What Clean Language Teaches Us about Leadership

I hear a version of the same story from almost every leader I coach through a change initiative. They rolled out the plan. They held the town hall. They built the FAQ, anticipated the objections, and answered them before anyone had to ask. And still, weeks later, the resistance shows up anyway: quieter, harder to name, and somehow worse than if someone had just said no in the room.

Here's what I've come to believe: the plan was never the problem. The problem is that people were working through the change in their own private metaphors and models the whole time, and no one ever asked what those were.

The Research Behind the Failure Rate

John Kotter has been studying this for three decades. In his original 1995 Harvard Business Review article and later in Leading Change, he found that most major transformation efforts fail. Not because the strategy was wrong but because leaders skipped the human work of building genuine buy-in before pushing for results. His eight-step model exists because urgency, a guiding coalition, and a vision people can actually repeat back to you matter more than the org chart redesign itself. Kotter's numbers have held up remarkably well since: leaders still reassure before they understand, argue logic against a feeling, and ask questions that hand people an answer instead of letting them find their own.

Ronald Heifetz gives us language for why this keeps happening. In "The Work of Leadership," his 1997 Harvard Business Review piece with Donald Laurie (and more fully in Leadership Without Easy Answers), Heifetz draws a distinction I use in my own work: technical problems have known solutions and can be solved by an expert with authority. Adaptive challenges require the people affected to change their own values, habits, or sense of identity. No restructure memo, however well-written, can do that work for them.

Most change initiatives get treated as technical problems. Almost none of them are.

Where Clean Language Comes In

This is where I've started leaning on Clean Language, a questioning method developed by David Grove decades ago and still criminally underused in leadership circles. The premise is simple: ask deliberately non-leading questions that use a person's own words back to them, rather than supplying your interpretation of what they mean.

"And when you think about the change, that's like what?"

That single question does something a status update never will. It invites metaphor instead of opinion, and metaphor is where the real information lives. I've had a team member describe a restructure as "walking into a room where the furniture's been rearranged in the dark." Nobody could have engineered that image through logic. But once it's on the table, you know exactly what to ask next: "And whereabouts do you notice that most?" "And what would you like to have happen?" Five or six questions later, "vague unease" turns into "I need fifteen minutes to understand the new reporting lines," a request nobody would have guessed from the outside, and one that's actually solvable.

The rule I keep coming back to, the one that's easiest to skip: "And is there anything else about that?" Ask it until they say no. The real answer rarely shows up on the first pass.

Why This Is a Systemic Nudge, Not a Soft Skill

It's tempting to file Clean Language under "nice to have," a communication technique for leaders who have the luxury of slowing down. I'd push back on that framing. Richard Thaler and Cass Sunstein's foundational work on nudge theory makes the case that small, well-placed changes in how a choice is presented can shift behavior more reliably than mandates ever do, because they work with people's existing motivations rather than against them. Applied to organizations, the research on nudging is clear that isolated interventions fade fast: what sticks is a system of small, consistent moves that reinforce each other over time.

A single Clean Language question is a small move. But used consistently, in every 1:1 ahead of an announcement, every time someone seems resistant but won't say why, it becomes exactly the kind of systemic nudge that compounds. You're not overhauling how your organization communicates. You're changing six words in every conversation you have this quarter, and letting that accumulate.

What to Avoid

A few traps, because I've fallen into all of them:

• Don't finish someone's sentence, even when you're sure you know where it's going.
• Don't ask "why" as it invites justification, not exploration.
• Don't rush to fix the thing they just told you; these questions are for understanding first.
• And don't reach for this when someone has a straightforward logistics question or has explicitly asked for your direct advice. Clean Language is for the conversations where the real answer isn't obvious yet, not for every conversation.

The Question I'd Leave You With

Kotter tells us why the plan alone isn't enough. Heifetz tells us why the problem is rarely technical. And nudge research tells us that small, repeated moves beat one-time overhauls. Clean Language is simply the tool that sits at the intersection of all three — a way to find out, before you roll out the next FAQ, what your team is actually standing in front of.

So, regarding the next conversation about change that is on your calendar this week: “And when you think about how change goes, that's like what?

Resources and further reading

• John Kotter, "Leading Change: Why Transformation Efforts Fail," Harvard Business Review, 1995, and Kotter's 8-Step Process
• Ronald Heifetz and Donald Laurie, "The Work of Leadership," Harvard Business Review; and Heifetz, Leadership Without Easy Answers (1994)
• Richard Thaler and Cass Sunstein, Nudge: Improving Decisions About Health, Wealth, and Happiness (2008); and Prosci's overview of nudge theory in change management
• David Grove's Clean Language questions, adapted in my Clean Language for Leading Change guide