2020 Main Conference Talk
Indeed’s mission is to help people get jobs, and we constantly strive to provide the best search experience. This presentation focuses on job seekers in the healthcare industry and describes how an agile collaboration among UX, Product, and Taxonomy enabled us to make sense of our users’ feedback and incorporate their mental models into the back-end.
Our journey started with the realization that, as Margaret Kelsey from InVision noted, “Design for everyone is design for no one.” Our categories, filters, and our search results were designed for an “average” job seeker, and they did not serve job seekers in the nursing field very well. To understand the nurses’ perspective we conducted in-depth interviews and synthesized their feedback into themes. The results prompted us to rethink how to approach our search experience and provided evidence that in order to fix the front-facing search product, we’d need to work in the back-end, at the taxonomy level.
We already had an existing taxonomy of nurse occupations, but we needed to be more specific. Using qualitative and quantitative approaches to sift through the millions of documents on our site, we developed an initial taxonomy of job attributes that could describe facets like medical specialty, patient type, or shift. We then exposed the job attribute taxonomy to our users, collected more feedback, and iterated. Although creating taxonomies is not usually thought of as an agile practice, we found a way to update our taxonomy by making fast, iterative changes that have a direct impact on our products.
Our talk is targeted at IA practitioners who want to create user-centered taxonomies in an agile fashion, are interested in performing open ended user research, and would like to learn about the communicative and collaborative processes required to manage UX research, product decisions, and back-end implementation.
Key takeaways from the session:
- Attendees will learn how to form a strategy that allows taxonomy creation to be iterative, data-informed, and user-centered.
- Attendees will learn the value of performing open ended user research which can make room for new ideas, even within an existing taxonomy.
- Attendees will gain insights into some of the communicative and collaborative processes required to manage UX research, product decisions, and back-end implementation.
About the speaker(s)
Alexandra Teodorescu, Ph.D. is a Taxonomy Analyst at Indeed where she develops metadata structures and ways to extract information from text. Before Indeed she taught and conducted research into Linguistics and Cognitive Science at The University of Texas at Austin, Princeton University, and McGill University.
Amanda Krauss, PhD, has been a researcher for over 20 years, and has worked in digital media, academia, and consumer-facing product development. She has given presentations on privacy, design ethics, and persuasive research. In her free time, she likes playing with structured data and Python. She also works as a technologist and researcher for hire, and particularly enjoys working with organizations that aren't necessarily "in tech."
Leora has a penchant for spicy salsa, good books, New Orleans jazz, and improv comedy. After graduating from Tulane University, she moved to Austin,TX, to serve with AmeriCorps. During that time, she taught in English Language Learners and pursued a Master’s in Education from The University of Texas at Austin. She then pivoted to high tech and currently works at Indeed, an employment related search engine, as a Segment Manager.