Where the bottleneck hides in a Nigerian SME.

Ask a founder where their business slows down and you will get an answer in under a minute. It will be specific, it will be delivered with conviction, and there is a good chance it is describing a symptom two steps downstream of the thing actually costing the money. This is not a failure of self‑knowledge. It is a structural consequence of where a founder stands relative to the work, and it is the reason the operations audit we sell is built the way it is.

What follows is the reasoning behind that method — what we look for, and why those things rather than others. It is a description of an approach, not a report of findings.

Why the named bottleneck is usually the wrong one

Founders see the parts of the business that reach them. What reaches them is whatever is loud — the thing someone escalated, the customer who complained, the week that went badly enough to require a conversation.

Expensive problems are frequently quiet. A step someone absorbed into their routine two years ago and stopped mentioning. A reconciliation that takes four hours because it has always taken four hours. A hand‑off between two people who have worked around each other so smoothly that neither of them files it as friction any more. None of that escalates, so none of it arrives at the founder's desk, so none of it is in the answer you get in the first minute.

Which means the value of an audit is rarely in discovering something nobody knew. It is in locating something several people knew separately, and putting it together in one document with a number attached.

The four questions

The core of the method is embarrassingly simple. Four questions, asked of the same business, and close attention paid not to the answers but to which ones can be answered at all.

• What did you turn over last month? Almost every business can answer this immediately. Money arriving is an event; events get noticed and recorded.
• What did it cost you to deliver that? Much harder. Costs are distributed across time, suppliers, and people, and nothing forces them into one place unless someone decided they should be.
• Which line makes you the most money? Harder again, because it requires the first two answers to be correct at the same level of granularity.
• How long does an order take from first contact to delivered? The one we expect to be answered with a range rather than a number — and where the real distribution is usually wider than the range offered.

The gap between question one and question four is the gap between a business that is running and a business that is understood. Both can be profitable. Only one of them can be improved deliberately, because only one of them can tell you whether a change helped.

Where we expect to find the mess

Three places, and we go looking in them first because in Nigerian SMEs specifically they are where the seams tend to be.

WhatsApp. For a lot of Nigerian businesses the operational truth — what was ordered, what was promised, what was renegotiated after the invoice went out — lives in chat threads on personal phones. The instinct is to treat this as a problem to migrate away from. We think that is wrong. WhatsApp wins because it is genuinely better than the alternatives at what operations needs: everyone already has it, it survives a bad connection, and adding a customer costs nothing. What is broken is not the channel. It is that nothing structured is ever extracted from it, so the business holds a complete record it cannot query.

The join between two systems. Wherever sales meets payments, or payments meets the ledger, there is usually a spreadsheet — and often a well‑built one, maintained by someone thoughtful. The spreadsheet is rarely the problem. The problem is that it is the only place the two systems meet, and the meeting happens inside one person's head once a week.

Growth, not age. The intuition is that older businesses carry more mess. We expect the opposite to matter more: process outrun by volume in the last eighteen months, with nobody having stopped long enough to notice. Age brings accumulated cruft. Growth brings load the cruft cannot carry.

A business that cannot answer a question about itself without three days of archaeology does not have a reporting problem. It has an instrumentation problem, and reporting is just where it shows up.

— The distinction the audit is built to draw

Why the output has to be checkable

An audit that produces adjectives is worthless. Ours produces a written report, a diagram of the system the business actually has rather than the one on the org chart, and a ranked list of automations with an hours‑saved‑per‑month figure against each. The constraint we hold ourselves to is that if a recommendation cannot survive being written down with a number next to it, it does not go in.

That constraint is also why the audit is paid. A recommendation of “change your invoice numbering and do not build anything yet” is a complete, valuable deliverable and a commercial write‑off — which is the argument we make at more length elsewhere.

The one thing worth doing regardless

Make the business queryable before you make it bigger. Not perfect, not automated, not migrated onto anything — just able to answer a question about itself without three days of archaeology. That is not something you buy. It starts with someone deciding that a number nobody can check is not a number.

Growth on top of a business that cannot be queried does not compound. It just makes the archaeology longer. If the ledger side of this is what worries you, the NRS compliance piece is the same argument with a deadline attached.