Company organisational structure: how to design it to grow without slowing down decisions

When a company with capable people keeps waiting for its managing director on almost everything, the reason lies entirely in its organisational structure: the system that distributes work, responsibility, authority and information. Designing it means establishing who decides, with what resources and within which boundaries. The company’s task is to create the conditions for it to be properly structured.

The signal usually arrives before the numbers do. A meeting postponed because one person is missing, a quote stalled awaiting a discount. Inside the company, everyone can show they did their part. Meanwhile, responsibility for the result remains suspended.

An effective structure makes the relationships between strategy, people and decisions understandable. It allows the leadership to oversee the choices that truly commit the business, and teams to resolve those close to their own work. In this guide, we will therefore start from how the organisation actually works, move through the elements that support it, and arrive at a method for checking whether the new set-up holds or not.

What a company organisational structure is

A company’s organisational structure defines how a business divides up activities, allocates responsibilities, assigns authority and coordinates the work needed to achieve its objectives. It includes formal roles, relationships between functions, decision-making rules and the mechanisms through which information reaches the people who need to use it.

We can examine it through four questions. Who is accountable for the result? Who controls the resources? Who can choose between alternatives? Who steps in when an exception arises? If the answers change depending on who you ask, the structure has a zone of ambiguity. That zone is usually filled with personal negotiations, informal requests and additional approvals.

Size matters, but it explains only part of the complexity. A relatively small company, operating in several markets with very different products, can have demanding coordination needs. A larger organisation, focused on repeatable activities, can function with a few stable rules. It is therefore worth looking at the variety of decisions, the frequency of unforeseen events and the number of dependencies between people, as well as total headcount.

The aim is to make the way the company operates less dependent on individual interpretation. A new joiner should be able to understand where their autonomy ends and how to obtain what they need. A manager should be able to explain their area without constantly having to add: ‘In theory it works like this, but in practice we ask the owner’.

When growth outpaces the organisation

In the early stages, the founder can hold together clients, product, priorities and people thanks to direct knowledge of the work. As volumes grow, this capacity meets a physical limit: available time. The business begins to accumulate a queue of decisions, even when it has already hired managers precisely to distribute them.

Waiting becomes a cost that can translate into lost commercial opportunities, urgent purchases, rework or experienced people tied up retrieving information. It is sometimes interpreted as a problem of attitude. But before asking for more initiative, it is worth checking whether those who should take it actually have the authority, data and budget.

Some signs to look for:

• The same decisions return to the top even after being delegated.
• Problems between functions are resolved mainly through personal relationships.
• Priorities change without timelines and resources being renegotiated.
• An ordinary absence interrupts activities that should be transferable.
• Meetings multiply, while it remains unclear who should decide.

Of course, a single episode may stem from a particular situation. It is the repetition of the same pattern, however, that points to a design problem. Understanding it requires recent examples: which decisions stalled? For how long? And at which step?

The diagnosis becomes more robust when facts, rather than impressions, are reconstructed.

From the organisation chart to how things actually work

The organisational chart makes roles, levels and reporting lines visible. It is a useful reference when you need to understand whom a function reports to or how organisational units are distributed. To assess the structure, you then need to follow the path decisions take within that design.

Take a non-standard commercial proposal. The sales manager may formally own the client, while the discount depends on the CEO, the timing on production and the payment terms on finance. The problem arises when these functions are consulted without a sequence, a deadline and a final decision-maker. The hierarchy on paper remains legible; the work stalls all the same.

A useful check is to set three concrete cases alongside the organisation chart: a complex sale, an investment and a quality problem. For each one, note the participants, the information required, the approval steps and the point of closure. The differences between the intended path and the actual path show where to intervene.

The chart should be updated once these relationships have been clarified. Changing the boxes first risks giving a new shape to the same obstacles. The in-depth piece on the organisation chart remains the place to choose the representation; here the concern is consistency between what the drawing promises and what people can actually do every day.

Job description: assigning the expected result

A job description helps translate an organisational position into an understandable assignment. Its usefulness grows when it specifies the purpose of the role, the expected results, the main relationships and the limits of authority. A list of skills, on its own, leaves the most important question open: what is this person accountable for?

For a marketing manager, for example, responsibility may cover the quality of the demand generated, control of the budget and the consistency of the positioning. It must also be clear, however, which commercial choices influence those results and what information must come from sales. Assigning an objective without defining the conditions needed to pursue it creates a responsibility that is difficult to exercise.

The document should be discussed with the person in the role and with adjacent functions. That is often where overlaps surface: two people convinced they approve the same work, or no one authorised to sign it off. The discussion makes explicit relationships that, until then, existed only in habit.

A review becomes appropriate when strategy, scope or responsibilities change. A description updated during recruitment and then abandoned gradually loses its fit with the work. Using it in regular reviews, by contrast, makes it possible to check how the role has evolved and what resources it needs to function.

Task manual: giving continuity to activities

The company job manual goes into the detail of tasks and helps make work transferable. It is particularly useful where errors, absences or handovers can interrupt a service. Its organisational function is to preserve operational knowledge that would otherwise remain in the memory of a few people.

The level of detail should follow the risk. A repetitive, sensitive procedure may require precise steps, checks and documents to be produced. For creative or negotiation work it is better to define objectives, constraints and quality criteria, leaving room for professional judgement. Standardising every action can make a document bulky and little used.

A good test question is this: could a competent colleague, with access to the necessary tools, carry on the work using these instructions? If credentials, references, examples or criteria for recognising an anomaly are missing, the documentation needs completing. Continuity depends on it being genuinely usable.

The job description and the duties manual can thus serve complementary purposes. The former clarifies the role’s contribution; the latter supports the execution of tasks. Keeping them consistent reduces the risk of giving someone broad responsibilities alongside instructions designed for a role that has since changed.

Delegated authority and decision-making powers: establishing who can decide

Organisational delegation translates authority into recognisable decisions. “You handle it” leaves room for very different interpretations. Workable delegation specifies the subject of the decision, the financial or operational limit, the information required and the cases in which a higher level must be involved. Any signing powers also require proper formal oversight.

To begin, list the recurring decisions that absorb time: discounts, purchases, suppliers, hiring, project priorities. Each is assigned an owner of the decision, and the people to consult are distinguished from those to inform. Having a say on an issue and holding a right of approval have different consequences for the process.

Thresholds should be built on the company’s situation. An amount that is manageable for one function may be significant for another; some decisions require central oversight even at limited cost, because they involve reputation, safety or contractual obligations. The nature of the risk therefore matters too.

A register of delegated authority should include a review date and a deputy for absences. After a few weeks, you can check which decisions still made their way up to the top. If escalation remains frequent, you need to understand whether the threshold is too narrow, trust is lacking or the available information is insufficient. Organisational delegation works when this distinction becomes applicable to the cases people encounter every day.

Cross-functional processes: governing hand-offs between departments

Organisational difficulties often concentrate at handovers. Sales promises a delivery, production redefines priorities, administration waits for a document. Each function has understandable reasons, but the customer experiences the overall effect. Designing cross-functional processes starts from that end result and reconstructs the path needed to achieve it.

For each handover it is useful to specify what is transferred, in what format, by when and against which completeness criteria. An order that has formally been sent may be unusable if an essential piece of data is missing. Measuring only the moment it is sent therefore hides part of the problem: one needs to observe when the recipient can actually start work.

The process owner oversees how the process works as a whole and makes obstacles visible. Their relationship with function heads must be clarified, especially when a choice has to be made between competing needs. Without this authority, they risk becoming someone tasked with chasing colleagues whose priorities they cannot influence.

The initial map can be simple: starting point, expected outcome, steps, responsibilities and main exceptions. Documenting every variant before verifying the frequent ones can slow the work down needlessly. Cross-functional process mapping serves precisely to identify the dependencies that the hierarchy alone leaves uncovered.

Autonomy and responsibility: making flexibility sustainable

Organisational autonomy concerns the freedom to choose how to carry out work within shared objectives and boundaries. It requires clear agreements on deliverables, required availability, quality and the handling of the unexpected. Flexibility becomes sustainable when people can organise themselves without passing uncertainty on to colleagues.

In a project-based team, this may mean agreeing on check-in points, response times and interim deadlines, while leaving freedom over hours of individual focused work. In a service requiring continuous cover, presence must instead be coordinated with the needs of the service. The arrangement must fit the nature of the work and the applicable agreements.

Management is party to the same pact. It must provide understandable priorities, avoid mutually incompatible requests and make it possible to flag overload before it turns into delay. An objective assigned alongside five other urgent matters, with no decision on precedence, leaves the individual with an unresolved organisational problem.

To assess autonomy, it is worth looking at reliability of delivery, the ability to anticipate obstacles and the quality of coordination. Mere availability says little about these aspects. The topic deserves dedicated attention because it brings together organisation, corporate culture and trust: three dimensions that directly affect the ability to work well together.

Onboarding: bringing people into the system

Onboarding shows with particular clarity how well an organisation can explain itself. Newcomers must understand their remit, find tools and information, know their points of contact and learn how decisions are made. When these elements are missing, integration depends on the occasional availability of colleagues.

An initial pathway can distinguish access, orientation and progressive autonomy. The first phase requires tools and working conditions; the second, examples, shadowing and knowledge of processes; the third, decisions the person can take with an increasing level of responsibility. The duration depends on the role and the complexity of the work.

It is useful to assign a point of contact and set concrete checks: which tasks can be carried out without assistance? Which questions keep recurring? Where does the available material not match practice? The answers help both the individual and the organisation, because they bring to light knowledge that has remained implicit.

The same principle of continuity applies to changes of role and to departures from the company. Handing over documents, transferring relationships and managing access prevent every transition from becoming an emergency. Onboarding and handovers therefore belong to the ordinary functioning of the structure, even when they are handled at different stages of the employment relationship.

Change management: supporting the adoption of the new structure

An organisational decision begins to take effect when it changes repeated behaviours. Change management helps design this transition: understanding who is involved, which habits must change, which capabilities are needed and how to monitor adoption over time.

The initial communication should explain the problem being addressed and its practical consequences. If the person responsible for a decision changes, it must be made clear what happens from the following day, how work already under way is handled and whom to contact in case of doubt. A general presentation often leaves precisely these points unresolved.

Resistance can contain useful information. A person may object because they are losing a habit, or because they see an operational risk the project has overlooked. Telling the two apart requires listening and verification. Labelling every difficulty as a lack of openness to change impoverishes the diagnosis.

Leadership must then respect the new system. If it continues to authorise directly requests that should follow another route, it teaches the organisation that the old path remains the most effective. Adoption takes hold when stated rules and observable behaviour converge, with explicit room to correct what the first period of application reveals.

Corporate reorganisation: deciding what to change and when

Corporate reorganisation means revising the structure in response to specific needs: growth, acquisitions, new markets, overlaps or changes to the business model. Before acting, it is worth establishing which results you want to achieve and which capabilities must be preserved during the transition.

Moving people and responsibilities also changes relationships built over time. A client may lose their point of contact, a team its point of reference, a function its informal access to important information. The plan must make these dependencies visible, so that they can be transferred in an orderly way rather than their value being discovered after the disruption.

Sequence matters. Some changes can be piloted in one area; others need a common date because they involve systems, budgets or responsibilities that cannot be separated. Any period of parallel running must be bounded: if it lasts too long, people may choose the most convenient route case by case and make the set-up even harder to read.

Reorganisation and change management meet, while addressing different questions. The first defines the structure to be built and the operational transition; the second supports people through adoption. Keeping these tasks separate helps to assign clear responsibilities and to avoid a design problem being left to internal communications alone.

Corporate culture and internal branding: making the criteria credible

Corporate culture shows itself in the decisions that are rewarded, tolerated or corrected. Internal branding connects the identity the business declares with the experience of the people who represent it. For the organisation, the practical point is to check whether the company’s promise is supported by the rules of work.

A company that values initiative should make it possible to propose alternatives and experiment within understandable limits. One that promises quality must provide for consistent timescales, skills and controls. When incentives systematically push in the opposite direction, people quickly learn which criterion really prevails.

This also applies to how managers behave. How are mistakes dealt with? What happens to those who flag a risk? Does useful but less visible collaboration receive recognition? These are organisational questions, because they influence the flow of information and the willingness to take on responsibility.

Culture can reduce the need for approvals when it offers shared criteria for handling new cases. It does, however, require examples, discussion and continuity. General statements gain meaning through recognisable situations. Documenting a few well-handled decisions turns them into learning material, without claiming to capture every behaviour in a procedure.

Which organisational model to choose

Organisational models help distribute complexity. A functional structure groups similar skills; a divisional one gives greater autonomy to products, markets or territories; a matrix coordinates several dimensions at once. A process-based approach puts the flows that cut across functions at the centre. Actual configurations may combine these criteria.

The choice should start from the prevailing type of interdependence. If competitiveness depends on depth of expertise, functional specialisation can offer advantages. If different markets require fast, specific responses, greater divisional autonomy can help. If a project needs several specialist skills, coordination between functional and project responsibilities has to be designed.

Every configuration carries a cost. Specialisation can widen the distance between departments; divisional autonomy can duplicate resources; a matrix requires clear criteria for resolving conflicting priorities. The model should be assessed alongside the available management capabilities and the information systems that will have to support it.

A practical criterion is to compare two options using the same operational cases. Who decides on an urgent request? How is a shared resource allocated? Where is a conflict between margin and service resolved? The most promising configuration is the one that makes these answers sustainable in the company’s context, with explicit costs and trade-offs.

How to design a company’s organisational structure, in six steps

A programme of work can begin with one important process and widen after the first review. This gradual approach makes it possible to observe the effects and correct assumptions before extending the new structure. The method below is a working proposal, to be adapted to the size and needs of the organisation.

1. Clarify the strategic priority. Reducing response times, supporting a new market or improving margin calls for different organisational choices. The priority must be concrete enough to guide trade-offs.
2. Reconstruct the actual work. Examine recent cases, documents and handovers. Interview both those who make the decision and those waiting on it. Note where the accounts diverge and verify the facts before choosing a solution.
3. Assign outcomes and authority. Define owners, decisions, resources, thresholds and relationships. For each objective, check that the person in charge can influence the main variables and knows how to manage dependencies outside their own role.
4. Align tools and information. Make data, documents, budgets and the status of activities accessible. Formal authorisation is of little use when exercising it means asking the person at the top, every time, for information only they hold.
5. Test the new set-up. Apply it to a defined set of activities, with a start date and an agreed review. Record exceptions, delays and decisions that were later corrected. Distinguish design problems from learning problems.
6. Stabilise and update. Transfer what has worked into role documents, processes and onboarding. Make someone responsible for keeping them up to date. The structure continues to evolve together with the work it must make possible.

The outcome of this process should be usable: a legible map, a few clear decision-making agreements and accessible information. Highly detailed documentation that no one can consult when it is needed adds a further step to the original problem.

Before the trial, it is also useful to agree on the conditions for stopping it. If the new process exposes the company to an unforeseen risk, who can suspend it? How do you avoid leaving activities without an owner while corrections are made? Settling these answers makes it easier to try out a change, because people know how difficulties will be handled. An organisational project can thus be ambitious while keeping its scope of application under control.

The review should also involve those who receive the output of the process. If order management is changed, for example, the view of the sales function alone offers a partial reading. Production, finance and customer service may observe different consequences. Gathering this feedback through consistent questions shows whether a local improvement has simply shifted work elsewhere. It is a particularly useful check when functional indicators show progress while the customer’s overall experience remains unchanged.

At the end of the trial, it is worth distinguishing three outcomes: rules to keep, elements to correct and issues still open. The latter need an owner and a next review. Organisational decisions become more credible when the company can recognise what it has learnt and what it still needs to understand. The team then sees that flagging a difficulty produces a response, and that agreements can evolve on the basis of real work.

Organisational KPIs: understanding whether the system is improving

The link with company KPIs serves to verify the effects of the set-up. Measures should be chosen according to the problem being addressed, defined with stable criteria and read together with the quality of results. Speeding up a decision that generates constant corrections can push the cost further down the process.

• Decision time: the interval between a complete request and the communicated decision.
• Escalation rate: decisions escalated beyond the intended level as a share of the total observed.
• Rework: activities to be repeated because of missing information, errors or changed priorities.
• Dependence on key people: activities that stop when a specific point of contact is absent.
• Delivery reliability: commitments met against those agreed in the same period.

Every indicator requires a denominator and a collection rule. If time is counted from the first request in one department and from the availability of all the data in another, the comparison loses meaning. It is advisable to start from a baseline, observe a comparable period and note changes in volume or complexity.

The numbers support the discussion without replacing the reading of individual cases. A rise in escalations could signal difficulties, or better surfacing of previously hidden risks. Management should ask which decisions have changed and why, using the measures as a way into the reality of the work.

A concrete test before changing the whole company

Imagine a services business in which every bespoke proposal has to go through the founder. This is an illustrative example, with no data relating to Bliss clients. The diagnosis shows that the real doubt concerns three aspects: minimum margin, available capacity and non-standard terms. Leadership is involved because no one holds all of this information at once.

The first trial can give the sales manager the decision within agreed criteria, have Operations verify capacity and reserve defined exceptions for the founder. A response time is agreed for complete requests. The team logs cases that exceed the limits and any corrections that emerge after acceptance.

At review, a shorter average time would be a positive signal only alongside sustainable margins, promises kept and manageable workloads. If sales speeds up by passing problems on to production, the design needs revisiting. If exceptions always concern the same point, perhaps that rule no longer reflects the real work.

This kind of test makes the discussion less abstract. People discuss an observable process, can point to where it jams and take part in correcting it. Any extension to other processes builds on what has been learnt, keeping the differences between activities visible.

Three errors to look for during the review

The first mistake is to assign incompatible objectives without a criterion for priority. Sales is assessed on revenue, production on consistency and finance on collections; an urgent order with special terms can set all three functions against one another. What is needed is a rule for assessing the company’s overall interest and a person authorised to close the decision. Otherwise each department rightly defends its own indicator, while the shared result is left to negotiation.

The second mistake is confusing involvement with mandatory consensus. Consulting those with relevant expertise improves a decision. Requiring everyone’s approval can leave it undefined, especially when nobody knows the deadline for stepping in. A simple agreement distinguishes who provides information, who gives a binding opinion on a specific risk and who decides. The distinction should be explained before conflict arises: in an emergency, every ambiguity becomes more costly.

The third mistake is designing around an exceptional individual. Some managers manage to make fragile systems work thanks to their memory, experience and relationships. Their effectiveness can conceal a continuity problem. To identify it, we try asking which activities would come to a halt during a prolonged absence and which knowledge cannot be found elsewhere. Documentation, shadowing and planned cover reduce this dependency without diminishing the professional value of those who have held the work together.

These checks can feed into a periodic review among managers, drawing on a few well-documented episodes. It is also worth recording the cases that went well: they show which agreements allowed people to work together and which information made a timely decision possible. The review thus takes on a learning function as well as a control one. Participants should leave knowing which rule is confirmed, which one changes and who will follow up the change. Without this kind of closure, the meeting risks becoming yet another place to describe the same obstacles.

The role of senior leadership remains decisive

Distributing authority requires the leadership to clarify strategy, constraints and control criteria. Some decisions still require central oversight because of their impact on capital, exposure to risk or difficulty of reversal. Others can be assigned to those with the information closest to the problem, accompanied by appropriate accountability and checks.

The G20/OECD Principles of Corporate Governance provide a general reference for the distinction between direction, oversight and management responsibility. Their application must be read in the context of the business and the relevant regulatory framework; an organisational project also requires the necessary professional checks on its formal aspects.

For a director, the final question is very concrete: which decisions will the company be able to make well even in my absence? The answer shows how much capability has been built into people, processes and information. That is where growth starts to become sustainable: when the leadership’s expertise helps others decide and work can continue with a clear direction.