Most owners of growing businesses know the feeling. The monthly numbers look acceptable, the weekly meeting reports no major issues, and then a large customer complains, a supplier stops deliveries or a project turns out to be weeks behind. The information existed somewhere in the business. It simply did not reach the people who needed it until it was too late to act cheaply.
The usual response is to ask for a report or a dashboard. Sometimes that helps. Often it adds another view of the same incomplete information. Poor visibility is rarely a reporting problem. It is a consequence of how work is recorded, how status travels and where managers stand in relation to the work.
Management cannot see what the business does not record, and it cannot see in time what the business records only after the fact.
Six causes of poor visibility
1. Systems record transactions, not work in progress
Accounting systems record invoices and payments. CRMs record deals won and lost. Most business systems are very good at recording the end of a piece of work and quite poor at recording the middle of it. A quote that has been waiting for technical input for nine days, a delivery held because one item is out of stock, an approval sitting in someone's inbox: none of these appear anywhere until they turn into a late invoice or a lost deal.
The Kanban Guide is useful here because it insists on defining when a work item starts and when it finishes, and on tracking work item age, the elapsed time since a piece of work started. A business that cannot say how long its open items have been open cannot see trouble building up.
2. Status lives in conversations
In many small and mid-sized businesses, the real status of a job is in a WhatsApp group, an email thread or a supervisor's head. Asking "where is this?" triggers a chain of messages. That works while the owner can ask everyone personally. It breaks once the business has more customers, jobs and staff than one person can keep track of. The pattern is described in when a status update requires a staff member to investigate.
3. Reports describe outcomes, not signals
Revenue, margin and debtors are outcome measures. They tell management what has already happened. By the time a month-end report shows a drop in gross margin, the jobs that caused it are finished. Leading signals, such as the number of jobs waiting on materials, the age of open quotes or the volume of rework, show trouble while it can still be prevented. Most businesses report the first kind and rarely the second.
4. Bad news is filtered on the way up
Information passed through layers of people is summarised, softened and delayed at each step. No one needs to be dishonest for this to happen. A supervisor who expects to fix a problem by Friday sees no reason to report it on Tuesday. When the fix fails, the problem reaches management late and already larger. Systems that make status visible directly, without requiring someone to choose to report it, reduce this filtering.
5. Numbers are assembled by hand
If the weekly report needs someone to export from three systems and combine the results in a spreadsheet, it will be late, it will be produced less often than it should be, and it will contain errors. It also depends on one person being available. The cost of that pattern is set out in why one number requires three people and four systems.
6. Managers are far from the work
The Lean Enterprise Institute describes the gemba as the place where value-creating work actually occurs, and argues that real improvement needs direct observation of conditions where the work is done rather than decisions made remotely. As businesses grow, owners naturally move away from the floor, the workshop, the call centre or the site. Reports replace observation. Without a deliberate habit of going to see, management's picture of the business becomes the picture its reports were designed to show.
An illustrative example
Consider an illustrative Cape Town shopfitting business with about 60 staff. Its owner received a monthly management pack showing revenue, gross margin by project and debtors. For most of a year, the pack looked healthy. Then three projects ran over budget in the same quarter.
When the business looked back, the warning signs had been visible on the factory floor for months: rework on joinery had been rising, materials were being ordered late because drawings were approved late, and installers were waiting on site for items that had not been dispatched. None of these appeared in the pack, because none of them was a transaction. The project managers knew. They expected to recover and did not escalate.
The business did not add a dashboard. It added three things: every project was given a visible stage with an age, a short daily list showed items waiting on drawings or materials, and the owner walked the factory and one site each week. Within a quarter, the management meeting was discussing jobs that were at risk, not jobs that had already gone wrong.
What good visibility looks like
The Lean Enterprise Institute defines visual management as placing tools, activities and indicators of performance in plain view so that the status of the system can be understood at a glance by everyone involved. The words that matter are "at a glance" and "everyone". Good visibility does not require an analyst to interpret it, and it is not reserved for management.
In practice that usually means:
- every important piece of work has a recorded state, an owner and an age;
- the states are defined so that "in progress" has a clear start and "done" has a clear finish;
- a small number of leading signals are shown alongside the outcome numbers;
- exceptions, such as anything older than its normal cycle time, are highlighted rather than buried;
- the numbers are produced by the systems, not assembled by a person.
Having data is not the same as having visibility, a distinction explored in the difference between having data and having visibility.
A practical diagnostic
Ask these questions about the three or four processes that matter most to the business:
- Can I see, without asking anyone, how many items are currently in progress and how long each has been open?
- When did I last learn about a problem from a customer before I learned about it from my own team?
- Which of my regular reports would warn me about a problem before it affected revenue or margin?
- How much of my management pack is assembled by hand, and by whom?
- When did I last watch the work being done, rather than read about it?
Answers to the first two questions usually show whether the gap is in recording work in progress or in how status travels. The third shows whether reporting is all lagging. The fourth shows fragility. The fifth shows distance.
Trade-offs
Visibility has a cost. Recording states and ages asks staff to update systems at the moment work moves, and that only happens if the system is part of how the work is done rather than an extra step afterwards. Too many measures also blur the picture. A small set of signals that people trust is worth more than a large set they ignore; more dashboards do not automatically create better decisions.
The goal is not total surveillance of every task. It is that management finds out about the problems that matter while they are still small, from the business's own systems and people, rather than from a customer. The broader cost of not achieving that is covered in what management cannot see usually becomes expensive.
Sources and further reading
- Visual Management - Lean Enterprise Institute (last verified 10 October 2026)
- Gemba - What Does It Mean? - Lean Enterprise Institute (last verified 10 October 2026)
- The Kanban Guide - May 2025 (latest) - Kanban Guides (last verified 10 October 2026)
Related Mellorca reading
- Why Work Gets Lost Between Teams
- Poor Data Is an Operations Problem, Not Just a Data Problem
- How Much Manual Work Is Hiding Inside Your Business?