Crushed, Problems, Implemented: Where Your Culture Sends New Ideas

Published October 12, 2026

Part 6 of The Generative Culture Series

I would mention an idea to my team or my director, and the next thing I knew, a senior vice president was asking me about it. My director would say yes, go ahead, try it. People were open to new ideas and excited by them, and an idea could just spread.

What happens to an idea after someone says yes?

Ron Westrum studied safety in aviation and health care and sorted organizational cultures into three types, and his table gives a new idea a different fate in each. In a pathological (power-oriented) culture, the entry reads “Novelty crushed.” In a bureaucratic (rule-oriented) culture, it reads “Novelty → problems,” an arrow he never explains; I read it as approved ideas that run into trouble. In a generative (performance-oriented) culture, it reads “Novelty implemented.” (Ron Westrum, “A typology of organisational cultures,” Quality and Safety in Health Care 13, Suppl. II, 2004.)

Crushed

In a culture organized around power, the answer depends on who gains from the idea. You get a yes when someone can build their standing on your change, and you meet resistance when it would spread power around or take some away, whatever the change would do for the work. You learn to ask who an idea would make more powerful before you ask whether it would work. The idea is settled before anything has been tried.

Problems

I have watched approved ideas go through so much proper process that they were delayed for a long time or derailed outright. The approver had said yes and the team was willing, but the organization kept running on rules that made the change hard to carry out, and the agreement stood for a long time without changing any work. Westrum describes bureaucracy as “often a compromise between competing interests,” which is the fair reading of those rules: they were written for other reasons, and nobody has to oppose an idea for them to slow it.

Sometimes the team gets as far as a pilot, a limited trial of the change, and meets resistance there. People have no time for it. Everyone else keeps following the existing rules. The person who raised the idea still holds a yes, and the team never gets to find out whether the idea helps.

Some ideas are adopted in name only. Ask about one and the answer is “Oh yeah, we do that,” when nobody does, or nobody does it in any way that matters. The organization counts the change as made, the existing work goes on as before, and the idea never produces a result anyone could evaluate.

Katherine Klein and Andrew Knight draw the line this way: adoption is the decision to take up an innovation, and implementation is getting people to use it skillfully and consistently. Many organizations adopt innovations and never see the expected benefits, they find, and a key reason is that implementation fails. (Katherine J. Klein and Andrew P. Knight, “Innovation Implementation: Overcoming the Challenge,” Current Directions in Psychological Science 14, no. 5, 2005.)

DORA, the research program that adopted Westrum's typology for software delivery, asks people in its culture survey whether “On my team, new ideas are welcomed.” (DORA, “Generative organizational culture”.) That measures reception. A team can welcome an idea and still never get it used.

The same article reports that putting a new practice into use costs performance at first, even when the practice will eventually help. That early dip is where a bureaucratic culture turns on an idea. I have seen a single hiccup send an idea everyone supported back to the drawing board. “We have to evaluate this.” “We need more discussions before we do anything with it.” Nobody declares the experiment over; the team simply loses the chance to work through the difficulty it had only begun to meet, and the dip Klein and Knight describe gets read as a verdict.

The dip is also the time when a new practice gets fixed. Marcie Tyre and Wanda Orlikowski studied new production and service technology in three organizations and found that most adjustments came soon after introduction; as people grew used to the new way, unresolved problems settled into everyday practice and stopped getting fixed. (Marcie J. Tyre and Wanda J. Orlikowski, “Windows of Opportunity: Temporal Patterns of Technological Adaptation in Organizations,” Organization Science 5, no. 1, 1994.) If the same clock runs for a new practice, a team sent back to the drawing board at the first hiccup never gets that window at all. Westrum's line applies: “The ability to think is often coextensive with the ability to act.” A team that cannot act on an idea cannot finish thinking about it either.

Then the idea changes hands. When existing processes take back over, it becomes nobody's idea; the organization's momentum absorbs it until little of the intended change remains and no one can say when it ended.

In a power-oriented culture, ownership follows credit and blame. An idea that makes progress becomes the idea of whoever can claim the most credit, a leader or an executive or a department, and the person who raised it drops out of the story. An idea that is rolled back belongs to whoever will take the fall. Credit flows up, blame flows to the convenient, and neither follows who actually raised the idea.

Implemented

In a generative organization an idea can come from anyone, and the people who hear it answer with small yesses: try it, let's find out. The first try usually stays inside one team and takes on one problem. It is cheap because it is small, and its job is to find out whether the idea deserves more effort.

The clearest one I ran myself began with EventStorming, a workshop format in which the people who run a business process map it on a wall, event by event, with everyone who touches the process in the room. (Alberto Brandolini, “EventStorming”.) I asked for permission to try it and kept the first run small: one project, one team, one application. The next run was a multi-day offsite for leaders and executives who were in town for it. After that, people asked me about it and asked me to present what we had done.

A first try should be as small as it can be and still answer the question. Eric Ries makes the same point about new products: a minimum viable product, in his sense, is the fastest way to test an assumption about how the product will work, and he is explicit that it is “not necessarily the smallest product imaginable.” The term is often used for the smallest thing a team can ship, which answers a different question. (Eric Ries, The Lean Startup, Crown Business, 2011, 77, 93, and 294 (ch. 6, n. 3).)

From there the idea spreads by being seen. When I changed how my team labeled its work in Jira and how it ran its board, other teams asked about it. When I wanted to try new facilitation and coaching with a DevOps team, I got approval, tried it with my own team first, and spread it from there with my director's support and other leaders'. The rest of what spread from where I sat was the same kind of thing, practical and close to the work: how meetings were run, how people managed their calendars, how teams kept their backlogs (the ranked list of work waiting to be done), how teams reported and communicated, and how teams were organized. Some practices were normal almost at once; others took much longer to settle.

Everett Rogers, whose Diffusion of Innovations is the best-known account of how new practices spread, calls the end of that process routinizing: the practice becomes part of regular activity and loses its separate identity. He also explains who comes to own an idea that spreads. When the people adopting an innovation change it as they take it up, “they begin to regard it as their own,” and they are more likely to keep using it once the special resources that launched it are gone. (Everett M. Rogers, Diffusion of Innovations, 5th ed., Free Press, 2003, 428–429.) In a generative culture, the idea belongs to everyone who has shaped it since.

Westrum's own evidence for this entry is a hospital study. He describes a “culture of conscious inquiry,” in which people are encouraged to speak up and think for themselves, and cites the surgical teams that made a new cardiac surgery technique routine while others failed to. Amy Edmondson, Richard Bohmer, and Gary Pisano found that the successful teams learned differently: they prepared together, ran practice sessions and early trials, and reflected on what happened. (Amy C. Edmondson, Richard M. Bohmer, and Gary P. Pisano, “Disrupted Routines: Team Learning and New Technology Implementation in Hospitals,” Administrative Science Quarterly 46, no. 4, 2001.) A surgical team is not an office team, but the sequence is familiar: try small, look at what happened, try again. The study measured whether a team learned to run the technique; it did not ask whether the technique was worth running.

A generative team inside an organization that is not generative can pay for it. Westrum asks whether such an isolated unit becomes a “Cinderella,” the target of hostile jibes and political actions, and says it happened to a high-performing naval laboratory he studied. An idea can be fully implemented inside a team and still be endangered by the organization around it.

Not every idea deserves to spread

Some ideas should stop, and Rogers names the assumption that gets in the way. Pro-innovation bias is the belief that a new idea ought to spread through a system, preferably quickly, and that rejecting it is a failure. By his own account, the research built on his work has asked why people adopt far more than why they reject, and why they keep using an innovation far more than why they stop. (Rogers, Diffusion of Innovations, 106–107, 111.) Organizations make the same mistake when they treat continued use as proof of value. Rogers allows that an innovation should sometimes be discontinued, because it does not solve the problem it was introduced to address, and says sustaining it “is not universally a positive outcome.” (Rogers, Diffusion of Innovations, 430.)

Generative organizations are, in Westrum's words, “probably more creative,” and he notes that effective organizations exist that are not generative, carried by “a brilliant algorithm or a charismatic leader.” A generative culture improves an idea's odds of being tried. Whether the idea is any good is a separate question.

Running the experiment

Propose the idea if you think it would help the team. In a culture where an idea is judged by whom it would empower, none of what follows comes first; safety does. Make the first try the smallest experiment that can still show an impact: change how one meeting runs, or how the team tracks its work. Small tweaks first, then more as each one holds up.

Before you start, agree with the team that this is an experiment, name the hypothesis (what you expect the change to do), and settle what a bad first week means: evaluate, adjust, or try again. A rough first week is for learning and reflection; sometimes you don't know what won't work until you try it. Agree, too, on what failing the test means. If the idea does not solve the problem it was brought in for, you stop, and the experiment has done its job. An evaluation agreed before the first week asks what to adjust. One called after the first hiccup asks whether to continue.

An approver reads for the same things. When I am the one approving, I say yes quickly to a proposal that is thought through and designed as an experiment: a hypothesis, the reason for trying it, how the team will test it in small steps, and what it will do if the test fails. A proposal with those four parts gets a fast yes from me because the hard conversation, what happens if it fails, has already happened.

After I say yes, I ask the team to put up a plan: the timeline, what it will take to test the idea and find out whether it is worth more, and what would tell us to experiment more or differently. When the first few weeks are rough, that plan is what gets reopened. The idea may need more time to shake out. The team may need more room to experiment, or the test may need a different shape, or the work may need a different owner. What goes back to the drawing board is the plan. The idea keeps its yes.

Where the evidence stops

Westrum's table gives each culture a word for what becomes of a new idea and no account of how it gets there. The research on how ideas spread has a gap of its own, and its authors say so. Trisha Greenhalgh and colleagues reviewed more than two hundred studies of innovation in service organizations and reported that only one had set out to follow whether an adopted innovation was later dropped. (Trisha Greenhalgh et al., “Diffusion of Innovations in Service Organizations: Systematic Review and Recommendations,” Milbank Quarterly 82, no. 4, 2004.) That review is two decades old and counted one kind of study. The link between a bureaucratic culture and those patterns is my observation laid over Westrum's table, and none of these studies tests it. A study that followed new ideas through all three cultures and recorded which were kept and which were dropped would be the test. My own view stops short of it too: I left the role where I watched those practices spread before I could see which of them were kept.

Back to Field Notes