We introduce the notion of a data-first design study which is triggered by the acquisition of real-world data instead of specific stakeholder analysis questions. We propose an adaptation of the design study methodology framework to provide practical guidance and to aid transferability to other data-first design processes. We discuss opportunities and risks by reflecting on two of our own data-first design studies. We review 64 previous design studies and identify 16 of them as edge cases with characteristics that may indicate a data-first design process in action.
Fig. 1. Refined and extended framework for data-first design studies. A new acquire stage is added for both obtaining and abstracting data. \textit{Discover} is moved and renamed to elicit, to emphasize the elicitation of tasks from potential stakeholders. Winnow focuses on analyzing the match between the abstract tasks of these stakeholders and the data abstraction. Nuances differ in the cast and design stages to incorporate the specific characteristics of the data-first approach.
Fig. 2. Simple conceptual model illustrating the challenges of finding an intersection of relevant tasks and data while keeping the problem space constrained. (a) The data and task axes both range from peripheral to core. (b) In the elicit stage, the visualizer examines the match between the initially acquired abstracted data and stakeholder tasks. (c) In the winnow stage, the visualizer can assess the benefits and risks of expanding the scope to additional data to expand the set of tasks and thus stakeholders.
