Indicator design and measurement

Most reporting problems are definition problems that surfaced late. An indicator that seemed obvious at design turns out to mean three different things to three field teams, and by then a year of data has been collected against all three readings.

The work in this section is mostly upfront and unglamorous: deciding what precisely is being counted, at what level of the results chain, from which source, how often, by whom, and disaggregated how. Done properly it produces a definition a new officer can pick up and apply identically in year three. Done poorly it produces numbers that cannot be defended when a funder asks how they were derived — and the answer arrives too late to fix.