Creating a Culture of Experimentation: A Conversation with MongoDB’s Mark Porter and Accenture’s Michael Ljung


Asked about his failures, Thomas Edison famously replied, “I have not failed 10,000 times. I have successfully found 10,000 ways that will not work.” Today’s organizations need that same spirit of experimentation to build the innovative products demanded by customers and by the health of their businesses. But while we’re all familiar with the doctrine of failing fast, few organizations are able to provide and protect a culture that allows people to rapidly test and discard ideas, burning through the ones that don’t work and finding the best ones quickly.

In a live conversation, Mark Porter, MongoDB CTO, and Michael Ljung, Global Software Engineering Lead and Chief Software Engineer for Accenture, examined this phenomenon, suggesting ways to encourage a culture of experimentation and to normalize experiments that don’t work as expected.

“A lot of our clients expect it to be right the first time,” said Ljung. And that’s true even when working on large, new-build projects that can last two years. “So by the time we get to the finish line, maybe the business context or the circumstances or the requirements have changed.”

To make progress in such an environment, Ljung and Porter suggested that leaders encourage their teams to experiment, rather than emphasizing fast failure (or success). With an experiment, it’s clear that an idea might not work. It’s also clear that everyone will learn from it, adjust, and move on quickly. To emphasize that the organization is committed to learning, Porter suggested that teams think about what they will learn if the experiment fails – and then commit that to paper.

Experimentation isn’t optional, said Porter. A lot of organizations have legacy processes and procedures that may have outlived their usefulness. “If you’re going to move to the cloud, or you’re going to transform yourself into a modern organization using modern software, you have to break a lot of that stuff,” said Porter.

Porter and Ljung agreed that psychological safety is critical to enabling a mindset that seeks out experimentation, and that enables people to try new things. That’s a challenge for many organizations and teams, and Porter and Ljung presented some strategies that can help. Porter thinks of psychological safety as a result of candor, context and accountability. Citing Netflix CEO Reed Hastings, Porter said, “Executives need to speak truths all the way down to the organization. What is going well, what is going poorly. And the organization needs to speak truth back up.” Context is often business context – the importance of the current initiative to the overall organization and marketplace. These principles will encourage teams to request accountability – something that just about every manager wants.

Ljung spoke in terms of empowerment. He notes that agile gives product owners the ability to make a wide range of decisions, but product owners can be stuck in mindsets that don’t allow that. He recalled one project, when a product owner had to make a decision near the end of a sprint. Rather than make the call, the project manager punted to the next steering committee meeting. “The great thing about agile is you have been empowered to make the decision for the steering committee,” noted Ljung. That decision can be implemented quickly and at low cost. And if the steering committee doesn’t like it, “We can change it and it doesn’t cost a lot of money to change… By experimenting and having those small bumps along the way, we can get a much better result at the end.” Which is exactly what modern organizations – not to mention Edison – are all about.