Behavior change is harder than tech nerds think

David Novák

Senior Consultant

3 October 2026

Thirty Years of Corporate Culture in Czech Companies and the Bill That Arrived This Year

Nobody Founded the Culture Here

When you talk to the CEO of a large Czech company today about why people aren't doing things differently, sooner or later the word culture comes up. Usually with a sigh and the implicit assumption that culture is something the company has because it just happened. It didn't just happen; it accumulated.

Every stage of the last thirty years brought difficult decisions that were reasonable at the time. Not stupid or cowardly—reasonable. Culture is what remains in companies after thirty years of locally correct decisions that no one added up. And because it arises from summation, it cannot be changed with posters or speeches.

Six Waves of Corporate Culture

Before 1989, the enterprise was an institution. It provided food, recreation, housing, and daycare. At the same time, it declared official values that were publicly nodded to and privately disbelieved. One thing remained from that era that isn't talked about much: the word "values" has a longer and worse history in the Czech Republic than in the West. When a poster with five values arrived in 2004, it didn't hit a blank slate. It hit a time-tested Czech reflex.

The 1990s were about survival and pleasing the new owner. Values didn't enter big companies—processes, reporting, and compliance did. The top three banks were privatized within three years: ČSOB to KBC in 1999, Česká spořitelna to Erste in 2000, and Komerční banka to Société Générale in 2001. The reasonable choice back then was not to invent anything of one's own, but to adopt from elsewhere. What remained was a learned reflex that what matters is decided elsewhere. To this day, the phrase "it came from headquarters" closes a debate instead of opening it.

Around 2003, group HR sends documents with values, mission, and vision to branches. They are translated, often literally and poorly. Competency models, assessment centers, and 360-degree feedback arrive. Culture gets a name for the first time: "corporate values." What remained from this were values as a graphic element and the first measurable phenomenon: the gap between what hangs on the wall and what actually gets done on Friday afternoon. People remember that gap longer than the values themselves.

After 2008, culture disappears from the agenda. HR switches to cost reduction and restructuring. Only one thing survives because headquarters demands it: the annual employee satisfaction survey. In my view, this is the most lasting damage from this entire history. For fifteen years, we have been asking people questions, and no decisions come out of their answers. The results are presented, three focus areas are named, and next year we ask again. Over that time, the organization learned that filling out a survey is a form of politeness. When management says "we'll do a survey" today, half the employees hear "nothing will happen." And they have fifteen years of empirical evidence for that.

In the mid-2010s, the labor market flipped. The Czech Republic has long had the lowest unemployment rate in the EU, and the fight for talent becomes the primary constraint on growth. Culture gets a budget for the first time, but only as a recruitment argument. Benefits, MultiSport cards, offices, employer branding. In everyday corporate speak, the word culture began to mean environment. Money went to what was visible on the career page. The manager's behavior in a meeting didn't get any.

During the same period, agile transformations took place in large Czech companies. They weren't called "culture." Yet they touched precisely what culture actually is: who decides what, how often it is reviewed, who it is done with, and what it means for something to be "done." It is our hypothesis—and I have no proof for it—but it is quite possible that the most effective cultural intervention of the last twenty years in our country was called agile.

In two years, COVID reduced culture to the question of how to keep people together when they don't see each other. A massive investment in internal communication and care arrived; wellbeing entered the HR vocabulary and never left. And then came the return to the office, which showed that culture is a power issue decided from the top down, even if it is discussed as a dialogue. That experience is fresh in organizations and lies just beneath the surface of every subsequent cultural initiative.

And over the last few years, everything is happening at once. Inflation shock, energy costs, ESG, digitization, cost-cutting programs, and now AI. And in many companies, an inevitable and long-planned headcount reduction is happening simultaneously.

The State of Culture in Companies Today

When you line up those six waves side by side, a few statements emerge that most practitioners will nod along to before you even finish speaking. Nobody founded the culture here; it was imported and translated. Most large companies in the Czech Republic have a culture written by no one who actually lives in it. Every wave was called something different and did the exact same thing: it spoke to the people without touching the system. Values, surveys, employer branding, wellbeing, and so on. Four different names, but one identical pattern.

The survey became a ritual of politeness. For fifteen years, we have asked employees questions, and no truly difficult decision comes from their responses. And no one ever decided who owns culture at the decision-making level. It was only decided who owns it at the agenda level. Those are two different things, and one of them currently has no leverage.

The Seventh Wave Is Called AI

Now AI joins this line, and it is different in one key aspect. For the first time in history, performance is demanded from culture. Adaptability, willingness to let go of routine, the ability to operate with fewer people, and trust in a tool that occasionally makes mistakes. None of this is a technology question. All of it is a behavior question. More is being asked of culture than in all previous waves combined, yet culture was never given the leverage to carry that weight.

Interestingly, those building the technology have hit the exact same wall. In an August interview with David Senra, Sam Altman admitted he was wrong about a few things: "The economy has huge inertia. People keep doing the same things, buying from the same companies, wanting to use their tools the exact same way." He added that "we were all too ambitious regarding time estimates." As we at Kogi have known well for over 10 years: changing people's behavior "is much harder than tech nerds realize."

It is worth noting that this is also a sales argument. For a year and a half, Altman has been building the narrative of a slow, manageable ramp-up to superintelligence, so slow adoption suits him quite well. Even so, it is a remarkable admission. A company with the best models in the world, unlimited capital, and half a billion users is bumping into the exact same issue as the CEO of a logistics company in Zlín. Behavior doesn't change simply because a better option becomes available.

Culture Runs Continuously; an Initiative Has an End

One common lesson emerges from all six waves, and it is uncomfortably simple: a cultural initiative is doomed from the start because culture runs continuously, while an initiative has an end. A continuous process cannot be overcome by an event, no matter how good it is.

This means that the failure of cultural programs is usually not a matter of execution. If the workshop had been better, the facilitator stronger, and leadership more engaged, the result would have been the same—just slightly delayed. This also applies to AI programs; it's just visible faster there because licenses are being counted.

The only thing that can change culture is something that also runs continuously. In every organization, such a thing exists, and no one calls it culture: it is the moment a project is assigned. The scope, objective, definition of done, and success metrics are written down. It happens every quarter, in every division, regardless of whether anyone has culture in their scope. The rule we insert into that moment at Kogi is simple: a project with an impact on people's behavior is not assigned without a named target behavior. One card, no new forum, no new committee. That is the entire governance.

For the first time on this topic, a definition of done is created. The system is installed the moment a project manager fills out that card without anyone asking them to. That can be measured, and it cannot be faked.

I must add where we currently are in this process. This distinction between an initiative and a system is our work-in-progress framework, not published research. So far, we have experience with it from one large project that has not yet concluded, so I describe it as a work in progress, not as something we have already proven. I also know of two risk areas: the card can degenerate into a formality (the only safeguard is that it must genuinely block something), and applying it across all projects will create bureaucracy—the exact thing we promise not to add.

The rest is simpler than it seems. Your culture will make its decision this year regardless. Not in a workshop, but during budget cuts, during AI rollouts, and when deciding who stays. The question is not whether you will "do" culture. The question is whether you will be there when it happens.