We study whether visualisation can be used to access information more efficiently and improve decision making. We test our primary hypotheses that visualisation improves performance in economic choices and can be taught through a randomised control trial with 1967 aspiring entrepreneurs in Colombia. Half our sample receive a ten-session training programme designed to use visualisation in business decision making. To control for the informational content of the programme, the other half of our sample receive either business-as-usual training or assigned to a pure control group. We expect treated individuals to be more likely to use visualisation and with higher quality, resulting in higher performance in choice experiments and better outcomes on average. Furthermore, our sample is characterised as “vulnerable”, with approximately 30% reporting that they have been exposed to armed conflict or migration from the Venezuelan crisis. Consistent with recent findings in neuroscience and psychology that past traumatic experiences make it particularly costly to visualise, we expect individuals previously exposed to trauma to benefit most from the training.