Developer experience · Qualitative interview need-finding study; semi-structured interviews with API designers, coded into themes and opportunities for tool support.
Enabling Data-Driven API Design with Community Usage Data: A Need-Finding Study
CHI '20: Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · 2020
Summary
Designers primarily relied on bug reports and peer reviews because formal usability testing was prohibitively expensive. They wanted real-world use cases and users’ mental models, including where users got stuck, workarounds, mistakes, and corner cases; the paper proposes mining and elicitation tools.
For your marketing
Possible founder application
Build feedback loops around issue data, usage traces, and support questions; mine recurring mistakes and workarounds, and use them to prioritize API redesign and documentation instead of relying only on feature requests.
Reported result
What the study found
Designers primarily relied on bug reports and peer reviews because formal usability testing was prohibitively expensive. They wanted real-world use cases and users’ mental models, including where users got stuck, workarounds, mistakes, and corner cases; the paper proposes mining and elicitation tools.
Research question
What the authors investigated
How do API designers gather and interpret user feedback, and what unmet needs exist for data-driven API design tools?
Design
Qualitative interview need-finding study; semi-structured interviews with API designers, coded into themes and opportunities for tool support.
Sample and setting
23 API designers from 6 companies and 11 open-source projects.
Scope and transfer
Limitations to keep in view
Interview-based, self-reported needs rather than a controlled evaluation of a tool; six companies and 11 OSS projects may not represent all API organizations, and the study does not estimate causal effects on API quality.
Research sources
ACM publisher page gives exact bibliographic data, sample, method, and abstract: https://dl.acm.org/doi/10.1145/3313831.3376382 Firecrawl research job 01a14371-854f-71aa-ba3d-455ec397717d returned this citation and metadata.