The power of the right culture and leadership approach

What cultural aspects are key?

• Empowerment
• Engaged
• Transparent
• Respectful

Key Takeaway: Empowerment, engagement, transparency, and respect are the four cultural aspects an engineering team should build around.

What behavioral aspects are key?

• Blameless problem solving
• Takes ownership
• Open to feedback
• Progresses the team, not just themselves
• No ego
• Celebrates success

Culture shapes work enjoyment, work relationships, and work processes.

To ensure that good behaviours are the pillars of the building, it is key that those who are not adhering to the good behaviours are told of their impact in a constructive, not destructive way.

AI code assistants have changed what ownership means inside an engineering team. A blameless culture still holds, but the review step now carries more weight, because the engineer who merges a change owns it whether a person or a tool drafted it. Leaders should state this plainly. Teams that leave it unsaid drift towards approving code that nobody has read, and the blameless habit turns into an excuse rather than a way to learn.

Key Takeaway: Blameless problem solving, ownership, openness to feedback, and no ego are the behaviors that hold an engineering culture together, and the engineer who merges a change owns that change whether a person or an AI coding assistant wrote the code.

How do you measure whether your culture is working?

Measure culture with a short anonymous survey, not an impression. DORA scores organizational culture with six statements written by the sociologist Ron Westrum, each rated on a seven point Likert scale and averaged into a single number. The six ask whether information is actively sought, whether messengers are punished when they deliver bad news, whether responsibilities are shared, whether cross-functional collaboration is encouraged and rewarded, whether failures are treated as opportunities to improve the system, and whether new ideas are welcomed. DORA states that organizational culture is a perceptual measure and is therefore best measured using survey methods, and that the Westrum measures are valid and reliable statistically.

The score places a team in one of Westrum's three types. A pathological culture is power oriented and punishes messengers. A bureaucratic culture is rule oriented and neglects them. A generative culture is performance oriented and trains them. DORA links a high trust, generative culture to software delivery and organizational performance, so the number is not a vanity metric. Run the six statements every quarter, keep the responses anonymous, and watch the trend after a reorganisation or a bad incident. A falling average tells you more than an all hands meeting does.

AI makes this measurement more useful, not less. The 2025 DORA report describes AI as an amplifier that magnifies what a team already is rather than fixing it, and reports that 90% of respondents use AI at work while 30% have little or no trust in the code it generates. That gap sits inside the culture questions above, not inside the tool. A team that punishes messengers will not report a doubtful AI generated change until it reaches production.

Key Takeaway: A short anonymous survey built on the six Westrum statements used by DORA, rated on a seven point scale and repeated each quarter, measures engineering culture far more reliably than an impression does.

What can help you improve your culture?

Communities of Practice (CoP) are sessions of interest, where the aim is learning an area of mutual interest, which can span multiple team disciplinaries. What is key is that membership is not voluntary.

Where there is learning, there is always an improvement in team culture. Even learning new skills together can build a bond between the team. I have seen communities around 'what is automation testing' and it has helped to get engineering, product, quality assurance to understand automated testing better and this meant that everyone in the team could support and understand this interesting area.

Should you give CoP's a try? Yes! They are easy to set up and will give maximum benefit to your culture.

I personally have learned lots from communities of practice, for instance a talented QA engineer taught the team how to have fun with some HTML, CSS and JavaScript, I even did a community of practice in my first role after University on 'How to juggle?', though I have forgotten now. Communities are all about learning and the more learning the stronger the culture.

Distributed teams need the culture written down, not only felt. Write a short note that records the decision from each design discussion, record demos so people in other time zones can watch them, and agree a small block of overlap hours where synchronous work happens. Culture that lives only in a meeting room disappears the moment half the team joins from somewhere else.

Another aspect of a high-performing team is the leadership approach. I have been very careful to say 'leadership' rather than management. Why? As leaders are those members who we remember when reflecting on our careers, ultimately leaders exist in an awesome culture.

In the past, I have seen many managers who act like army generals and sit on a comfy chair, and exercise their power to get something done. How was the culture in those teams? Pretty awful, and for me, that is why most members were low-performing.

Key Takeaway: Communities of Practice improve engineering culture because shared learning across disciplines builds bonds, and distributed teams also need decisions written down, demos recorded, and an agreed block of overlap hours.

How should leaders handle AI inside an engineering culture?

Treat AI as an amplifier of the culture you already have, and give it somewhere solid to land. DORA guidance on platform engineering states that when platform quality is high, the effect of AI adoption on organizational performance becomes strong and positive, and that when platform quality is low, the effect is negligible. The same tool, dropped into two different teams, produces two different outcomes.

That guidance also names the trap a leader has to watch for. Individual productivity gains are frequently lost to what DORA calls downstream disorder, where faster coding is swallowed by bottlenecks in testing, security review, and deployment. A team can feel much quicker while nothing reaches customers any sooner. DORA also warns of a J-curve, where a platform shows early gains and then dips as complexity rises, which is the exact moment a nervous leader cancels the work.

The cultural work is the same work as before. Agree out loud who reviews AI generated changes and what that review has to check. Keep the six Westrum statements running, because they are what tells you whether people feel safe reporting a change they doubt. Ask for a balanced scorecard rather than a single speed number, since DORA notes that platforms can improve productivity while decreasing the throughput and change stability tracked by the DORA metrics if they are not carefully managed. A leader who adopts AI without naming the review rule and the measure is not leading the change, only permitting it.

Key Takeaway: AI amplifies an existing engineering culture rather than repairing it, so leaders should pair adoption with a stated review rule, a quarterly Westrum survey, and a balanced scorecard instead of a single speed number.

What is a leader vs a manager

As per the above, in the great cultural environments that I have worked in, I have always seen leaders rather than engineering managers. I know who I would rather work for! Who would you rather work for?

Cultural improvement takes time, though we can all work hard to make it a better place to work. I'm an avid grower of vegetables and take a humble tomato plant, with nutritious soil, water being plentiful and the right amount of sunlight, you will have tomatoes to fill a bucket. People are similar, in the right environments they will flourish and as a leader it will be one of your proudest achievements!

Let's summarise 5 key points from this article:

• Mantras and values will shape your culture and behaviors
• One bad apple can spoil a barrel. Toxic behavior will turn a positive culture into a negative one in an instance.
• Communities of practice will really help to push your culture in the right direction. Giving people an opportunity to learn more skills will always be welcomed by a team.
• Leaders will also drive cultural improvement. A team that is micromanaged will not allow members to grow and this will not help your culture.
• Cultural improvement takes time, though it will be worth all of the wait, just like butter paneer, if made from scratch (my favorite).

Key Takeaway: A leader asks questions, shows people how, and says 'we', while a manager gives direction, holds authority, and says 'I'.