Ask a fourteen-person business to draw itself and you will get a chart of people. Names in boxes. Lines running upward to the founder. Everyone recognises it and nobody argues with it.
Ask the same business to draw how one job moves from first enquiry to final invoice — every pair of hands it passes through, every point where it sits waiting, every decision that has to come back to the founder — and the room goes quiet.
The second drawing is the business. The first is the seating plan.
Most founder-led businesses between £250k and £3m have never made the second drawing. Not through carelessness — nothing in conventional business advice asks them to. They inherit a way of organising built for railway companies in the 1850s, adapted for factories, adapted again for offices, and now stretched over businesses that operate nothing like any of them.
What follows is a different organising logic. I call it the Orchestration Model: designing a business around the movement of work rather than the arrangement of people. It explains why capable founders end up as the mechanism their business runs on, and why buying better tools rarely changes it.
Why doesn’t an org chart describe how your business actually works?
An org chart describes arrangement — who reports to whom. It says nothing about movement: how work travels, where it changes hands, and who owns it from start to finish.
These are two entirely different questions, and only one of them determines whether your business runs well.
A chart answers who is responsible for which area. The question that governs your week is what happens between the areas. Value is produced in sequences that cross the boxes: an enquiry becomes a proposal, a proposal becomes a project, a project becomes an invoice, an invoice becomes a renewal. No box owns that sequence. Each one owns a segment of it.
At enterprise scale, the consequence is silos. At your scale, the consequence is you.
There is a further wrinkle at SME scale: many businesses in this band have no chart at all. They have a mental one. Sarah does the invoicing. Dave handles anything technical. Anything that fits neither description finds its way to the founder. That is not a structure. It is a memory — and it works perfectly well at six people. At fourteen it produces a business that only one person can explain.
What is the Orchestration Model?
The Orchestration Model is a way of designing a business around how work moves through it — its flows, handovers, ownership and decision rights — rather than around who reports to whom.
It is not a flatter hierarchy or a rearrangement of the same boxes. It is a structure of movement rather than a structure of position, and it asks you to design four things deliberately.
- Flows. The repeatable sequences that produce value. Most businesses of this size have between five and nine that genuinely matter: winning work, delivering it, bringing new clients on, getting paid, hiring, and putting things right when they go wrong.
- Handovers. The points where work changes hands — person to person, team to team, human to software. Almost all structural loss happens here rather than inside the work itself. Things do not usually fail while someone is doing them. They fail in the space between two people who each assumed the other had it.
- Ownership. Who holds a flow end to end, as distinct from who performs the tasks inside it. Most businesses have clear owners for tasks and no owner at all for the sequences those tasks belong to.
- Decision rights. Where judgement sits, what has to escalate, and what does not. Undefined decision rights are the most common single reason a founder cannot step back from a business that is otherwise running well.
A worked example. A proposal is accepted by email on a Friday afternoon. The person who sold it knows. The person who has to deliver it does not, because nothing formally connects the two — only a habit of mentioning things. On Monday the client rings to ask when work starts, the founder takes the call, and the job gets carried across the join by hand. Nothing was done badly. Nobody was slow. The sequence simply had no design in it.
The four are rarely neglected equally. In most businesses I look at, flows are broadly understood, handovers are assumed rather than designed, ownership is absent, and decision rights have never been discussed at all. That order tends to be the order of repair.
Set against the chart, the difference in what each one asks you to think about is stark.
| Chart thinking asks | Flow thinking asks |
| Who reports to whom? | What happens between them? |
| Who is responsible for this area? | Who owns this sequence end to end? |
| Do we have the right people? | Does work move without someone carrying it? |
| Who should I hire next? | Where does work currently stall? |
| What is everyone’s job? | What has to come back to me, and why? |
Why does the founder end up holding the business together?
When flows are not designed, the gaps between people default to the founder — the only person in the structure who touches every sequence.
This is not a discipline problem or a delegation problem, though it is almost always diagnosed as one. It is structural. If nobody owns a flow, somebody has to carry the work across each join, and that somebody will be whoever can see the whole thing. In a founder-led business, that is one person.
I map this for other people for a living.
When I ran it on my own consultancy, I did not get a number. I got every point. There is no piece of work in that business that continues if I stop touching it.
But that is a business where I am deliberately the product. It is not the same as a business that quietly became dependent on its founder while nobody was designing anything.
One of those I chose. The other chooses you.
Accumulated dependency has a specific fix, and it is not delegation. Delegation moves tasks. Orchestration moves ownership. Hand someone the tasks inside an undesigned flow and you have not removed yourself from it — you have simply added a person who now needs you before anything can finish. Which is why the next hire so often fails to buy back the time it was meant to.
It also explains why the pressure intensifies rather than eases as the business grows. Every additional person adds joins, not only capacity, and joins are precisely the thing nobody has designed. Six people generate a manageable number of them. Eighteen generate considerably more than three times as many, and all of them run through the one person able to see the whole sequence.
What changes when AI enters an unorchestrated business?
AI does not create the problem — it prices it. Agents work inside flows rather than roles, so a business that cannot describe its flows cannot deploy them well.
Larger organisations are discovering this loudly. Much of the current commentary about the org chart no longer holding is really an observation that software has started doing work which answers to no box, crossing every system at once, owned by no department. That is the enterprise version of a problem you already have at fourteen people.
The practical version is simpler. You cannot automate a business you cannot describe.
Every serious implementation begins with an honest account of how work currently moves — what triggers it, what it produces, where the decisions sit, what happens at the exceptions. Businesses that already hold that account can act on it in weeks. Businesses that do not spend the first month building it, or skip that step and buy tools that speed up individual segments while the handovers carry on leaking. Faster work, moving through the same undesigned joins, arriving at the same bottleneck.
The structural argument here stands with no AI in it at all. But the window in which an undescribed business is merely inconvenient rather than expensive is closing.
What replaces the org chart?
Not another chart. What replaces it is an Operating Map: a working description of the flows the business actually runs on.
For each flow, an Operating Map records what sets it in motion, the sequence of steps it moves through, every point where it changes hands, who owns it from start to finish, which decisions it contains and who is entitled to make them, and where it currently fails. It is a picture of movement rather than position.
The org chart tells you who works here. The Operating Map tells you how the work works.
The chart does not disappear. Employment, line management, pay and reporting all still need one. It simply stops being the primary description of your business and becomes what it always was underneath — a personnel document.
The first thing an Operating Map tends to expose is not inefficiency. It is that two or three flows have no owner at all, and that everybody had quietly assumed they belonged to somebody else — most often to you.
Building one is a discipline rather than a template. The value is not in the drawing; it is in the arguments it forces while you are drawing it, most of which are about ownership and none of which resolve themselves quietly. It is the artefact I build with clients inside the Business Architecture Intensive, and I have yet to see one produced without at least one uncomfortable discovery.
Where orchestration sits in the Architecture Gap
The Architecture Gap Framework assesses five structural dimensions: Business Model, Offer Architecture, Client Journey, Revenue Design and Operational Structure.
Orchestration is what Operational Structure is really measuring. Not headcount. Not which tools you have bought. Not how much of your process has been written down. Whether the movement of work through the business has been designed — or whether it accumulated, one workaround at a time, until it took the shape it currently has.
A business can hold a sound model, strong offers, good clients and sensible pricing and still be weak in this dimension. It shows up in a recognisable way: everything running through one person, work that stalls in the gaps rather than in the tasks, and a growing suspicion that the business is harder to run than its size warrants.
The question worth sitting with
The future of business design is not a better chart. Charts describe arrangement, and arrangement was never the constraint.
Every business already has an orchestration model. Most were never chosen. They were accumulated — a workaround here, a favour there, an exception that became the rule — until the founder became the mechanism holding the whole thing in sequence.
The question is not whether your business is orchestrated. It is whether you designed it.
Operational Structure is one of five dimensions assessed by the Architecture Gap Scorecard.
It is a straight answer to whether the movement of work in your business was designed or inherited — and where the structural constraint actually sits.
If you would rather talk it through, a Discovery Conversation is the direct route. Bring the flow that frustrates you most and we will look at what is holding it in place.


