This note shows how big data can help combine experimental with non-experimental approaches in impact evaluations when take-up is low. In this study, author have access to a large administrative data set (of 660 MB), which follows the monthly financial indicators of each client for up to 18 months prior to the intervention and 6 months after it. Moreover, from the experimental approach their also had a large pool of clients randomly assigned to the control group. This data enables us to obtain credible estimates by combining the experiment with two non-experimental approaches. Their first use propensity score matching to find, among the clients in the control group, a subset of clients that best mimics the pre-intervention financial trajectories of clients in the treatment group that received treatment. The effects of the workshops on the treated clients are summarized. Under our preferred specification, the author finds that participating in the workshop increases by 11 percentage points the likelihood of paying more than the minimum payment, and reduces by 3.4 percentage points the likelihood of delaying payment. Monthly credit card spending increases by 63.7 percent, and the likelihood of owning a deposit account with our partner bank also increases by 2.7 percentage points. The two financial education interventions help clients reach the minimum payment and pay their bills on time more often, without reducing their credit card spending. Both interventions increase the likelihood that clients are profitable for the bank.