Meet Whaley, the Chatbot on a mission to ease the experience of users hunting for jobs in DC Government.
A UX Case Study
Introduction: Objectives, Research and Synthesis
In continuation of client work for the City of the District of Columbia’s Service ReDesign Initiative, I was tasked with the design of a robust working chatbot to solve a pain point that was discovered during phase one. To learn more about phase one research and synthesis of employee experiences, please read my case study here.
Employees within the Health and Human Services Agency cluster had described pain points during recruitment, hiring, on-boarding and promotion. Research of the application portal experience (Careers.Dc.Gov) and Reddit discussion threads further revealed applicant frustration and dissatisfaction. Applicants described uncertainty and confusion about where they were in the process. Were their applications still being reviewed or had they been passed over?
A research plan and the design thinking process shaped my ability to meet the production timeline of 3 days. The primary objective was to understand if the customer experience might be improved for job applicants and employees by the use of a Chatbot. And, how a chatbot could be designed to answer questions such as status inquiry, timeline updates and overall user experience.
User Personas
Maria is the User, a future employee of the District of Columbia who demonstrates frustration about the application hiring process. She is a recent graduate of a Master’s Degree program and has just completed applications for grade 7 and 9 positions. She is motivated, organized and seeks transparency around process. She is uncertain about the status of her application or timeline of the hiring process.
The chatbot will meet the needs of the user by providing guidance. The persona for the bot is Whaley using the Whale as metaphor for its stability in navigating a space as vast as the ocean, or government hiring. Whaley has the primary function of aiding the user in obtaining updates on status and timeline of the application but will also collect user satisfaction ratings and will teach information on the value of the whale as a geo-engineer in managing atmospheric carbon levels. The conversational tone is playful to ease the tension of the user who may be feeling tired or anxious around seeking employment.
Build better voice apps. Get more articles & interviews from voice technology experts at voicetechpodcast.com
Design and Rapid Prototyping in Low and Mid Fidelity
Early ideations of use cases and flows informed my choice to address timeline, status, service rating and information with the bot. I prototyped using the white board and early usability tests to understand and refine decision flow. I employed an additional round of rapid “paper” prototyping with the white board to determine and test dialogue flow and CUI architecture.
Next, I built a mid fidelity prototype using Chatfuel and deployed it on Facebook messenger. More user tests revealed flow glitches for the Rate Your Experience feature as well as collecting user information for email notifications of updates to the tracker and timeline features.
Repeated user testing informed the need for more contextualization of the choice to use the Whale persona. I added more introductory language to anchor the meaning of the Whale as a guide and friend in a new, vast environment such as the ocean or government hiring.
Summary & Next Steps
In summary, by using the Design Thinking Process, I produced a functional conversational user interface (CUI) or Chatbot, to solve for user pain points within the District of Columbia’s service ecosystem. Empathy and research were employed to identify and understand the needs of the user. User pain points were alleviated by creating a means for identifying application timeline, status tracker and gathering user data about overall service satisfaction. The Whaley chatbot functions in a future state wherein an integrated tracking system can be resourced to create answers for the user.
Rapid ideation, low to mid fidelity prototyping and user tests informed redesigns leading to a functional bot for demonstration to the client. Early use tests revealed an increase in satisfaction by users. They noted relief at their problem being solved with the tracking and timeline tools. They were willing to rate the service experience and also described enjoyment of interacting with Whaley’s persona and learning about the greater ocean ecosystem.
If the client is interested in further use of CUI’s, I recommend secondary research into existing on-boarding challenges and the potential design of a bot to interact at the on-boarding stage as the 39 chapter employee manual has been described as overwhelming.
Additionally, employees reported customer dissatisfaction when having to repeat their personal information each time their call was transferred during attempts to obtain services. A bot that integrates to a future state centralized customer database could expedite retrieving client information for employees fielding calls.
While the government service ecosystem may not be ready to incorporate CUI’s quite yet, the lean UX process used here demonstrates some of the user pain points could very well be alleviated by the bots when the time arrives for their integrated functionality.
