Back to Blogs
Amy Edmondson’s Right Kind of Wrong is essential reading for CMOs who need innovation without recklessness. Her message is not “celebrate failure.” It is build the psychological safety required for candor, smart risks, better experiments, and faster learning. For marketing leaders, that means creating teams brave enough to test, learn, and win.
Amy Edmondson’s Right Kind of Wrong should be required reading for any CMO trying to build a culture of experimentation without accidentally creating a culture of chaos.
I came into our CMO Huddles conversation as an admiring student of Amy’s work. Her research on psychological safety has already changed how many leaders think about teams, candor, and learning. But what struck me most in this conversation was how directly her ideas apply to modern B2B marketing.
CMOs are under enormous pressure to deliver today while preparing the company for tomorrow. They need to test new channels, pilot AI use-cases, sharpen positioning, explore new customer experiences, and make budget bets in uncertain markets. That requires experimentation. And experimentation requires the possibility that something may not work.
The point is not to fail. The point is to succeed by taking smarter risks.
Amy was quick to clarify that high achievement and intelligent failure are not opposites. In fact, the best performers often learn faster because they are willing to enter new territory.
“If you're not taking what we might rightly call smart risks, you're not doing your job at the very highest level.”
That is a helpful reframe for CMOs. Playing it safe may protect this quarter’s optics, but it can quietly create a bigger risk: Irrelevance.
Amy defines a failure simply as “an undesired result.” But not all failures are equal. She distinguishes among basic, complex, and intelligent failures. Basic failures are preventable mistakes in familiar territory. Complex failures come from multiple factors interacting in messy systems. Intelligent failures happen in new territory, in pursuit of a worthwhile goal, with a thoughtful hypothesis, and at a scale no bigger than necessary.
That last part matters. Amy is not inviting CMOs to bet the brand on an untested idea. She is inviting them to create experiments that are “just big enough to learn.”
Psychological safety is often misunderstood as niceness, softness, or protection from consequences. Amy’s version is much more demanding.
“Psychological safety refers to a climate where people believe that candor is expected.”
That sentence should stop every CMO for a moment. Psychological safety is not a room where everyone feels cozy. It is a room where people tell the truth early enough for the truth to help.
Amy described it as a learning environment where people disagree, raise concerns, and take interpersonal risks in order to reduce business risks. In marketing, that might mean a demand gen leader challenging a campaign assumption, a brand leader questioning a launch message, or a junior team member pointing out that an AI pilot has privacy or quality risks no one wants to discuss.
This is where experimentation and psychological safety meet. If people are afraid to speak up, the team does not get smarter. It just gets quieter.
Many CMOs already set aside part of the budget for experimentation. Amy’s guidance makes that practice more rigorous.
“If you know in advance that your experiment is going to work, it's not an experiment.”
That sounds obvious until you watch how organizations behave. Many teams label safe extensions as “experiments” because the term feels innovative. But a real experiment includes uncertainty. The discipline is to make uncertainty useful, bounded, and connected to strategy.
For CMOs, that means each experiment should answer four questions:
This applies to campaign tests, positioning work, AI pilots, new event formats, customer community ideas, and cross-functional go-to-market experiments. “Let’s play with the tool” is not enough anymore. The better question is: What are we trying to learn that could improve the business?
One of the hardest parts of leadership is that experience can become a reflexive no. A team member brings an idea. The CMO has seen something similar before. The answer comes fast: Tried it, didn’t work, next.
Amy called that the opposite of a high-quality conversation.
“You've got to be willing to listen with curiosity.”
That may be the micro-shift many CMOs need most. The goal is not to abandon judgment. Judgment is part of why leaders become leaders. But judgment has to be moderated by curiosity because markets change, teams change, customers change, and tools change.
As Amy put it, “We haven’t stood in this river before.”
The CMO’s job is not to be the smartest person ending every conversation. It is to create better conversations so the team can make better bets.
Generative AI makes Amy’s framework even more relevant. Most marketing teams are experimenting with AI because the protocols are still forming. The tools are changing quickly, the risks are real, and the opportunity is too large to ignore.
Amy’s advice is to make AI experimentation more purposeful.
“Gen AI, clearly, almost by definition — what most of us are doing is experimenting.”
That experimentation now needs to mature. CMOs should move from tools-first play to strategy-first learning. Are we using AI for convergence, meaning efficiency, speed, and cost reduction? Or for divergence, meaning new ideas, new experiences, and new ways to create value?
Both matter. But they require different teams, different hypotheses, and different measures of success.
Amy also warned that AI experimentation requires systems thinking. If entry-level work is automated away, where do future mid-level marketers learn judgment? If marketing builds AI workflows faster than governance can catch up, what downstream risks are being created?
In other words: Experiment, yes. But do it with eyes open.
One of the most useful sections of the conversation came from Amy King’s question about attribution, complex systems, and skeptical cross-functional stakeholders. Modern marketing outcomes rarely have a single cause. Campaign design, sales follow-up, market timing, product readiness, customer behavior, and external conditions can all collide.
Amy’s answer was beautifully practical.
“The answer lies in the quality of the debrief.”
She advised leaders to start with “what happened?” not “who did it?” That distinction is everything. The wrong first question turns a learning conversation into a trial. The right first question opens the system.
For CMOs, after-action reviews should apply to both successes and failures. If the team invested money, time, or trust, the organization should get the learning value back.
A good debrief asks:
The goal is not blame. The goal is to do better next time.
CMOs live in the tension between the quarter and the future. Sales needs pipeline now. The CEO needs answers now. The board wants confidence now. But the experiments that create tomorrow’s advantage rarely feel urgent today.
Amy named this clearly:
“The pull of the now will always win, unless we are thoughtful and mature and override it.”
That is a leadership sentence. The future will not protect itself. CMOs have to make experimentation explicit, budgeted, and visible. They have to educate stakeholders on the risks of not taking smart risks. They have to ask, as Amy suggested, what happens if we keep doing things the way we have always done them?
And if the CEO does not fully embrace experimentation? Amy offered two paths: Try to influence through inquiry, and if that fails, build the best learning-oriented department possible.
“Then buffer your people.”
That may be the quiet work of great CMO leadership: Creating enough psychological safety inside marketing that the team can learn, even when the broader organization is still catching up.
Amy Edmondson is not asking CMOs to celebrate failure. She is asking them to design better learning systems.
That means distinguishing preventable mistakes from intelligent failures. It means creating psychological safety so people speak up before small risks become large ones. It means making experimentation strategic, not performative. It means asking better questions, debriefing honestly, and protecting the future from the urgent.
Her simplest micro-habit may be the most powerful:
“Ask more questions. And I mean good questions.”
For CMOs, that might sound like:
Right Kind of Wrong is an essential read or listen for marketing leaders who are serious about innovation. The goal is not to fail more. The goal is to learn faster, take smarter risks, and create the conditions where success has a better chance of showing up.
Psychological safety helps marketing teams surface concerns, challenge assumptions, and share learning before problems become expensive. It is especially important when CMOs are asking teams to test new ideas, adopt AI, or make bets in uncertain markets.
An intelligent failure is an undesired result from a thoughtful experiment in new territory. It is connected to a goal, based on a reasonable hypothesis, and kept small enough that the organization can learn without creating unnecessary damage.
Start with a clear business or marketing goal, define the hypothesis, decide what success and learning look like, limit the downside, and debrief afterward. Experiments should be designed to improve future decisions, not simply prove that the team was right.
AI pilots should move from random tool testing to strategy-led experimentation. CMOs should clarify whether the goal is efficiency, creativity, customer experience, or new capability, then build small tests with the right mix of marketing, technical, legal, and business input.
Ask better questions. Instead of “Who caused this?” ask “What happened?” “What are we learning?” “What are customers seeing?” and “How can I help?” Genuine curiosity is one of the fastest ways to create psychological safety.
Want to hear more? Listen to the full conversation on Renegade Marketers Unite.
CMO Huddles helps B2B marketing leaders navigate high-stakes transitions, executive alignment, team design, board pressure, and growth expectations with support from peers who have been there.