What We Learned Restructuring the Team Mid-Growth

What We Learned: Restructuring the Team Mid-Growth

The design takes an afternoon. Everything that decides whether you were right happens in the year after

A restructure takes an afternoon to design and about a year to find out whether you were right.

That gap is the whole subject of this article. Everything I have read about organisational change focuses on the afternoon, the design, the logic, the new lines on the chart. Almost nothing prepares you for the year. I want to write about the year.

Not because the design does not matter. It does. But the design was the part we found manageable. The part that kept us awake was the long stretch where you have already made the decisions, people are already living inside them, and you have no way of knowing yet whether you were correct. One of the hardest parts of leading change is that only time can tell you whether you made the right call. You do not get immediate validation. You get responsibility first and clarity later.

Key Takeaways

• Change is an afternoon. Transition is a year. You can move the boxes on Monday; nobody transitions on Monday.
• You will undercommunicate the reasoning. Leaders under-explain the why by an order of magnitude and are surprised people did not hear it.
• Performance dips before it improves. Judge the reorg at the bottom of the dip and you will reverse a correct decision.
• People adopt change at different speeds, and speed is not loyalty. The slow adopters are often the most invested.
• Be impatient with the problem, patient with the people. Most leaders do the exact reverse.

Why the Structure Had to Change

We reached a point where the company had outgrown the structure that got us there. At the time it arrived as a series of small, unrelated irritations that took a while to recognise as one problem.

Processes that worked perfectly with a smaller team started producing bottlenecks. Not because they were badly designed, but because they had been designed for a different size of company and nobody had noticed them quietly expiring. Ownership became blurry at the edges, which is where ownership always goes first. And too many decisions relied on a handful of people.

That last one is the diagnostic I would give anyone reading this. Count how many decisions in a normal week cannot be made without three or four specific individuals. If that number is growing faster than your headcount, your structure is already behind your business, whatever the org chart says.

The temptation in that situation is to hire. It feels like the obvious answer, it is easy to justify, and it produces a visible sense of momentum. It also does not work, because adding people to a structure that is already the bottleneck adds coordination cost without adding capacity. You end up with more people waiting on the same four humans. So it became clear that if we wanted to keep growing, we could not simply hire more people. We had to redesign how the organisation operated.

Here is what I have come to believe. Companies rarely outgrow their people. They outgrow their systems. A structure is fitted to a size, a speed, and a set of assumptions, and all three change underneath you without any announcement. The structure that got you here is almost never the structure that keeps you here, and the reason it is so hard to see is that it is still working, right up until it is not.

What We Changed

Four things, and they are unremarkable when you list them.

1. We redefined responsibilities. Who does what, stated properly rather than assumed.
2. We introduced clearer departmental ownership. Which department owns which outcome, including the awkward ones that sat between two teams and had therefore belonged to neither.
3. We documented processes that had lived in people's heads. This one deserves a note. Undocumented knowledge is the most dangerous asset a growing company holds. It looks like efficiency, the person just knows, so why write it down, and it works beautifully until that person is on holiday, or ill, or busy, or gone. Every process living in one head is a single point of failure wearing the disguise of competence.
4. We changed reporting lines. The hardest of the four by a distance, because a reporting line is not an administrative detail. It is a relationship.

On paper, all four decisions were logical. I would defend every one today. But on paper is exactly the problem, because change management is never logical. It is deeply human, and the two are not the same thing at all.

The Org Chart Is the Easy Part

Here is the single most useful idea I have encountered on this subject, and I wish I had found it before rather than during. Change and transition are not the same thing, a distinction the change theorist William Bridges built an entire model around.

Change is external. A new reporting structure, a merger, a new role. It can happen on a specific date, at a specific time. An organisation can implement change in a single afternoon. Transition is different. It is the internal process people go through before that change becomes real to them, and it does not follow your timeline. It follows theirs.

It usually unfolds in three stages. First an ending, where people let go of the way things used to be. Then the uncomfortable middle, where the old way no longer works but the new one has not taken shape. Only after that a new beginning, when people start to feel confident in the new reality and genuinely commit to it.

The part that stayed with me most is that transition begins with an ending. Not with something new. With a loss. That explains something I could not understand for a long time. You can announce a change that is objectively good for someone, more ownership, a clearer role, a better path, and still be met with what looks like resistance. Most of the time it is not resistance. It is someone processing what they are about to lose before they can appreciate what they are about to gain. Maybe a manager they trusted. A part of their role they were proud of. Or simply a way of working that made them feel capable.

You can implement a new structure on Monday. Nobody transitions on Monday.

Change versus transition: the chart moves in a day, people move through endings, a neutral zone, then a new beginning

People Move at Different Speeds, and That Is Normal

One thing I underestimated is how differently people process change.

Some see the opportunity immediately. They are already asking what this means for what they can build next, sometimes within an hour. Others see uncertainty. Not opposition, uncertainty. They want to understand what happens to the things they were relying on, and until they know, they cannot really engage with the upside. And some need time before they can even understand why the change is happening at all. Not because they are slow. Because they were not in the conversations where the problem became visible, so they are being handed a solution to a problem they have not yet seen.

The mistake we made early was treating the first group as the baseline and the other two as friction to be managed. That is wrong, and wrong in a way that costs you. The fast adopters are not more committed or more capable. Often they are simply the people whose position improved, or who happen to be comfortable with ambiguity. Meanwhile the people who take longer are frequently the ones with the deepest investment in how things worked, which is not a flaw. It is usually why they were good.

There is a practical consequence. If you calibrate your communication to the fast group, you will stop explaining far too early, because the loudest feedback you are getting is "yes, makes sense, got it."

People adopt change at different speeds, and the slowest are often the most invested

Communication During Change Is a Rate, Not an Announcement

Looking back, I do not think we got the organisational decisions wrong. What I would change is this: I wish we had invested even more time in explaining the why behind every change, repeating it more often, and creating more opportunities for people to ask the difficult questions, including the ones that are uncomfortable to answer.

Communication during change is not something you do once. It is something you keep doing. Research on organisational change keeps finding the same mistake: leaders dramatically underestimate how much they need to communicate a vision. John Kotter's work put it as undercommunicating by a factor of ten or more. A leadership team spends weeks shaping where the company is going, explains it in one meeting or one message, uses a tiny fraction of its communication to reinforce it, and then wonders why people still do not understand. The numbers vary from study to study. The conclusion never does. Once is almost never enough.

Another well-known study explains why. In Elizabeth Newton's 1990 experiment, "tappers" tapped the rhythm of a famous song on a table and "listeners" tried to name it. Tappers predicted listeners would guess right about 50% of the time. The actual figure was 2.5%, three songs out of 120. The tappers were not overconfident. They could hear the full melody in their heads while they tapped. The listeners heard only disconnected knocks.

The same thing happens inside organisations. By the time leaders communicate a decision, they have debated it, refined it, and lived with it for weeks. For everyone else it is the first time. What feels repetitive to the leader often feels like the beginning to everyone else. So repeating the vision is not repetition. It is clarity.

Now picture what a leader is doing when they announce a restructure. You have lived with the problem for weeks. You have run the options, argued them, discarded three versions, and arrived somewhere you are confident about. The melody has been playing in your head for a month. Then you present it, clearly and carefully, in forty minutes. You are the tapper. You think you are getting the message across one time in two. You may be getting it across one time in forty.

This reframed the whole thing for me. When somebody asks the same question for the third time, the instinct is to hear it as resistance, or as not listening. It is almost never that. It is a person hearing taps and trying to find the tune.

The curse of knowledge: the tapper hears the melody, the listener hears knocks, 50% predicted versus 2.5% actual

So the practical version is simple and slightly boring.

• Explain the reasoning, not just the decision. The decision is a conclusion. The reasoning is what lets people apply it to situations you did not anticipate.
• Say it in more formats than feels necessary. Written, spoken, in the all-hands, in the one-on-one, in passing.
• Invite the hard questions explicitly. People will not volunteer them into a room that feels settled. "What is the part of this you are least comfortable with" gets you something. "Any questions?" gets you silence and a private conversation you never hear.
• Answer the same question the fifth time as if it were the first, because for that person it is.

That last one is harder than it sounds, and it is where most of us fail.

Not Everyone Stayed

This is the hardest reality of scaling, and there is no honest way to write about restructuring without addressing it. Not everyone stayed.

Some people looked at the new direction and realised it was not where they wanted to grow. Others found the pace of adaptation genuinely difficult. Those conversations were hard in a way that performance conversations are not, because they were not about performance. They were about fit within a company that was evolving.

I want to be precise about why that is harder, because it is underappreciated. In a performance conversation there is something to fix. There is a gap, you can name it, and both people can work on it. It is uncomfortable but it has a shape. When a good person's role changes underneath them and the new version is not what they want, there is no gap to close. Nobody did anything wrong. There is no villain and no correction available. You are two people acknowledging that the thing you both signed up for has become a different thing. That is a harder conversation to have and a much harder one to carry afterwards, because there is nothing to resolve. You cannot fix it with better feedback and you should not try to.

The other thing I would say, carefully, is that a departure during a restructure is not automatically evidence the restructure was wrong. Sometimes a company changes and a person does not want the new version of it, and both of those facts can be legitimate at once. Treating every exit as a failure of the change is its own kind of dishonesty. It pressures you into pretending the company has not changed, which is unfair to the people who stayed and unfair to the person leaving, who deserves a straight answer about what the role has become. That does not make it feel any better. It just makes it true.

Not Every Change Can Be Gradual

Here is one of the hardest things we have had to say to a team. As leaders, we would all love to introduce every change slowly, with unlimited time for people to adjust. That is the humane version and it is what everyone wants. Reality does not always allow it. Sometimes the business moves faster than people's comfort zones and you have to ask your team to adapt sooner than they would like. Not because you do not care about the discomfort, but because the alternative is a slower change that arrives too late to solve the problem it was meant to solve.

This is the message leaders are most tempted to soften, and softening it is a mistake. If you imply a gentler timeline than you can deliver, you buy about three weeks of calm and then spend six months rebuilding trust. People forgive a hard timeline explained honestly. They do not forgive a comfortable timeline that turns out to have been a comfortable fiction.

I would add one piece of self-criticism, because "the business demands it" is an argument that can be abused. Sometimes speed is genuinely required by circumstances outside your control. And sometimes an urgent change is the direct consequence of a leadership team not deciding something six months earlier, and the urgency is your own doing passed downstream. Those two situations feel identical from inside the leadership seat and completely different to everybody else. Being honest with yourself about which one you are in does not change what you have to ask of people. It changes how you ask, and whether you are willing to say the quiet part: that this is faster than it should have to be, and here is why.

Patience Is the Skill Nobody Lists

If you asked me which leadership quality mattered most through all of this, I would not say decisiveness or vision. I would say patience, and I think it is chronically underrated.

Patience to answer the same question for the fourth time without any edge in your voice. Patience to listen to concerns without becoming defensive, which is harder than it sounds. Patience to accept that people adopt change at different speeds and that the speed is not a measure of loyalty or ability.

But patience does not mean delaying necessary decisions. That distinction is everything, and it is where the word usually gets misused. Patience is not slowness. It means giving people the support and communication they need while still moving the business forward. The cleanest way I can put it is this: be impatient with the problem and patient with the people.

Most of us instinctively do the reverse. We are patient about fixing the structure, because that is uncomfortable and confrontational and there is always a reason to wait another month. And then, having finally decided, we are impatient with everybody who has not caught up within a fortnight. Both are the wrong way round.

The Part Nobody Sees

The hardest part was not making the decisions. It was carrying the responsibility of not knowing whether they were the right ones yet.

Leadership is usually described in a language of certainty. Conviction, clarity, decisiveness. Restructuring does not give you any of those on the timeline people imagine. You make the best decision available with the information you have, and then you wait. Sometimes weeks. Sometimes months. Before you can honestly say whether it worked.

There were plenty of late nights where I replayed conversations in my head. Wondering whether we had communicated something clearly enough. Whether someone had interpreted a change differently from how we meant it. Whether we could have made the transition easier for a particular person if we had said it another way, or earlier, or in a different room.

You also come to understand, in a way no framework prepares you for, that every decision touches real people rather than boxes on a chart. Every new reporting line changes who someone talks to when they are stuck. Every shift in ownership changes what someone is proud of. Every process change affects somebody's daily rhythm, their confidence, and their sense of whether the ground under them is stable. None of that shows up in the design document. All of it shows up in the year that follows.

There is a specific trap in here that I have watched catch leaders. The discomfort of an unresolved decision is genuinely worse than the discomfort of a bad one. Not knowing is harder to sit with than knowing something went wrong, because at least a wrong answer can be acted on. That creates enormous pressure to resolve the uncertainty artificially, to declare the restructure a success early, point at two encouraging signals, and move on. Declaring victory too soon is one of the most common errors available here. It is not usually dishonesty. It is a leader trying to make an unbearable ambiguity stop. The discipline is to let it stay open. To keep saying, internally and sometimes out loud, that we do not know yet.

Restructuring Is a Process, Not an Event

The biggest lesson for us was this: restructuring is not something you do. It is something you are in, for a while.

You rarely know whether the decisions were right immediately. Sometimes it takes months before the organisation settles enough to evaluate anything objectively, because for the first stretch you are measuring a system that is still in the neutral zone, where the old way is gone, the new way is not fluent yet, and almost everything looks worse than it did before.

That point is worth sitting with, because it produces the most expensive mistake available. Performance during a transition usually dips before it improves. People are learning new reporting relationships, new ownership boundaries, new processes that used to be automatic. Of course it dips. If you evaluate the change at the bottom of that dip, which is precisely when the pressure to evaluate is highest and when the people who disliked the change are most confident, you will reverse a correct decision at the exact moment before it starts working.

The neutral-zone dip: performance drops before it climbs, and the bottom is when pressure to judge peaks

The 30-90-180 Check

Two practical things I would do differently, and now do. Together they are the framework I would hand any operator running a reorg. Call it the 30-90-180 Check.

1. Define what you expect to see, in advance, at 30, 90, and 180 days. Not KPIs necessarily. Observable things. At 90 days, are decisions still routing through the same four people? Are the Heads of Department escalating less? Has the number of "who owns this" questions dropped? Writing these down before you start is what stops you evaluating on mood, and mood during a transition is a terrible instrument.
2. Name the neutral zone out loud. Tell people, at the beginning, that there will be a period where things feel worse and less clear than before, that this is expected, and that it is not evidence the decision was wrong. This costs you nothing and it changes how the middle period is experienced. Discomfort that was predicted feels like a stage. The same discomfort unpredicted feels like a mistake.

The 30-90-180 Check does one useful thing above all. It moves the decision about whether the restructure worked from the moment you feel worst to a set of observable signals you chose while you were still thinking clearly.

The 30-90-180 Check: pick observable signals in advance, and name the neutral zone before it starts

Getting the structure right is only half of it; getting a team through the transition is the other half. It is the same discipline behind scaling delivery without losing quality. If you are building an operation that has to keep growing, see how AP Collective works.

What I Would Keep

I do not think we got the organisational decisions wrong. The bottlenecks we set out to remove are gone. Ownership is clearer than it was, the same clarity we describe in our piece on team ownership. Decisions no longer pile up behind a handful of people, which was the thing that started all of it.

What I would change is almost entirely in the transition rather than the design. More explaining. More repetition. More deliberate space for the difficult questions, earlier, and from the people least likely to volunteer them. More willingness to say "here is the part I am not sure about yet" instead of presenting a finished object.

Today we are over 60 people across seven departments, fully remote and spread across most of the world's time zones. The structure fits the company we actually are rather than the one we used to be. That is worth what it cost. It did not feel worth it in month two.

Which brings me to the thing I would most want another operator to take from this. Successful change management is not about getting everyone comfortable. That is not available, and chasing it will make you slow and vague and eventually untrustworthy. It is about helping people understand where you are going, why you are going there, and giving them enough trust to walk through the uncertainty with you.

Anyone can announce a change. Building the understanding, trust, and commitment around it is the real work, and it happens over months, in dozens of unremarkable conversations, long after the announcement everybody remembers.

Frequently Asked Questions (FAQs)

Final Word

You get responsibility first. Clarity comes later, if you can hold your nerve long enough to receive it.

Related Reading

• Fixing "I Thought You Owned That". Clearer ownership was one of the outcomes this restructure was built to produce.
• Managing People, Not Just Headcount. Why a reorg is planned around capacity and motivation, not bodies.
• Scaling Marketing Delivery Without Losing Quality. The delivery system a growing structure has to protect.

Sources

• John Kotter on undercommunicating the vision, and PMI on communication and project outcomes link: pmi.org/learning/thought-leadership/pulse/essential-role-communications
• William Bridges, the transition model (endings, neutral zone, new beginnings) link: en.wikipedia.org/wiki/William_Bridges_(author)
• Elizabeth Newton, 1990 tappers and listeners study, on the curse of knowledge link: ssir.org/articles/entry/loud_and_clear

Disclaimer

This article reflects operating experience at AP Collective and is general management guidance, not a guarantee of results. Cited studies and models are attributed to their original researchers and publishers; figures were accurate as of the last-reviewed date.

Written by Lui, Chief Operating Officer at AP Collective.