEs play a crucial role in maximizing developer productivity. Ideally, teams should standardize their development tools and develop a high degree of familiarity with their chosen IDEs. Engineers should also be well-versed in shortcuts and efficiency features within their environments.

I recall an experience where my machine would freeze every time I launched our platform application along with the accompanying database in our selected IDE. It would then take about 10 minutes to deploy and become accessible. Testing changes was a huge pain. Now, imagine an engineering organization with 50 engineers experiencing this problem several times a day. That’s easily over 8 hours of lost productivity daily. This is a clear example of engineering waste—difficult to track and quantify unless properly surfaced.

Eventually, we discovered that the issue was a bug related to the IDE calling home, and all we needed to do was update the system configuration. This reduced load times from 10 minutes to 2 minutes—a simple fix that significantly improved productivity.

Communication and Team Sizing

Communication within engineering teams is one of the most critical factors in determining team effectiveness. Nothing results in more waste than engineers working in the wrong direction due to a misunderstanding of business requirements.

One of the challenges of team sizing is that as team size increases, the number of communication channels grows exponentially. This is given by the formula:

Communication Channels = n(n - 1)/2

Where n is the number individuals on the team. This may be easier to grok when illustrated graphically:

This highlights the importance of keeping team sizes as small as possible to enable effective communication. Beyond the complexities of communication in large teams, stand-ups also take longer longer and become less effective, simply because people may tune out or may not even involved in some of the updates been given.

To add to this complexity, in remote work environments, most communication is asynchronous and happens over chat tools. This can become chaotic if multiple adjacent channels are created, making it difficult to track conversations and follow through with required actions.

To address this, teams should standardize on a single communication tool and incorporate intentional redundancy - repeating key information across multiple channels and validating that messages are received as intended.

Coding Standards and Documentation

Clear coding standards and up-to-date documentation are superpowers for any engineering team. They ensure consistency in code implementation, creating a shared vocabulary and conventions within the team. This also simplifies the on-boarding process for new engineers.

One key consideration is that documentation should be kept as close to the code as possible to prevent it from becoming stale as the code-base evolves.

DevOps and Deployment Infrastructure

For any engineering team, tooling includes the artifacts and pipelines necessary for deploying software solutions. The best tools should never get in the way but should make deployment as smooth as possible. However, faster isn’t always better. Some pipelines have quality/testing gates that prevent deployment if a certain threshold is not met, and that’s a good thing.

That said, we must ensure that the deployment process is optimized to eliminate inefficiencies that don’t serve anyone’s best interests. A common issue arises when DevOps and infrastructure teams are misaligned with engineering teams. Without aligned KPIs and goals, these teams may work in opposing directions, leading to an antagonistic relationship that benefits no one.

While this may seem to contradict the earlier point on team sizing, the best solution is to incorporate DevOps and infrastructure members into engineering teams, even if only in a fractional capacity. This ensures alignment among key stakeholders.

Generative AI

Steve Jobs once said that computers are the bicycles of the mind. Today, there are many discussions about Generative AI (GenAI) and whether it will eliminate the need for programmers. I don’t believe this will be the case. As with every technological revolution, the nature of work will evolve, inefficiencies will be removed, and new markets and opportunities will emerge.

Most people don’t realize that ‘computer’ used to be a profession, referring to people responsible for extensive calculations. The invention of computers made that job obsolete, but in doing so, it created the need for computer programming as a profession.

While I don’t know exactly what the future holds, it’s already clear that AI is a force multiplier for developer productivity and should be embraced by engineering teams. Every organization should invest in enabling their engineers with AI-powered tools to maximize the return on investment (ROI) from their engineering teams.

Summary

There’s a popular saying: “A workman is only as effective as his tools.” Another states: “If all you have is a hammer, everything looks like a nail.” Both highlight the importance of having the right tools and a broad skillset in any engineering endeavor.

This isn’t a comprehensive list, but the message is clear—tOnce a good foundation is set, the right tooling is guaranteed to bring out the best in software engineering teams.