Articles about using data, measurement, and uncertainty to make better decisions.
21 articles
A skills pilot does not need to demonstrate the final impact of the entire initiative immediately. Before it is scaled, the pilot needs to produce enough evidence to show whether the process is understood, can be carried out, yields consistent information, and improves real decisions.
The most productive contribution of How to Measure Anything is not the promise of a number for everything. It is the demand that we specify which uncertainty we want to reduce before we begin collecting data.
A dashboard shows the same result for two business areas: 72% meet an expectation. The apparent equivalence vanishes once we learn that each percentage was built from different people, time periods, and sources.
A dashboard can display 73.4 and still fail to explain what was observed, how the evidence was combined, or how much the conclusion could change. Visual exactness does not guarantee conceptual precision.
Noise brings an uncomfortable idea to talent decisions: two competent assessors can apply the same criteria and reach different conclusions for reasons the system never deliberately chose to accept.
Missing evidence, conflicting evidence, and uncertainty are not the same problem. When they are collapsed into a generic category, an organization may manufacture certainty, defer a decision indefinitely, or treat something it has not yet observed as a gap.
An average can accurately represent the center of several results while concealing the difference that matters most to the decision. Its usefulness depends on which components it combines and which of them may compensate for one another.
More measurement can broaden understanding while making a decision harder. The difference depends on whether each signal has an explicit function, a defined audience, and a recognizable use.
Two people can observe the same behavior and reach different conclusions. The usual response is to ask which one of them got it wrong.