Why Human-Centered Design Is the Missing Link in Organizational Change

How teams that think together outperform those that simply execute

Over the past decade, working across some of the most capable organizations I know, I’ve noticed something consistent. The ones that hold up under pressure, that make better decisions, adapt faster, and don’t fall apart when complexity compounds, aren’t necessarily the ones with the smartest people. They’re the ones that have built something specific: the capacity to reason together.

That’s harder to build than it sounds, and rarer than most leaders realize.

Human-centered design

Human-centered design has been one of the most reliable levers I’ve seen for developing this capacity. HCD is a set of practices rooted in deep observation of how people actually work, iterative problem-solving, and collaborative sense-making. Most people associate it with product design.

But in the organizations I’ve worked with, its real power shows up elsewhere: in how teams frame problems, surface assumptions, and make decisions together under uncertainty. In this sense, HCD is really a normalized way of aligning teams on how to solve problems together.

Change management

Classic change management models tell you what needs to happen at a high level. HCD gives you the ground-level methodology for how behavior actually changes, group by group, team by team, in ways that earn adoption rather than demanding it.

Large organizations don’t transform as a monolith. They transform subculture by subculture, each with its own starting point and resistance. HCD, applied to organizational change, is how you navigate that. It isn’t a diet pill, it’s a fitness program. Consider some of these examples.

At Autodesk, HCD methods spread across the organization not because leadership mandated them but because they worked. When software architects used them to tackle the re-architecture of AutoCAD for cloud computing, one of the company’s most technically complex projects, even the engineers were convinced. The methods eventually reached sales, marketing, finance, and senior leadership. By 2016, 90 percent of Autodesk’s participatory research projects were conducted collaboratively with customers, nearly double the rate just a few years earlier.

At Cox Enterprises, a program that started with a single designer teaching stakeholders his process eventually reached 55,000 employees across a $24 billion company. Today Cox has 800 certified practitioners managed by a core team of three, and more than 100 people apply each year for 15 instructor spots. It became self-sustaining not because it was mandated but because teams experienced what it felt like to reason together more effectively.

At Intuit, “Design for Delight” wasn’t really about design. It trained thousands of employees in deep customer empathy, rapid experimentation, and iterative learning. The goal was changing how teams thought together about problems that didn’t have obvious solutions. Having worked with teams across the organization, I’ve seen firsthand how this shifts the way people reason together, not just how they design.

Different industries, different entry points, same result: teams that could think together under pressure, not just execute under direction.

The obvious objection is that this works in small teams but can’t survive contact with a large organization. IBM’s transformation starting in 2012 trained more than 150,000 employees in structured collaborative practices, not as a training event but as a sustained shift in how teams worked. I observed that transformation firsthand.

Practices like “playbacks,” structured critique sessions where teams present work-in-progress for candid feedback, gave those employees a common operating system for how to think together. The result was measurable: across dozens of account teams, design thinking maturity increased 42 percent on average, with voluntary enrollment above 90 percent. The program spread because it worked, not because it was required.

This matters more now than it ever has.

AI in the mix

AI dramatically increases what individuals can produce. But organizations don’t succeed because individuals are productive. They succeed because people can think, decide, and act together. And AI doesn’t automatically help with that. In fact, recent research from the Wharton School found that when people use AI that gives confident but wrong answers, they adopt those wrong answers roughly 80 percent of the time, and report feeling more certain than if they’d had no AI at all. Speed without collective judgment doesn’t produce better decisions. It produces faster misalignment.

The organizations I’ve described didn’t stumble into this capability. They designed for it, invested in it, and treated it as strategic work rather than a training program or a culture initiative.

That’s the pattern worth recognizing. In an era when AI is raising the ceiling on individual output while quietly making team reasoning harder, collective judgment is the capability that separates the organizations that adapt from the ones that just accelerate in the wrong direction.

The question isn’t whether your teams are using AI. It’s whether they’ve built the capacity to question it. HCD is how.

---

About the author

Jim Kalbach is a noted author, speaker, and instructor in innovation, design, and the future of work. He is currently Chief Evangelist at Mural, the leading online whiteboard.

Jim is the author of several books: Designing Web Navigation (O’Reilly, 2007), Mapping Experiences, 2nd Ed. (O’Reilly, 2020), and The Jobs To Be Done Playbook (Rosenfeld, 2020). In 2023 he co-authored Collaborative Intelligence (Wiley, 2023) with Mariano Battan. Jim is also the Co-founder and Principal at the JTBD Toolkit, an online resource with learning, trainings, and content.