It’s common for policy briefs and research papers to end with recommendations that flow from the analysis. However, if the recommendations aren’t closely linked to the needs of the people that they’re trying to benefit, the solutions may fail to get political purchase or popular support, hit barriers during implementation, or fall flat with the target audience. On the flip side, if recommendations are too expensive or require big shifts in behavior, they may not be realistic. Human-centered design methods can produce many potential solutions, zero in on the highest potential ones, and test them so recommendations to policymakers are strong. Additionally, many of the methods below involve community members in the creation and testing of ideas so there’s a closer link between policies and the people most affected by them.