The Gap Between What the Dashboard Says and What the Floor Says

Every organisation of meaningful size runs two versions of itself. There is the reported version, which travels upward through the management layers and arrives in the board pack. And there is the observed version, which is what the people doing the work actually experience.

These are never identical. In a healthy organisation the gap is small and stable. The useful insight is that the size of the gap is itself measurable, and that it behaves like a leading indicator.

Why a gap forms at all

Reporting distance. Each layer between the work and the board is a summarisation step, and summarisation discards variance. The exception that would have been informative gets rounded into the average that is not.

Incentive design. People report against what they are measured on. That is the intended function of measurement. The side effect is that anything unmeasured goes unreported, and the unmeasured is where new problems live, because problems are novel before they are recurring.

Rational editing. A manager with a problem they believe they can fix by next quarter will usually fix it rather than escalate it. This is often correct and produces good outcomes. It also means the board sees the problems that could not be quietly solved — a biased sample.

Time. Reported figures describe a closed period. Observed reality is now.

Nobody lies. Every layer rounds. Rounding in one direction, repeated four times, is indistinguishable from a lie by the time it reaches the top.

Treating the gap as the metric

The conventional response is to improve reporting accuracy. This has limits, because the mechanisms above are structural rather than procedural.

The more useful move is to stop trying to close the gap and start measuring it. Run one quantity through both channels and watch the difference.

Pick something both layers can assess — confidence in the quarter's forecast works well. Ask both populations the same question, on the same scale, on the same cadence. Do not reconcile them. Plot the difference.

A stable gap is normal and tells you little. A widening gap is the signal. It means information is being lost or edited on its way up at an increasing rate, and that happens when news is getting worse faster than the reporting culture can absorb it.

Reading the direction

Reported above observed is the common case and the dangerous one. The floor is more pessimistic than the board pack. This precedes misses by roughly the length of your reporting cycle multiplied by the number of layers.

Observed above reported is rarer and usually means operating teams see improvement that has not reached the financials. It is frequently ignored, and it is a chance to lean in earlier than competitors.

A gap that suddenly narrows after widening deserves inspection. Occasionally the problem was resolved. More often the people reporting the discrepancy stopped bothering.

Running it without breaking trust

Aggregate only, with a floor. Never report a result for a group small enough to identify individuals. Set the floor before you start and hold it under pressure.

Ask about the work, not about people. Confidence in a forecast is a question about a forecast. Confidence in a manager is a performance review conducted by ambush.

Publish the aggregate back. The strongest predictor of whether people keep answering honestly is whether they ever see the result.

Act on it visibly once. One documented instance where the gap changed a decision does more for participation than any assurance about anonymity.