MELLORCATalk to Mellorca
Patterns

Reporting & Operational Visibility

When Your Dashboard Is Really a Manually Updated Spreadsheet

A dashboard that is filled by someone pasting figures into a spreadsheet looks live but is not. Here is why that matters for decisions, and how to connect the data or make its age visible.

Mellorca Patterns2026-10-06

Observable condition

The business has a dashboard. It sits on the screen in the boardroom, or it is shared as a link on a Monday morning, and it looks the part: traffic-light indicators, sales by region, a trend line for debtors, a gauge for capacity. Visitors are shown it. Managers refer to it.

Behind it, every Friday afternoon, one person opens three exports, copies figures into a hidden tab of a spreadsheet, checks that the formulas still pick up the right rows, and saves the file in the shared folder. On weeks when that person is on leave or busy with month-end, the dashboard quietly shows last week's numbers. Few people notice, because nothing on the screen says when the figures were last updated.

The dashboard looks live. It is actually a manually maintained report presented in a dashboard's clothes.

Realisation

The realisation is that the format has created a false impression. People treat a dashboard as a current view of the business, because that is what dashboards are for. Microsoft's Power BI documentation describes dashboards as a way to monitor your business and see your most important metrics at a glance, with tiles that update as the underlying data changes. When the underlying data is a spreadsheet updated by hand once a week, that promise is not being kept, and decisions are being made on the assumption that it is.

The gap is not only one of freshness. A manually updated dashboard also depends on one person's routine, carries every copy-and-paste risk of a spreadsheet, and gives no warning when something has gone wrong. The visual polish makes the figures look more authoritative than the process behind them deserves.

Realisation prompt: if you looked at your dashboard right now, could you tell when each number was last updated, and by whom?

Identification

This is a presentation-versus-pipeline mismatch. The visible symptom is a dashboard that sometimes shows stale or inconsistent figures. The structural issue is that the business invested in the display layer but not in the data flow beneath it, so a manual spreadsheet process sits where an automated connection should be.

It is a specific form of the broader risk described in when a spreadsheet becomes a business-critical system. The difference is that here the spreadsheet is hidden behind a screen that suggests otherwise.

Diagnosis

  • The dashboard was built before the data flow. A dashboard tool was bought or a template was adopted, and the quickest way to fill it was to paste figures in. The temporary arrangement became permanent.
  • Sources cannot be connected easily. One or more systems has no convenient connection, so its data arrives by export. Once one source is manual, the whole update becomes a manual ritual.
  • Refresh is a person, not a setting. Excel itself can refresh external data connections when a workbook opens or at a set interval, and Power BI can refresh imported data on a schedule. When those options are not used, freshness depends on someone remembering.
  • No freshness indicator. The dashboard does not show the time of the last update, so stale figures look identical to current ones.
  • Personal ownership. The file lives in one person's folder or workspace. Microsoft's guidance on personal analytics advises limiting the use of personal workspaces and keeping mission-critical content out of them, because storing critical content there represents risk to the organisation.

An illustrative example

The following scenario is illustrative and does not describe a specific organisation. A Pretoria-based security services company displays a contracts dashboard in its operations room. It shows guards deployed, sites covered and overtime hours. The figures come from a rostering system, a payroll export and a client spreadsheet, combined weekly by the operations administrator. During a fortnight when she is ill, the dashboard is not updated. Overtime rises sharply on two sites during that period, but the screen continues to show the earlier, lower figure. The overspend is found only when payroll runs.

Commercial impact

The commercial value at stake is decision speed, risk, management capacity and responsiveness. Decisions made from a dashboard that looks current but is not carry a hidden risk of being wrong in timing or in substance. Overtime, stock, cash and service levels can all move significantly within a week.

There is also a capacity cost. The weekly update ritual is recurring manual work, and it ties the business's view of itself to one person's availability. When the dashboard falls behind, trust erodes, and managers start keeping their own figures, which brings back the very inconsistency the dashboard was meant to remove.

Common misidentification

The usual conclusion is that the dashboard tool is inadequate and a better one is needed. A new tool fed by the same weekly paste will behave the same way. The constraint is the pipeline, not the presentation.

Another misreading is that a weekly figure is good enough, so freshness does not matter. That may be true for some measures. The problem is not weekly data as such; it is weekly data presented as if it were current, with no indication of its age and no protection against a missed update. The question of when a spreadsheet is the right tool is explored in when a spreadsheet is a useful tool and when it becomes infrastructure.

Possibility

In a better state, every figure on the dashboard has a known source and a known refresh time, shown on the screen. Sources that can be connected are connected and refresh on a schedule. Any source that must remain manual is clearly labelled, with a named owner and a backup.

The dashboard then becomes what it appears to be. Readers can trust its timing, and when something is late, the screen says so. Each number can be traced back to where it came from, a problem explored further in why business users cannot trace a dashboard number to its source.

Intervention

  1. Trace every tile. For each figure, record its source system, how it reaches the dashboard and who updates it.
  2. Show data age. Add a visible last-updated time for each source. This alone changes how people read the screen.
  3. Connect what can be connected. Replace manual exports with direct connections and scheduled refresh where the source supports it.
  4. Isolate the manual remainder. For any source that stays manual, define the update routine, the owner, the deputy and a check that warns when the update is missed.
  5. Move the file to shared ownership. Keep business-critical dashboards and their data in a shared, governed location rather than a personal folder.
  6. Match freshness to use. Decide for each measure how current it needs to be, and design the refresh accordingly.

Diagnostic questions

  • Which numbers on your dashboard are typed or pasted in by a person?
  • Does the dashboard show when each figure was last updated?
  • What happens to the dashboard when the person who updates it is away?
  • Which decisions have been taken from the dashboard in the past month, and how current did the data need to be?
  • Could your current tools refresh any of these sources automatically if configured to?

A dashboard is a promise of currency. When a hand-updated spreadsheet sits behind it, the business should either keep that promise by connecting the data or make the real age of each number plain to see.

Sources and further reading

Related Mellorca reading

Article summaryA dashboard that is filled by someone pasting figures into a spreadsheet looks live but is not. Here is why that matters for decisions, and how to connect the data or make its age visible.
Explore our solutionsStart a conversation