Observable condition
When the business had twelve people in one building, the owner knew how things were going without opening a report. She could see who was busy, hear which customers were unhappy and notice when a job was stuck. Now there are fifty people across three sites, two shifts and a handful of systems, and she finds out about problems when they become large enough to reach her. A branch manager mentions in passing that the backlog has been growing for a month. A customer cancels with a reason that nobody inside the company recognised as a trend.
The business is not failing. It is growing, and growing in the right direction. But the sense of being in control has gone, and the people at the top are making decisions with a feeling that something important is happening out of sight. They respond by asking for more meetings, more reports and more updates, none of which seems to restore the old clarity.
An illustrative example: a Port Elizabeth company that repairs and services industrial pumps. When it had one workshop, the owner knew every job on the floor by sight. After opening a second workshop and a field-service team, the only view of the whole operation is a Monday spreadsheet compiled from three people's notes.
Realisation
The realisation is that visibility in a small business is a by-product of proximity, not a system. It comes free when everyone is within earshot. It does not scale, because it depends on one person's attention and memory, and that attention cannot be spread across more people, places and processes.
As the business grows, three things happen together. The volume of events rises. The distance between those events and the people who need to know about them lengthens. And the number of hand-offs, where information can be lost, increases. Without a deliberate replacement, the business becomes less visible to itself in direct proportion to its success.
This is why many owners describe growth as making the business feel harder to run even when the numbers are good. A part of that feeling is simply a loss of sight.
Identification
This is a visibility-scaling problem: informal observation has stopped working, and no designed alternative has been put in its place. It is related to, but distinct from, growth making the business feel less efficient, which describes the general sensation, and having plenty of data but no single view, which describes a data-consolidation gap. This Pattern focuses on what changes as scale increases and why the old ways of knowing stop working.
The visible symptom is surprise: problems discovered late, by accident or from outsiders. The common assumption is that managers need to communicate more. The structural cause is that the signals which used to reach the owner by proximity are not being captured, summarised or routed anywhere.
Diagnosis
- Signals exist but stay local. The branch knows its backlog. The workshop knows its rework rate. Each is held in a notebook, a spreadsheet or a person's head, and nothing connects them.
- Reporting measures outcomes, not the health of the process. Revenue and margin are reported monthly. Queue length, waiting time, error rate and utilisation, which show trouble forming weeks earlier, are not.
- Averages hide the problem. Google's site reliability guidance, written for software services but instructive for any operation, notes that tracking only an average can mask the cases that matter, and recommends separating successful from failed requests and watching the slowest, not just the typical. A workshop's average turnaround can look steady while one customer category waits three times as long.
- Alerts do not exist. The same guidance says monitoring should tell people when something is broken or about to break, and that a page should be urgent and require human judgement. Most growing businesses have no equivalent: nothing tells a manager that a threshold has been crossed.
- Complexity grows faster than headcount. IBM describes observability as the ability to understand what is happening inside a complex system from the data it emits, and notes that modern systems generate data in volumes and at speeds that periodic sampling cannot keep pace with. A fifty-person business is not a cloud platform, but the principle carries: a system that emits no data about itself cannot be understood from outside it.
- Visibility is treated as a reporting task. It is given to whoever has time on Fridays instead of being designed as part of how work is done.
Commercial impact
The commercial value at stake is scalability, speed, customer experience and risk.
Scalability is the central issue: a business that cannot see itself cannot add sites, services or people without losing control. Speed suffers because trouble is addressed after it has grown. Customer experience declines when issues that used to be caught informally now surface as complaints. Risk grows quietly as backlogs, compliance gaps, ageing receivables and key-person dependencies pile up unobserved. The general pattern, where small problems become expensive as volume rises, is covered in Why Small Operational Problems Become Expensive at Scale.
Common misidentification
The first response is usually more meetings. Meetings do spread information, but they do so slowly and from memory, and they scale poorly: more sites mean longer meetings with less attention per site.
The second is a dashboard of everything. Without a view of what the business needs to see, the result is a wall of numbers that no one can read. The Insight What Operational Visibility Actually Means sets out the difference between displaying data and being able to see the operation.
The third is to hire a layer of supervisors. That may be right in some cases, but it replaces one person's limited view with several, and the underlying information is still carried by word of mouth.
Possibility
A better arrangement treats visibility as something designed into the way work is done. A small number of measures describe the health of each process: how much work is arriving, how long it is waiting, how often it goes wrong, and how close the team is to its capacity. These are captured as a by-product of the work, not compiled afterwards. They are visible to the people who run the process and summarised for the people above them. Thresholds trigger notifications, so that a growing backlog reaches a manager before it reaches a customer.
The four categories above are an adaptation of the "golden signals" that Google's guidance recommends for monitoring a service: latency, traffic, errors and saturation. They translate well to a pump workshop: jobs arriving, days waiting, jobs returned for rework, and the share of workshop hours already committed.
Intervention
- Write down what you used to know by walking around. List the things the owner could previously sense: who is overloaded, which jobs are stuck, which customers are unhappy.
- Turn each into a measure. For each item, decide what could be counted, where it is already recorded and who needs to see it.
- Start with four measures per process. Volume in, time waiting, rate of errors or rework, and load against capacity.
- Capture at the point of work. If a status has to be typed into a separate spreadsheet at the end of the week, it will be late and incomplete. Where a field in the existing system can record it as the work happens, use that.
- Show the spread, not just the average. Include the slowest cases or the oldest open items.
- Set thresholds and recipients. Decide what level of backlog or delay requires a response and who is told.
- Review the set quarterly. Drop measures nobody used and add those the business discovered it needed.
Diagnostic questions
- What did the owner or senior managers know instinctively two years ago that they now find out later?
- Which problems in the last six months were first discovered by a customer or by chance?
- Do you measure how long work waits, or only how much is finished?
- Would you hear about a growing backlog before it affected a customer?
- If the person who holds the whole picture in their head were away for a month, what would the business not know?
Bottom line
Growth removes the informal vision that small businesses rely on. Replacing it with a few well-chosen signals, captured as the work happens and routed to the right people, lets the business keep growing without losing sight of itself.
Sources and further reading
- Google SRE monitoring ditributed system - sre golden signals (last verified 10 October 2026)
- What Is Observability? - IBM (last verified 10 October 2026)
- What is Data Governance? - IBM (last verified 10 October 2026)
- Set data alerts in the Power BI service - Power BI (last verified 10 October 2026)
Related Mellorca reading
- Why Real-Time Operations Still Depend on Friday Reports
- Why Finance Reconciles Information That Systems Already Hold
- When Your Metrics Arrive After the Decision Was Needed