The third way to introduce DevOps is to create a culture of trust that supports experimentation and risk-taking. This is the Third Way in Gene Kim’s Three Ways framework — and it is what puts the team on a learning curve steep enough to outpace the competition.
Why Culture Comes Last in Order, Not in Importance#
The Three Ways are sequenced for a reason. Flow first, because without flow there is nothing to feed back on. Feedback second, because without feedback there is nothing to learn from. Learning third, because learning is what compounds the first two over time.
But culture being third in sequence does not make it less important. It is the multiplier. A team with mediocre flow and feedback but a strong learning culture will out-improve a team with great tools and no psychological safety. Tools you can buy. Culture you have to build.
A Culture of Trust#
Trust is the foundation. Trust between team members, trust between teams, trust between teams and leadership. Without it, people hide problems, defend decisions instead of revising them, and avoid anything that might make them look bad. With it, problems surface early, decisions get challenged, and people share what they learned — including the embarrassing parts.
Building trust is not a one-off exercise. It is the cumulative result of how leaders respond when something goes wrong, how teams handle disagreement, and how the organisation rewards honest reporting. One blame-driven incident review can undo a year of trust-building.
Experimentation and Risk-Taking#
A learning culture treats every change as an experiment. You have a hypothesis, you ship a small test, you measure, you decide. Some experiments work. Some do not. The ones that do not are not failures — they are data points that cost less than the wrong decision would have.
Experimentation only works if risk-taking is genuinely allowed. If every miss gets punished and every win gets credited to the loudest voice, people stop experimenting. They stick to the safe path. The safe path is also the path where nothing new is learned and the competition closes the gap.
Steeper Learning Curve, Faster Market Adjustment#
The point of all this is speed of learning. Markets move. Customer needs shift. Technology stacks evolve. The team that learns faster than the market changes stays ahead. The team that does not falls behind, regardless of how good its initial product was.
This is what makes the Third Way the durable competitive edge. Anyone can copy your features. Anyone can hire engineers as good as yours. Few organisations can copy a learning culture, because culture is not a feature — it is the accumulated way the organisation actually behaves under pressure.
Make Learning Explicit#
Learning does not happen automatically just because experiments are run. It has to be made explicit. Blameless post-mortems on incidents. Retrospectives that produce concrete changes, not just venting. Game days and chaos experiments that surface unknown failure modes before customers do. Internal sharing — brown bags, demos, write-ups — so what one team learns becomes available to the whole organisation.
This is also where leadership matters most. Leaders model what learning looks like. If they admit mistakes openly, others will too. If they hide them, everyone else will hide theirs.
Key Takeaways#
- Culture is third in sequence, first in importance. Tools you buy; culture you build.
- Trust is the foundation. Without it, problems hide and learning stops.
- Treat changes as experiments. Some work, some do not. Both produce signal — if punishment does not poison the well.
- Speed of learning is the durable edge. Features and engineers are copyable. Culture is not.
- Make learning explicit. Post-mortems, retros, game days, internal sharing — without ritual, learning gets lost.
- Leaders model what is allowed. If leaders admit mistakes, the team will too. If not, the team will not.