Everybody, it seems, is urging us to embrace data-driven decision making. The logic is inescapable - it is far better to base decisions on data than on intuition. But data is not the solid empirical basis it is often made out to be. It would be well worth reading R.J. Ackermann, Data, Instrument and Theory, or this discussion from Ian Hacking: "The common image of science remains a modified version of Enlightenment science. The humanities have clung to the Enlightenment image of science as a grand unifying intellectual adventure, one that strives to find the ultimate theory of everything... But science has become as multifloriate as the humanities. It has become the domain in which there can be stability without foundations, sharing without commensurabiity... It has become a domain in which there can be coherent action within a thoroughly disunified world picture." Words of caution well worth heeding.