comm lead chatbot

The problem

1 student advisor 🆚 hundreds of prospective students

The growing popularity of the Communication Leadership (Comm Lead) program has led to a surge in graduate applications. The traditional email response method by advisors is now drowning the Student Advisor, as manual replies with identical content are inefficient. Prospective Students, who I was in this position, seek a more comprehensive understanding of the Comm Lead program during the application process. There's a rising expectation for accurate and fast responses, emphasizing the need for an improved communication system that meets the evolving demands of both applicants and the advisor.

The goal

Transitioning from manual email answering overload questions to efficient communication

Initial solution 1

In a brainstorming session, we explored using Mailchimp for email automation to save advisors' time. However, this could still be time-consuming for students writing emails and challenging to manage for advisors due to complex automation triggers.

Selected solution

Initial solution 2

Automated response mechanism + Staying at CommLead website = Chatbot

Initial solution 3

We also considered redesigning the website's information architecture to include search and filtering features. However, keeping the content updated remains a challenge for student advisors.

The research - understand student advisor

Conducted user interview with advisor, learned their workflow, and aligned with the expectations of a chatbot

FAQs are primarily text-based

The advisor's email FAQs are designed to prioritize clear and accessible communication, mainly comprising plain text with the occasional inclusion of links for additional information or resources.

Budget friendly

Advisor mentioned that the project has a limited budget, emphasizing the need to not only reduce the costs of building the chatbot but also to minimize the expenses associated with its ongoing maintenance.

Easy to operate and data report

Advisor specified that the chatbot needs to be able to integrate into WordPress, ensuring it's easy to update and maintain. Additionally, they expressed a desire to obtain data reports on student interactions with the chatbot.

We avoid complex interactions like embedded too much multimedia, focusing instead on straightforward information delivery.

Due to budget constraints, hiring an engineer to build from 0 to 1 is not feasible. Therefore, we need to conduct research on third-party chatbot plugins available in the market.

This finding helps us narrow down the scope of third-party plugins. Ideally, maintenance should be straightforward, requiring advisors to only update text.

The research - competitive analysis

Tidio wins among 5 chatbot solutions

Competitor analysis

The research - understand prospective students

Google Analytics shows that 32% of students are international, and they frequently click 'make an appointment' on the website.

Consider two student personas

International students and domestic students seek different information, with international students more concerned about issues like Visas. Thus, when designing the chatbot, we need to consider the needs of both groups. Additionally, international students face time zone differences, making email inquiries less effective for timely responses. This indirectly validates our chatbot solution.

Provide a "make an appointment" entry point

Many students schedule virtual meetings with student advisors through the website to get their queries answered, so we need to include a direct appointment-booking service in the chatbot.

Through an online survey, we collected the FAQs that prospective students are most interested in.

Admission

Financial

Degree

Career

We tried to categorize the FAQs received from students into the four categories provided by the student advisor, but found that these categories didn't include information about the program itself, such as events and campus life. Therefore, after discussing with the advisor, I decided to add an 'Events' category.

Events

The design decision #1

Should students start by choosing whether they are international or domestic?

Entry point user flow I started

Pros

1. Facilitates our advisor to subsequently conduct data analysis separately for the two categories of students.
2. Allows international students to find answers more quickly and directly, such as regarding visas and related issues.

Cons

1. Initially treating them differently can make students, especially international students, lose a certain sense of belonging.
2. Both FAQ categories of students have a large overlap, so it's unnecessary to add an extra step by separating them.

Final entry point user flow

Users no longer need to select the type of student they belong to; questions specific to “international students” are integrated into the subdivided categories, making the entire Information Architecture (IA) and user flow more streamlined and consistent.

The design decision #2

Students' preference of booking meetings 🆚 The advisor doesn't want to answer questions can be found on the website on the meeting

Through a survey, we found out that students prefer to get answers to their questions by directly making appointments with advisors, as they believe such answers are more effective and authentic. However, contrary to this, most of their queries can actually be found on the website, and advisors also wish to avoid repetitive and inefficient virtual meetings.

The design decision #3

Text-based conversation needs joy and character too

We presented this question-and-answer chatbot wireframe to five prospective students in CommLead Program. They found the categorization of questions logical and everyone was able to find the answers they needed. However, they mentioned that the process was quite dull and expressed a desire for some joy or entertainment within the experience.

We adjust voice and tone

Freindly

Welcoming

Optimistic

Helpful

The design - final deliverables

CommLead style guide and fonts

Tidio backend flow charts and operation handbook

Reflection

When the product service targets more than one type of user, and these target users differ significantly, such as students and advisors in a chatbot, it's crucial to clearly define the persona for each type of user and strive to balance their user goals.
This is my first time balancing business goals in design, such as budget, acquiring user behavior data, etc. From this experience, I learned the importance of conducting comprehensive competitor research and interviewing stakeholders to understand their objectives.
To reduce development costs and time, it's essential to flexibly use low-code platforms and third-party plugins. Perhaps in the era of AIGC, everyone can be both a developer and a designer.

CommLead Chatbot