Project Overview

Context

This is a 10-week practicum project that my teammates and I created a chatbot answering the primary questions prospective students have when exploring the UW Comm Lead graduate programs, which will be embedded in the Comm Lead Website. This is a Design Award project at Comm Lead Screen Summit 2022.

Problem Statement

The information that prospective students need as they apply for the graduate program is spread out across multiple Comm Lead and UW websites, making it difficult to locate the answer to any specific question. A chatbot embedded in the Comm Lead website would help answer primary questions and/or direct prospective students to the information they need. If they need additional help, the chatbot would also assist them in contacting the program.

“How might we design a chatbot experience that allows prospective students to get FAQs answered 24/7 and have access to the resources and help they need?”

Selecting the Chatbot

Competitive Analysis

We first conducted a quick round of research comparing 18 available Chatbot options in the market. Then we screened 5 Chatbot products and carried out a competitive analysis comparing their prices and features, and finally decided to use the product TIDIO.

Why Tidio?

Use Research to Inform Design Strategies

Secondary Research

We conducted secondary research to scope out users’ interests and needs. Key findings from the research helped us develop personas and the basic structure of the chatbot.

Sources:

• Comm Lead website Google Analytics data.

• Youcanbook.me appointment information and common questions from faculty Heather.

• Prospective COMMLD research and survey conducted by UW instructor Misty Weaver and her co-contributors.

Persona

As domestic students and international students usually have different needs and concerns while applying for graduate programs, we created two personas, one stands for the goals of domestic prospective students and the other stands for the goals of international prospective students.

Basic Structure of the Chatbot

Based on research, we came up with 4 main areas of FAQs to include in the chatbot: Admission questions, Funding questions, Degree questions, and Work opportunities.

Online Survey

Before designing the complete User Flow, we ran a quick round of online survey to further understand users’ needs and gain insights about their experience interacting with a chatbot.

Key Findings:

• 60% of users prefer selecting questions from a menu than typing in questions directly.

• Information that users frequently have trouble finding on the website include: Opportunities After Graduation, Student Experiences and Faculty & Staff, etc.

• Users want more distinct personality for the chatbot.

User Flow

Combining all research findings, I mapped out the complete user flow of the chatbot.

Designing & Testing

Chatbot Demo

View demo →

Visual Identity

Usability Test

Background & Participants:

• 4 domestic and 6 international current Comm Lead students.

• All participants have previous experience using chatbots.

• Participants were given an instruction sheet for the unmoderated test and filled out a survey in the end.

Future Recommendations:

Due to the tight timeline, we don’t have time to implement some solutions after the test. But we come up with some great strategies for staff who will continue to work on the chatbot in the future.

• Provide smoother navigation: less scrolling up, slower moving chat, options to return to menu after thank you note, etc.

• More unified voice & tone

• Include more categories and FAQs: Student experiences, Housing, Fee breakdown, etc.

• Support typing-in queries

Going Forward

Feedback

• The project won the Design Award at Comm Lead Screen Summit 2022, which is the annual student project showcase at UW Comm Lead graduate programs. Industry leaders were there to judge student presentations, and awarded prizes to the top projects in each of the three categories (Strategy, Design, and Story).

• 80% of users said a chatbot would’ve been useful when they were applying.

• One quote from our clients when reviewing our personas:

  “I feel like I know the 2 students so well LOL.”

Reflections

• All participants of the user testing are current students. Current students were all prospective students once, so their insights are definitely valuable. However, if we had more time, I hope we also recruited some real prospective students and listened to their perspectives. Then there would be more useful data from the testing.

• There’re still a lot we can improve to make the end-to-end experience more seamless and complete. At this stage, users may still encounter small “frictions” as they navigate through the chatbot. And we don’t have time to fully polish the language to make sure it is all unified due to the tight schedule. I will definitely follow up with staff who will continue working on the chatbot to get to know the timeline and give a hand if needed.