Chatbot Prototype ProjectÂ
Project Overview
The MSOE marketing department reached out, during the 2024 fall semester, to the course Designing for Chat & Voice, prompting us to create a prototype chatbot within the software Voiceflow that expands upon the university's chatbot's current capabilities. The project's cross-functional teams collaborated with the Milwaukee School of Engineering (MSOE) marketing department, the Director of Enrollment Marketing and Events, and the Director of Digital Marketing to create the MSOE chatbot. Our primary audience was prospective students; the project lasted 12 weeks. This concludes with an extensive, 15-min, data-driven presentation with a live demo for project stakeholders.
Goal
My team's chatbot prototype focused on expanding the university's chatbot's ability to assist prospective students with scheduling a tour, accessing resources, and signing up for SMS text messaging. The SMS feature was not a component of the prototype, but a Figma prototype that demonstrated the feature.
Design Process
How to Understand Your User?
My team started by identifying our primary users' needs, wants, and limitations through preliminary research, interviews, and surveys. I conducted competitive market research on how other university employs chatbots to assist students. I identified key features of these chatbots that we could incorporate into MSOE’s chatbot. Then, proceeded to determine potential new features to add to the chatbot. This was then followed by the team brainstorming potential conversation structures, tones of voice, and chatbot framing. We created multiple user stories, Journey maps, conversation flows, test scripts, and chatbot iterations. To determine, we conducted usability testing of the chatbot prototype with our primary and secondary audiences.
Defining our primary audience and their journey
I created a value proposition, use case, and identified potential problems for our target audiences. The team created a persona within Figma outlining the key characteristics of our primary user demographic. We proceed to create a user journey map that identifies our users' actions, goals, feelings, pain points, and potential areas of opportunity.
How do you effectively create conversation flows?
Expanding upon our user journey maps, each team member completed a test script for how our users could interact with the chatbot. We then evaluated each script to determine overlap, tone of voice, and proper language usage. Then, we merged the conversation flows in FigJam and walked through them with the project stakeholders to identify areas of improvement.
Creating the Chatbot
-
Implemented MSOE data sources
-
Used an in-depth workflow
-
Utilized a variety of triggers, functions, components, Api calls, conditions, buttons, and prompts
Testing the Prototype
After completing the initial prototype, we proceeded to conduct user testing. I wrote a test script that contained qualitative and quantitative questions, which the team used in their interview. I conducted 4 in-person and 2 virtual interviews. We then conducted 7 usability tests with stakeholders and peers. The data collected from these interviews was analyzed and influenced the final design of the chatbot prototype.
Results
The final deliverable was presented to the project stakeholders in a 15-minute presentation with a demo, covering the design process over 12 weeks. The project was highly praised, and the prototype features will be integrated into the MSOE chatbot, and because of our prototype, the chatbot is now multilingual.
Lessons learned
-
Effective cross-functional collaboration techniques
-
How to apply effectively conduct A/B user testing, user testing, in remote and in-person environments
-
Effectively analyze data to identify optimization opportunities
-
How to present deliverables to stakeholders effectively. Through implementing shared terminology, remaining concise and clear, while providing in-depth justifications for design choices through applying insights from research
-
Team conflict resolution techniques and collaboration strategiesâ
Chatbot Presentation
