A dashboard aggregates what each owner reports about their own work, which is the one source with a reason to soften.
Why it matters when the plan changes
Dashboards are the main instrument by which leadership knows how a programme is going, and they inherit the distortion of every reporting line they aggregate. Each owner softens slightly, reasonably, and the softening compounds. Reliance runs high within a reporting line and low across it, so the version that reaches the top is strongest exactly where independent verification is weakest. What arrives is a picture nobody falsified and nobody on the ground would recognise.
The tension is that the dashboard's virtues create the problem. Comparability across workstreams requires a common scale, and a three-state scale has nowhere to put an unknown, unlike a fourth state built to hold insufficient evidence. Timeliness requires self-report, because independent verification is too slow to run every period. Forecasts that carry explicit uncertainty and are later scored improve over time; a dashboard with no such mechanism cannot learn from its own misses.
In practice
A programme reports fourteen workstreams, eleven green. Two of the eleven have had no substantive update in six weeks because the owner changed role. They are green because nothing has been reported against them, and they read identically to the nine that were genuinely assessed.
Evidence
Reliance is high within a reporting line and low across it, so aggregated self-report is strongest exactly where verification is weakest.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)Forecasts that carry explicit uncertainty and are scored later improve, which requires somewhere to record an unknown.
Superforecasting and the Good Judgment Project (2026)
What it cannot tell you
A transformation dashboard cannot distinguish a workstream that is genuinely stable from one that has simply gone unassessed; both report the same colour. It shows what owners say about their own areas, not what independent verification would find, and a three-state scale has no place to record an unknown.
Questions
Because each reporting step softens slightly and a three-state scale has nowhere to record thin evidence. Sull, Homkes and Sull found that 84% of managers say they can rely on their boss and direct reports all or most of the time, so amber carries a cost and green becomes the default until something undeniable happens.
A fourth state for insufficient evidence, a confidence on each rating, a named owner and a date beside each colour, and some proportion of entries verified independently rather than aggregated. Each is cheap and each reduces the comfort the instrument currently provides.
Not entirely, and it does not have to be the only input. The Good Judgment Project's approach, where predictions are scored later using brier scores, shows that forecasts carrying explicit uncertainty improve; sampling a few workstreams independently each period has a similar effect on how owners report.
Decisions. It should surface what has to be decided and by whom, rather than only reporting what state each thing is in. A report that generates an hour of discussion and no decision has consumed programme capacity in order to describe the programme to itself.
As a question. A workstream reporting the same colour for months is either genuinely stable or not being assessed, and the two look identical. Asking which, for the oldest unchanged entries, is usually the highest-yield question in a steering meeting.