Growth that is outrunning the structure underneath it

Revenue is up and everything is harder. Decisions that used to happen in a corridor now need three meetings and a document. Nobody can point to the week that changed, and every individual meeting is justifiable. The organisation has outgrown the structure underneath it, and the structure has not been redesigned so much as accreted.

This is one of the few problems where an experiential session is not a substitute for real work but genuinely the fastest way to see the shape of it.

Growth is a load, and load has to be carried

The useful frame is structural rather than organisational. Revenue is what the organisation has promised to deliver. Delivery capability is what carries it. Systems are what hold the delivery together. Culture is what nobody measures and everything rests on.

Add promises faster than you add the capability to carry them and the structure does not fail politely. It bends somewhere specific — usually the same handful of people — and then it fails visibly, late, in front of a client.

What makes this hard to see from inside is that the failure is silent for a long time. The system shows stress well before collapse, and almost nobody inspects it, because everyone inspecting it is busy carrying load.

The four places it shows up first

One person is the interface. Someone knows how the whole thing actually works and is in every escalation. They are the reason it functions and the reason it cannot scale, and they are usually the last person anyone thinks to protect.

Onboarding has stopped working. A new starter used to be productive in three weeks and now takes three months. This is the clearest measure of whether systems exist independently of the people who built them.

Decisions climb. Things that were decided by the person doing the work are now decided a level up. Nobody mandated this. It happens because the cost of a wrong decision rose and nobody adjusted who is allowed to make one.

Quality is defended by heroics. The work still goes out well, and it goes out well because specific people stayed late. This reads as commitment and is actually the alarm.

What a session surfaces that a consulting exercise does not

An organisational design study produces a good document. The problem is not usually analysis — it is that the leadership team does not viscerally agree on what is load-bearing, and so the recommendations get implemented selectively.

Put the same group in a room where they have to buy materials, commit to revenue, and physically build something that carries what they promised, and three things happen that a document does not produce.

They over-commit, and watch it. Given a chance to promise more revenue, teams take it. When the promise is a weight hanging on a structure they built, the consequence is not a forecast.

They discover what they under-invested in. The parts that earn nothing — the trunk, the roots — are the parts that get bought last and cut first. Every team does this. Watching it happen to your own build is more persuasive than being told it is a pattern.

They argue about the right thing. The debrief is not about the model. It is about which part of their actual organisation is the trunk, and whether anyone has looked at it this year.

Three redesigns that make it worse

Each is a reasonable response to a real symptom, and each is the wrong one often enough to be worth naming.

Adding a layer to fix escalation. Decisions are climbing, so a management tier goes in to absorb them. The tier absorbs the decisions and adds a handover, and six months later the same decisions are climbing again from one level lower. The problem was decision rights, not headcount.

Reorganising around the current bottleneck. The structure gets rebuilt around whatever hurt most last quarter. It relieves that and creates a new interface somewhere else, because every reorganisation trades one set of handovers for another. Redesigning around a symptom guarantees another redesign.

Hiring ahead of the system. More people are added to carry more load, into an organisation whose onboarding has already stopped working. The new people take three months to become productive and consume the time of the very individuals who were already the bottleneck. Capacity goes down before it goes up, and often it does not come back.

The common thread is treating a load-bearing problem as a capacity problem. More weight and more scaffolding are not the same intervention.

Who should be in the room

The leadership team, and the people who currently absorb the strain — which is not always the same group, and the gap between them is frequently the finding.

Twelve to twenty-five, in three to five teams, so that teams build separately and can see each other's structures at the end. The comparison at scoring is a large part of the value: three groups given identical constraints produce visibly different structures, and the differences are the conversation.

One inclusion worth insisting on: whoever is currently the single interface. They will recognise their own position within the first round, and they are usually the only person who can describe what would actually have to change.

Turning it into a design decision

The session is diagnostic, and what it earns is a shorter list of real questions. Three worth leaving the room with:

What are we currently promising that we cannot carry? Usually there is a specific client, product line or commitment the room already knows about and has not said aloud.

Which single person is load-bearing, and what would it take to change that? Name them. Then decide whether the answer is documentation, a hire, or removing the thing that only they can do.

What earns nothing and holds everything? Whatever it is, it is under-funded, because it always is.

Where HDX fits

Rooted is the direct fit and is built on this metaphor without being coy about it. Leaves are revenue, branches are delivery, the trunk is systems, roots are culture and values and earn nothing at all. Materials are bought from the Bank on a purchase order and there are no returns. Requesting a leaf registers it as a commitment, and a leaf that cannot be hung legally by the end of the round costs half its value and earns nothing — so an ambitious order at minute two can be the thing that loses the round. A facilitator inspects every structure against the building code before scoring.

Where the structural problem is also a change problem — the shape has to alter and people are attached to the current one — Shift is the usual companion, and the two run well as a pair across a day and a half.

Two things worth reading first: when a physical game beats a digital one, because for this subject the physical version is not a stylistic preference, and measuring behaviour change for the ninety-day check that tells you whether the design decision actually got made.

---

If growth is outrunning the structure underneath it, tell us what the room needs to work out.