Every business has someone who just knows. They know which customer always pays on the twenty-eighth, which supplier quotes in rand but invoices in dollars, and which line on the monthly schedule is wrong every time. Often this person has been there for years. And often, when you watch what they actually do all day, most of it is retyping: moving figures from an email into the accounting package, from a bank statement into a spreadsheet, from a spreadsheet into a report.
This piece looks at why expertise ends up buried under data entry, what it costs the person and the business, and what a practical way out looks like. It is a hypothesis about a common pattern, not a judgement about any particular employee or manager.
How expertise gets trapped
Nobody sets out to waste a capable person. The pattern usually starts with a sensible allocation of work.
- The person who understands it gets it. When a new kind of transaction appears, it goes to whoever understands it best, because they will get it right. Over time, the knowledgeable person collects every awkward transaction type.
- Capture is invisible. Typing a figure takes seconds, so nobody records the hours. Multiplied by hundreds of entries a month, it becomes the largest part of the job without ever appearing as a task.
- Systems do not share information. If two systems cannot exchange a record, a person becomes the link. The hidden cost of apps that do not share data is paid in exactly this way.
- Judgement is bundled with typing. Because the expert is already entering the data, they also notice errors, which makes the arrangement look like a control rather than a waste.
That last point is the hardest to see. An experienced administrator catching a wrong supplier code while retyping an invoice is valuable. But the same catch could be made by a simple rule, a validation check or a review of exceptions, leaving the person free to do something more demanding.
What the research suggests
The UK Health and Safety Executive's management standards identify control, meaning how much say a person has in how they do their work, as one of the main factors affecting wellbeing at work. Its guidance describes good practice as giving people a say over the pace and method of their tasks and the chance to use their skills. A day made up of predetermined capture work offers very little of either.
The idea of deskilling, in the broader literature on work, describes the way tasks can be broken down and simplified until the skill of the person doing them is no longer needed or used. Data entry is a textbook case: the description of a data entry role is, by design, one where accuracy and speed matter and discretion does not. When a person with years of experience is placed in that role by default, the organisation is treating a rich resource as a simple one.
The opposite concept is job enrichment, which describes redesigning a role to give people more variety, responsibility and a visible whole piece of work. It is not a perk; it is a way of using skill that already exists. Neither source quantifies the effect in a South African small business, and this article does not claim that it can be quantified, only that the direction is well understood.
An illustrative case
Consider an illustrative example, not a real case: a wholesale distributor of building materials in Bloemfontein. The senior accounts clerk, who has worked there for eleven years, knows every contractor on the books. She knows who disputes freight charges, who needs a statement before a site meeting, and which orders are likely to be returned.
Most of her day goes on capturing. Delivery notes come in as photographs on WhatsApp and are typed into the order system. Supplier invoices arrive as PDFs and are keyed into the accounting package. At month-end, she rebuilds the debtors list by copying balances into a spreadsheet. A hurried month sees her working until eight in the evening.
The owner would be surprised if asked whether she is valued, because the answer is clearly yes. But what the business asks of her is typing, not knowledge. The credit-control calls that her knowledge would make most effective happen in the last hour of the day, if at all. When she suggests a change, there is no time to try it. The question the business never asks is what she would do with her time if the capture were not hers.
The human tax: autonomy and dignity
The first cost is autonomy. A day made of capture work is a day dictated by incoming paper: whatever arrives sets the agenda. The person has little say over what they work on, and almost none over how, because the method is fixed by the form. The article on why the person closest to the problem cannot fix it describes the frustration that follows when someone sees the improvement but has no time or permission to make it.
The second is dignity. Being asked to do routine work well below one's ability sends a message, however unintended, about how the business sees them. People tend to notice when the system uses only a fraction of what they can do. Some respond by becoming quietly disengaged; others leave, taking the knowledge with them. When that happens the business discovers how much of its operation lived in that one head, which is the situation explored in what happens when the only person who understands the system leaves.
What could realistically change
The goal is not to remove the person from the numbers. It is to remove the retyping and keep the judgement.
- Measure the typing. For two weeks, ask the person to note how long they spend on entry, on checking and on tasks that need their expertise. The result is usually a surprise to everyone, including them. The method is described in how to find the manual work nobody is measuring.
- Remove the easiest duplicate first. Pick one flow, such as supplier invoices or bank statement lines, where the same information is entered twice. Importing, forwarding to a capture tool or a bank feed can often eliminate it without a large project.
- Turn the person's checks into rules. Ask what they look for when they key in an invoice. Write those checks down as validation rules or an exceptions list, so the system flags the unusual cases and the person reviews only those.
- Give them the next job. Decide what the freed time is for: credit control, supplier negotiation, improving the month-end process, or training a junior. Make it explicit, or the hours will be absorbed by more capture.
- Keep a human check where it matters. Where a payment of a significant size is involved, keep a second person's approval, as the discussion of reducing human error without removing human judgment recommends.
POPIA is worth a line here. Where captured records contain customers' personal information, moving it between more systems or inboxes has its own obligations, so any automated route should be agreed with whoever is responsible for protecting that information.
The business outcome
The business gets faster, cleaner records and, more importantly, uses an asset it already pays for. Debtors are chased earlier, errors are caught by rules rather than late-night vigilance, and the knowledge that lived in one head becomes partly written down. The business also becomes less dependent on one person's stamina at month-end.
The human dividend
The dividend is autonomy and optionality. Autonomy means that the person's judgement is used and trusted, and that they have a say in how the work is done. Optionality means that the role can grow. Someone who spends their time on analysis and exceptions has more routes open to them, inside the business or outside it, than someone whose role is defined by capture.
The trade-offs
- Informal checks disappear. The accidental review that came from retyping must be replaced by explicit rules, or errors will rise before they fall.
- Set-up takes time. The person best placed to design the change is the person with the least spare time.
- Not everyone wants a wider role. Some people prefer predictable work. The aim is to offer the choice, not to impose it.
- Systems change ownership. Once someone is no longer the human link between two systems, the business must decide who looks after the connection.
Why it matters
A knowledgeable employee is among the most valuable and least replaceable assets a small business has. Using that person as a connector between screens is an expensive way to run a business and a poor way to treat someone. Removing the retyping, and deciding on purpose what to do with the time, lets their expertise do the job it was hired for.
Sources and further reading
- Deskilling - Wikipedia (last verified 10 October 2026)
- Data entry clerk - Wikipedia (last verified 10 October 2026)
- Management Standards - Control - HSE (last verified 10 October 2026)
- Job enrichment - Wikipedia (last verified 10 October 2026)
Related Mellorca reading
- The Difference Between Being Informed and Being in Control
- Why Clear Ownership Gives People Room to Decide
- The Manager Who Can't Act Without the Founder