Introduction

An SME increases sales, hires people, opens regional offices, launches products, creates departments and reporting lines. The company has grown. But decisions have slowed. The founder approves almost everything. Managers spend more time in meetings. Staff feed multiple systems. Data lives in spreadsheets, WhatsApp threads and processes nobody can explain. The company has more people, more technology, more information, more controls. But does it have more capacity? This is where what I call the Complexity Trap begins.

The thesis is straightforward: companies, like societies, need complexity to grow. The trap springs when complexity grows faster than the capacity it creates. For African SMEs seeking scale, this question deserves as much attention as sales, financing or technology. It is not enough to ask how much to grow.

In this article I invite readers to reflect — because it is not enough to ask: How much can we grow? We must also ask: How much complexity can we manage?

This article complements two earlier pieces in this column: “AfroEntrepreneurs? Between Necessity and Scale” and “Strategic Dilemmas in Management: Choices, Risks and Uncertainties (I/II)”.

“The real challenge is not to grow. It is to grow without creating unnecessary complexity.”

When the solution begins to become the problem

In The Collapse of Complex Societies, Joseph Tainter explains why sophisticated societies decline. His thesis is not that complexity is bad: societies become complex to solve problems, building structures and coordination mechanisms. In the early stages, it works. More complexity generates more capacity.

The problem emerges when each new layer of complexity begins to yield diminishing returns — what Tainter calls the declining marginal returns of complexity. The system invests more to obtain less additional capacity. Now substitute “society” with “company”. The first formal process may improve operations. A new manager may free up the founder. An integrated system may improve information flows. A financial control may prevent losses. Up to this point, complexity is working for the business.

But we keep adding to it. A problem arises. We create a process. The process fails. We add a control. The control requires a report. The report requires data. We buy a tool. And before long, we have created solutions to manage other solutions. Complexity begins to feed on itself.

Progressive Complexification: nobody decides to make the company bureaucratic

Management teams rarely decide: “Let us make this company heavier.” It happens gradually. This is what we might call Progressive Complexification: the slow, almost invisible accumulation of complexity.

One more approval. One more report. One more meeting. One more product. One more system. One more person between the problem and the decision. Each individual decision may make sense in isolation. The problem lies in the cumulative effect. Consider a Mozambican distribution SME that expands into four provinces and triples its turnover. The founder still approves purchases, discounts and payments. To strengthen control, spreadsheets, reports and meetings multiply — and then an ERP system arrives. But the spreadsheet does not disappear. Neither does WhatsApp. Neither do the old forms. Now the company runs on ERP plus spreadsheet plus WhatsApp plus paper. It has digitalised without simplifying. The business grew. Decision-making did not.

When the founder becomes the bottleneck

In the early stages, centralisation can be an advantage. The founder knows customers, suppliers, costs and people. Decisions come fast and cash is protected. But a structure suited to fifteen people rarely serves an organisation of one hundred and fifty. At a certain point, the founder can become the single greatest bottleneck to growth. The company scales; the operating system remains one individual.

Sound professionalisation replaces personal dependency with simple systems. Poor professionalisation layers bureaucracy on top of the dependency that already existed: directors, processes and controls are created, yet the founder continues to make the decisions. The result is bureaucracy and centralisation simultaneously. This is why some African SMEs suffer paradoxically from both informality and excess bureaucracy at the same time: unclear responsibilities, multiple approval layers, little delegation, excessive meetings and dispersed information. The answer is not to add more processes.

The Invisible Tax of Complexity

Let us call the Complexity Tax the cost a company pays solely to coordinate and sustain the complexity it has accumulated. This tax rarely appears in the accounts. It is hidden in meetings, in time spent waiting for a signature, in duplicated data, in reports nobody reads, in redundant systems and in low-margin products. This is why analysing only the financial return on investment of a growth decision can be insufficient.

A new branch, a new product or entry into another market may carry a positive ROI. But each decision adds complexity. The question we should ask is: what additional capacity do we receive in exchange for the additional complexity we create? That is the true ROI of Complexity.

Every company needs a Complexity Budget

Capital, talent, time and attention are finite. The capacity to absorb complexity is finite too. I propose calling the Complexity Budget the quantity of complexity an organisation can absorb without compromising execution. This reframes how we think about growth. Two excellent opportunities can together constitute a poor strategy when pursued simultaneously.

Consider an SME that plans, within the same year, to enter a new country, launch three products, implement an ERP system and open a manufacturing unit. Each project may have a sound business case. But all compete for the same managers, talent, capital and attention. The company may have the money to fund them. It may not have the capacity to absorb them. Before approving a new opportunity, the question should be: do we have the capacity to absorb the complexity this decision will create?

We need a Complexity Audit

We measure sales. We measure margins. We audit accounts. We monitor stock. We assess risk. Why do we not also audit complexity? A Complexity Audit should form part of the management discipline of any growing company: at minimum once a year, and after every expansion, acquisition, diversification or digital transformation.

It can begin with seven areas: processes, decision-making, reporting, meetings, technology, portfolio and structure. And with simple questions: how many approvals does a decision require? How long does it take? Which reports are actually used? How many systems are needed? Which products generate almost no margin?

Once the diagnosis is complete, five responses are available: eliminate, simplify, decentralise, integrate and automate. The order matters. Automation should not be the first response. Before automating a complexity, ask whether it should exist at all. Otherwise, we are merely making a bad process faster.

AI: solution or new complexity?

This question has become even more relevant with artificial intelligence. The promise is substantial: higher productivity, less administrative burden and better decisions. But none of this is automatic. A company can end up with multiple AI agents, copilots, vendors, databases and integrations. Each tool solves one problem. Together they may create another: more complexity. This is precisely what happened with digital transformation — we digitalised processes without asking whether they made sense. We now risk repeating the pattern by automating them.

Automating complexity is not eliminating complexity. AI should only receive credit when it increases the net productivity of the organisation.

Complexity has no right of tenure

Creating complexity is easy. Eliminating it is hard. A report is born after a crisis and survives for years after. A committee is created as a temporary measure and becomes permanent. The risk disappears. The control remains. New processes, committees, reports, controls and systems should therefore carry an expiry clause: every new layer of complexity should have an owner, a cost, an expected benefit and a review date.

The rule is simple: complexity should not enjoy automatic permanence. It must periodically justify its existence. Rather than asking “why should we eliminate this process?”, the question should be: “why should it continue?”

From scale-up to smart-scaling

African SMEs need to scale up — to raise productivity, create employment and compete in larger markets. But there is a condition. Scaling up is not enough. What is needed is smart-scaling: growing revenues, customers and operational capacity faster than costs, structure and complexity.

Consider two companies that each double their turnover. The first doubles its headcount, systems and structure. The second increases those resources by just thirty to forty per cent. Both have grown. Only one has truly begun to achieve scale. Simplification is not a return to informality. It is the removal of weight to free up capacity. And here companies hold an advantage over the societies Tainter studied: they can simplify before they are forced to.

Conclusion: growing without crossing the limit

Every growing company accumulates complexity. Some of it is necessary. Without processes, systems, people and controls, scale is rarely achievable.

The trap begins with Progressive Complexification: small layers accumulate without anyone questioning the combined effect. Then the Complexity Tax emerges: time, money and energy are consumed simply to keep the organisational machine running.

Meanwhile, the ROI of Complexity falls. Structure is added, but ever less additional capacity is returned. The company exceeds its Complexity Budget: it has the capital to pursue opportunities, but not the management capacity to absorb them.

At that point, a Complexity Audit ceases to be an interesting exercise and becomes a management discipline.

The objective is not to cut for the sake of cutting. It is to decide what should be eliminated, simplified, decentralised, integrated or automated — and to require complexity to justify its continued existence.

The destination is not to build simple companies. It is to build companies whose complexity is proportional to the value and capacity it creates.

Before the next branch, product, system or management layer, perhaps two questions are worth asking: How much value will this decision create? And how much complexity will we have to carry to create that value?

It is not enough to ask how much to grow. We must ask how much complexity we can manage.

Is your company scaling the business — or merely scaling the complexity?

Note: Joseph Tainter’s analysis of the collapse of complex societies focused comparatively on three cases: the Western Roman Empire, Classic Maya civilisation and the Chacoan system in the present-day south-western United States. | ERP — Enterprise Resource Planning: an integrated management system that consolidates, on a single platform, the principal processes and data of a company — including finance, procurement, sales, inventory, production and human resources — enabling greater coordination, control and information-sharing across all areas. | ROI — Return on Investment.

Source: Diário Económico
Original article: https://www.diarioeconomico.co.mz/?p=552652