Most founders can name what is wrong with their business. Very few can name where it is wrong.
That distinction does more work than it appears to. Ask a founder turning over £1.2m what their problem is and you will usually get a symptom: margins thinner than they should be, delivery permanently stretched, a pipeline that arrives in lumps, nobody able to make a decision without them. Every one of those observations is accurate. None of them is a cause. And a symptom treated as a cause generates activity rather than progress — which is why so many capable founders work harder each year for a business that behaves the same way.
Business architecture gives you somewhere precise to look. It has five structural dimensions, and every established business has all five, whether or not anyone sat down and designed them.
The five structural dimensions of business architecture are business model, offer architecture, client journey, revenue design and operational structure. Each depends on the one before it.
The useful question is not whether yours are good. It is which one is carrying the strain — and, more precisely, which one is carrying it first. Because these five are not a checklist of things to improve in parallel. They are load-bearing, and they bear load in a fixed order.
What are the five dimensions of business architecture?
If you have met business architecture in a corporate setting, you will have met capability maps, process models and information layers — descriptive artefacts built for enterprise architects to document an organisation at scale. These five dimensions are a different instrument. They are commercial rather than descriptive, and they are built for a founder-led business of between £250k and £3m, where the person reading them is also the person who has to change them.
Here is what each one governs, and what it inherits from the dimension above it.
| Dimension | What it governs | What it inherits from above |
| 1. Business Model | The commercial logic: who you serve, what you sell them, how the money is made, and what the business must be good at for that to work repeatably | Nothing. This is the foundation |
| 2. Offer Architecture | What you sell, how many things you sell, and whether they form a progression or a menu | The coherence, or incoherence, of the model |
| 3. Client Journey | How someone moves from stranger to client to advocate, and what happens at each transition | The shape of the offers being sold |
| 4. Revenue Design | How money behaves: pricing, margin, the balance of recurring against one-off, and concentration | Everything the model, offers and journey have already determined |
| 5. Operational Structure | How work, roles, ownership and decisions are organised to deliver all of the above | The full weight of the four dimensions above it |
Naming the dimension is what turns a vague sense that something is not working into a specific structural problem you can act on.
Why does the order matter more than the scores?
A business does not fail at its weakest dimension. It fails at its earliest one — because every dimension inherits the flaws of the one above it.
This is the Load Order, and it is the single most useful thing to understand about the five dimensions. Structural weight transfers downwards. A gap in the business model cannot be repaired at the level of operational structure, because operational structure is where the model’s consequences arrive, not where they originate. Better operations applied to an incoherent model produce something worse than inefficiency: they produce a business that is highly efficient at doing the wrong thing, and therefore much harder to argue with.
Founders discover this the expensive way. A business with two incompatible commercial models installs a project management system to fix delivery strain. The system works. Delivery strain returns within two quarters, because the strain was never operational — it was two models competing for the same capacity, and the system simply made the competition faster.
The lowest score tells you where it hurts. The earliest gap tells you where to work.
Which gives you the reading rule that matters when you assess your own five: the lowest-scoring dimension is your headline, but the earliest significant gap is your lever. Those are frequently not the same dimension, and choosing the wrong one is how a year of committed effort produces very little structural change.
The five dimensions, in order
What follows is not a description of each dimension — it is the question that tests it, and what it looks like when that dimension is the constraint rather than merely imperfect.
1. Business Model — is there one, or is there a description of your biggest client?
The business model is the commercial logic of the whole thing: who you serve, what you sell them, how the money is made, and what the business has to be genuinely good at for that to work more than once. Most established SMEs have a documented model rather than a designed one — an accurate account of what happened, written up after the fact.
The test: can you state your model in one sentence that your delivery team would recognise as a true description of what actually happens each week? If the sentence needs exceptions to survive contact with reality — mostly, except for those three accounts — then what you have stated is not a model. It is a description of your largest client, generalised.
The constraint signature is more specific than that, and it is the most common structural failure in the £250k–£3m band: not a bad model, but two. A productised model and a bespoke model, running inside one company, sharing one delivery team, one price list and one set of people. Each is entirely defensible on its own. Together they compete for the same finite capacity, and because neither is acknowledged as separate, neither gets designed for. The symptoms all surface downstream — pricing that resists standardisation, a client journey with two front doors, capacity planning that never quite works — which is precisely why the cause goes unexamined for years.
In practice, the dimension a founder names as their problem is almost never the dimension constraining them. They name the one that hurts, and pain surfaces downstream of its cause. I have sat with founders who described an operations problem in forensic detail for the better part of an hour, when the actual issue was that the business had quietly been running two business models since roughly 2023 and had never once said so out loud.
2. Offer Architecture — a progression, or a menu?
Offer architecture governs what you sell, how many things you sell, and whether those things relate to one another. It inherits the model directly: an incoherent model produces an offer range that cannot be rationalised, because each offer was a reasonable answer to a different commercial question.
The test: does each offer create a natural reason to take the next one? If a prospect could pick any item in any order without consequence, you have a menu. A menu is not wrong, but it will not compound, and it puts the entire burden of sequencing onto your sales conversations.
The constraint signature is widening. Under pressure, the instinct is to add an offer — and adding offers adds surface area rather than capacity. The range grows, the team’s attention divides, and the business becomes broader without becoming any more able to carry weight.
3. Client Journey — designed, or merely happening?
The client journey is how someone moves from stranger to client to advocate: the sequence of experiences and, critically, the transitions between them. It inherits the offer architecture, because the journey is largely determined by the shape of what is being sold and in what order.
The test: after their first substantive interaction with your business, could a prospective client accurately describe what working with you would look like and where it would take them? If that answer depends on which member of your team they happened to speak to, the journey exists but has not been designed.
The constraint signature is loss at the transitions rather than inside the stages. Businesses tend to be competent at the stages — the pitch, the onboarding, the delivery — and undesigned at the joins between them. That is where enquiries cool, where onboarding stalls, and where good clients quietly stop being clients without ever complaining.
4. Revenue Design — designed, or accumulated?
Revenue design governs how money actually behaves: pricing, margin, the balance of recurring against one-off income, and how concentrated your revenue is in a small number of relationships. It inherits everything above it, which is why it is so often misdiagnosed as a pricing problem when it is a structural one.
The test: if you stopped selling tomorrow, what would still be arriving in ninety days? If the honest answer is very little, your revenue is reactive by design — or, more accurately, reactive by default.
The constraint signature is revenue that arrived rather than revenue that was designed: pricing set years ago by feel and never revisited, discounting used as a reflex rather than a decision, margin that varies by client without anyone being able to say why, and a concentration risk everyone can see and nobody has structurally addressed.
5. Operational Structure — does the business hold without you in it?
Operational structure governs how work, roles, ownership and decision rights are organised so that everything above it gets delivered. It sits last because it carries the full weight of the four dimensions above — which is also why it is the dimension that visibly fails first, and the reason it absorbs so much founder attention that belongs elsewhere.
The test: can your most senior person make a good decision in your absence — and do they, routinely, without checking? Capability is not the question. Whether the structure gives them the authority and the information to use it is.
The constraint signature is the founder operating as the business’s decision-making layer: work that flows through one person because the handovers were never designed, and a structure that functions well precisely because that person is always available. It works. It simply cannot grow, and it cannot be sold.
Why does a strong dimension sometimes hide the real problem?
Because strength downstream compensates for weakness upstream. An exceptional client journey can carry an incoherent business model for years, at a cost nobody attributes to it.
This is the reason self-assessment so often points at the wrong dimension. You score highest where you have worked hardest, and you have worked hardest wherever the pain has been — which is downstream of wherever the cause sits. The dimension you are proudest of is frequently the one absorbing the failure of the dimension above it.
It shows up as a business that feels good and scales badly. Exceptional people compensating for undesigned structure produce an experience clients genuinely value, delivered at a margin nobody can explain, at a pace that quietly consumes the best people in the company. Nothing looks broken. The cost is paid in margin, in founder capacity, and in a growth ceiling that keeps arriving earlier than forecast.
What do you do once you know which dimension is weakest?
Work on the earliest significant gap rather than the worst score, and change one dimension at a time — downstream fixes get undone by the upstream cause.
Three rules make the difference between a diagnosis and an improvement.
- Read for the earliest gap, not the lowest score. The lowest score is your headline; the earliest significant gap is your lever. Start upstream even when the pain is downstream.
- Change one dimension, in order. Attempting all five at once is the most reliable way to spend a year and arrive with the same architecture you started with.
- Re-read afterwards. Downstream dimensions frequently improve without being touched, because they were never independently broken. That is your evidence that you worked on the right one.
It is worth being clear about what “getting it right” means here, because it is not perfection in five places at once. A dimension is sound when it can carry the weight your strategy puts on it. Scaling with clarity is simply the state of a business where each of the five is designed to hold what the next one places on it — which is why clarity at scale is a structural property rather than a state of mind, and why it cannot be restored by working harder inside a structure that was never designed to take the load.
The Architecture Gap Scorecard scores all five dimensions in about ten minutes and returns both readings: your weakest dimension, and the earliest significant gap in your Load Order. Most founders find those are not the same dimension, and that the second one is the more useful answer.
If you would rather work through it in conversation than in a questionnaire, a Discovery Conversation does the same job with more nuance and no scoring.
Which of the five did you name first — and is it the earliest, or just the loudest?


