You can't turn an organisation around by announcing where it should be. You find the north star, decide the hard things quickly, and then walk toward it in steps small enough that nobody gets lost on the way.
A few days ago a colleague asked me what I had learned over the past months, and I described it as a zigzag. Thinking about it afterwards, I don't think the word was right, even though it sounded good on the call. What I actually did was move in small and deliberate steps, and I chose each of them so that I wouldn't lose people along the way. What I was trying to say is that the direct route rarely works when an organisation is heading somewhere else, because you can know exactly where you want to end up and still fail by marching straight there, while every day a few hundred decisions are made by people who are still facing the old way.
When you join an organisation to change it, you never start from zero but in the middle of something that is already moving. There are structures, habits and unwritten rules that made sense to someone at some point. A previous manager may have tolerated a behaviour long enough for it to become normal, and somewhere a workaround turned into a process and the process into part of people's identity, until "we've always done it this way" stopped being an excuse and became part of how people think about their work. The unit I took over had been losing money for a long time and had been through a round of layoffs that ended in legal disputes. It also had a culture that called itself self-managing, although nobody had ever given it the framework that self-management needs, and one person described it to me as "no mama, no papa": people had been left alone, and that was called freedom. Several told me they were afraid to come to work, and the effort I saw was about as low as you would expect from people who believed the next termination could be theirs. I don't think any single person was to blame for that. It was a pattern that had grown over years, and patterns like that don't change because someone new arrives with a slide deck.
The idea that helped me most when I read up on organisational change is that there are two ways of looking at an organisation. The classic view treats it as frozen, so you unfreeze it, move it and freeze it again, while the other view assumes it is changing all the time anyway and the job is to redirect what is already under way. The second view described my situation much better, because this organisation wasn't standing still and waiting for someone to give it direction; it was moving quite fast, just away from where it needed to go. In a situation like that, pushing harder tends to make things worse. People who feel threatened don't learn faster, they become more defensive, and what really needs to come down is their fear of looking incompetent while they try something new. When you are new in a role it is tempting to arrive with the answer already in your pocket, and I felt that temptation as much as anyone, but a solution brought in from outside, however good it is, tends to set off something like an immune reaction. Ronald Heifetz said it in a way I keep coming back to: people don't resist change, they resist loss, and leadership means "disturbing people — but at a rate they can absorb".
I should be honest that step by step doesn't mean being slow on everything, because it is easy to make patience sound more romantic than it is. When I read about how companies actually went through big transformations, I was surprised to find that the real shifts usually happened in short and concentrated bursts, and that small changes did not simply add up to something bigger over time. That doesn't contradict moving step by step, but it made me realise that some of the steps have to be big and fast. So I tried to be fast on structure, money and people while staying patient with culture. A unit that loses money every month can't wait until everyone feels ready, but how people see their role, how they talk to each other and whether they dare to take initiative can't be ordered, and if you try, people learn to perform the new behaviour while they hold on to the old one. Since I'm writing about research anyway, you will probably hear at some point that seventy percent of change programmes fail. When researchers went back to the sources of that number, they found nothing behind it except an estimate its own authors had called "unscientific", which was then repeated often enough to sound like a fact, so as far as I can tell nobody knows the real rate.
The fast part included hiring and letting people go, and I think an organisation learns more about its leadership from how that is handled than from any all-hands meeting. Joel Brockner's research on layoffs has a finding I wish more managers knew: when people experience the process as fair, how bad the outcome is matters a lot less for how they react, and that is true for the people who stay as much as for those who leave. The people who stay watch very closely how the leavers are treated and draw their own conclusions about how safe they are. Cutting people alone rarely fixes a business anyway, and companies that only reduce headcount usually do no better than their competitors. When I looked more closely at our own situation, much of the bench that had led to the earlier layoffs turned out to be a problem of finding the right allocation rather than a people problem, and no number of terminations would have fixed that.
The hardest decisions, though, were the ones where I had to correct a call of my own. More than once I ignored a gut feeling because acting on it would have been inconvenient at the time, and later had to undo what I had set in motion. Correcting it was always the right thing to do, but it cost the people involved more than if I had listened to the first signal. Decisions like that are uncomfortable in a particular way because there is nobody else to blame, and I think they are also the ones people remember best, because they show whether you hold yourself to the same standards as everyone else.
Then there was the question of structure. Jo Freeman wrote about the tyranny of structurelessness in the early seventies, and her point still holds today: there is no such thing as a group without structure, there are only groups where the structure is invisible, and invisible structure "becomes a way of masking power". The so-called self-management I found worked exactly like that. Information travelled through personal back channels, decisions were made wherever someone happened to have influence, and people who wanted to take initiative couldn't tell whether it would be welcomed or punished. So I did something that might sound like the opposite of where I wanted to go and introduced clearer structure, with departments that had a named head, explicit escalation paths and decision rights that were written down. I said openly that this wasn't the shape I wanted in the end, and that it was there to give people firm ground before I asked them to walk further. How well a team is set up at the start matters more for how well it manages itself than any coaching it gets later, and enough companies have learned that the hard way. Buffer removed its managers entirely and brought them back after new joiners got lost and growth stalled. I believe an organisation needs some structure rather than a completely flat hierarchy, but only just enough of it to function and nothing beyond that, which for me means as few line managers, targets and formal appraisals as possible. I find that a good test for any structure, including a temporary one, because it makes you ask what each layer is actually there for.
During the same months I was also removing process wherever it made ownership impossible, and although that may look contradictory, both moves served the same goal. Some of what stood in the way was my own doing, for example a period in which every task assignment had to go through me, which made me the bottleneck for exactly the initiative I kept asking for. Other parts were inherited, like the rule that internal work had to be booked to customers, which meant that every attempt at knowledge sharing died as soon as billable hours got tight. A study from Germany kept coming back to me during that time. In the years after reunification, Michael Frese and his colleagues found that East German workers showed less personal initiative than West Germans, and that this had nothing to do with them being different people. It came from years in jobs with little control and little complexity, and as those jobs gained more room to act, the initiative came back. What I took from this is that you can't train initiative into people while their work stays controlled; you have to give them the room back and accept that it will take time. Looking at what actually survived in our organisation, every initiative that lasted had a named owner, time that was really paid for and visible backing from leadership. Opening up the numbers to the team belonged in the same category, because people can't take ownership of a business they aren't allowed to see.
I also learned that a north star and a plan are different things, and that the difference matters more than it first seems. When a plan doesn't survive contact with reality, most leaders, me included, react by adding more detail, more instructions and more control, and each of those tends to make the gap wider. What works better is a clear intent combined with freedom in how people get there. The intent I kept repeating was that we were moving from renting out a pair of hands to being experts who help customers create value. It was short enough to remember, and it gave people a way to check their own decisions without me making those decisions for them. Edgar Schein, who spent decades studying organisational culture, wrote a line I underlined twice: "It is the consistency that is important, not the intensity of the attention." In my experience one inspiring kickoff doesn't change a culture, but saying the same thing every week and acting on it when it is inconvenient slowly does.
The other sentence I kept repeating was "Every one of you is a grown-up, and I want to treat you that way", and I took it from my time at Equal Experts. I joined them at thirty-five and was the youngest in the room, and their own description is that being "low ego, grown-up experts with intrinsic motivation to learn and grow means we can minimise hierarchy". What I took from it is that being a grown-up has little to do with age, or even with how much someone is trusted, and a lot to do with attitude. Grown-ups take ownership of their customers, their projects and their own development, of sharing what they know and of the organisation they are part of. That is what I played back to the team, and it shaped the career levels we designed: levels that describe impact, to replace those that mostly measured tenure. Everyone is a consultant, whatever their craft, and leadership is a role you take on in addition to your level rather than a rung you climb to, so nobody has to stop doing the work to be recognised as senior. Ideally nobody would ever have a reason to hunt the next level, but we also had people at the very beginning of their careers, and a step that only works for experienced people would have lost them along the way. I wasn't willing to accept that, because people who are just starting out need a place in a transformation as much as anyone, and in a few years they are the ones who will carry the organisation. So the junior level stayed, and I still think a junior is a grown-up too, someone who manages themselves within their means, owns their learning path and asks for guidance early. They need more guidance than others, but they deserve the same respect.
My own preference would be as little hierarchy as possible, and I deliberately didn't try to get there in one go. After years of uncertainty, people first needed something stable to stand on, and taking away the remaining structure at that point would have repeated the original mistake of calling neglect autonomy. What I aimed for instead was a management layer that doesn't grow and has a little less to do every quarter, because people take over more of the management of their own work. The order I had in mind was to hand over control of how the work is done first, then of how the team is set up, and only then of the direction, one step at a time.
Which brings me back to "we've always done it this way". I still think it is the most dangerous answer you can get in an organisation, but not because the people who say it are wrong. We all stick with the status quo far more than its merits justify, and economists have measured that in decisions as real as health insurance and pensions, so it is a very human default rather than a flaw in someone's character. Resistance also carries information, and quite often the people driving a change create it themselves through broken agreements and poor communication, which I have been guilty of too. So whenever I heard that sentence I tried to treat it as a question rather than a wall and asked why we do it this way. Sometimes there was a good reason that nobody had ever written down, and then the habit stayed, but more often the reason had left the company years ago and only the habit was still there. Either way, asking the question was what started the conversation, while arguing with the sentence never got me anywhere.
I don't know yet how much of this will last, since culture changes over years rather than quarters, but there are signals that are hard to fake. A colleague who had been close to resigning told me recently that they felt seen and had stopped thinking about leaving, and the forced optimism of the early months has given way to something quieter that feels more real. The numbers moved as well, but what I'm proudest of is seeing people take ownership of things nobody asked them to own. None of that came from one big move. It came from small steps towards the same point, at a pace people could follow.
The direct route may be the shortest line on the map, but it is rarely the one people can actually walk.