Auto Club Roadside Assistance Chatbot
Roadside assistance has been a hallmark of AAA's membership services. If something breaks on the road, you call the number on the back of your membership card. But this journey seems a bit outdated nowadays. So how can we innovate?
I was a part of the first innovative project given to Clublabs to envision the future of roadside assistance.
Project overview.
Despite the fact that AAA members love the roadside service, the process of requesting the service has always been a love-hate relationship.
Auto Club of Southern California(ACSC) within AAA wanted to create an innovative way for users to request assistance more efficiently.
Clublabs felt this was a great opportunity to unveil the department's mission, which is to be an incubation lab. A team of a product manager, a UX researcher, a UX designer (me), and three developers was assembled to lead this project.
My Role & Responsibility
As a UX designer, I mainly led the concept designs and voice and tone design for the dialogue. Not only focusing on the design, I partnered up with the UX researcher to contribute in workshops and research.
My Role:
Contributed and participated in a discovery workshop
Conducted the SME interview with the Emergency Roadside Service(ERS) agents
Contributed to the in-person member interview
Participated in dialogue design workshop provided by Microsoft
Designed the dialogue flows for multiple scenarios
Co-created the journey map that covers the user, agent, truck driver, and business
Created a solution approach for the chatbot UI and the system.
Contributed to the chatbot platform selection for the Proof of Concept
Conducted quality check for the development of the proof of concept.
My Responsibility:
Design the proof of concept of requesting roadside service through chatbot.
Participate and co-own the user research and its artifacts.
Design the microinteractions for chatbot interface.
Define the voice and tone for the chatbot dialogue.
Design the dialogue to reflect the voice and tone guideline.
Support the implementation of proof of concept
Quality check the development of proof of concept
Demonstrate the proof of concept prototype and walk through the design process to the stakeholders.
(Initial) Project Goals.
Increase the member satisfaction score
There were tons of feedback collected through TQS4, surveys, Tier 2 customer support, all pointing at the negative experience members have with the emergency roadside service(ERS). This decreased the member satisfaction score and impacted the brand reputation.
Increase the completion rate
The online request form had a high percentage of requests that were not completed. A major overhaul was long overdue.
Design Process Overview
A UX researcher(VH) and I followed the double diamond process to provide the solution for roadside assistance chatbot.
Discover
Understanding the needs of multiple stakeholders and users.
Discovery workshop
SME interview
Observation
Member interview
Define
Defining the voice and tones for the dialogue, and the direction of VUI.
Journey maps
Voice and tone guidelines
Storyboarding
Develop
Designing the solution approach for one unique scenario for the proof of concept demonstration.
Dialogue design
Chatbot interface design
Interaction design
Deliver
Supporting the development of the proof of concept and overseeing the development of the proof of concept
Quality check
Discover.
Discovery workshop
To clearly understand the project goal, we suggested to run a one-day collaboration discovery workshop. The participants include ERS business owner, ERS call center manager, AAA National ERS team, VP of Clublabs, Dispatch management, UX designer, and engineering team. Led by a UX researcher(VH), the discovery workshop covered 3 things:
Identify the pain points and goals for each team
Identify the measure of success
Align in short-term and long-term vision
SME interview
After the discovery workshop, there were still some questions unanswered for us to understand the scope of the project and the goals. We scheduled the SME interview with the ERS product owner to understand where the business is suffering. Through this, we were able to gain the hidden goal - to reduce the cost by reducing agent interaction with ERS calls.
Observation
The UX researcher and I wanted to understand the requesting process and the journey that users are going through. We visited the call center in Long Beach, CA to listen in some of the member calls. Our observation points were:
The dialogue between the agent and users
The current journey and use cases with the emotion users are feeling
ERS workflow from requesting a dispatch - dispatch - managing the dispatch team.
In-person member interview
The UX researcher and I felt we should validate the chatbot technology is something that users are expecting when requesting the roadside assistance. Also we wanted to gain insights on their past roadside assistance experience.
Key findings
1. Members well adapted the concept of chatbot
2. Use online platform for non-urgent cases
3. Talking to an ERS agent has very positive experience - calms them down.
Project Goals. Again.
Increase the member satisfaction score
There were tons of feedback collected through TQS4, surveys, Tier 2 customer support, all pointing at the negative experience members have with the emergency roadside service(ERS). This decreased the member satisfaction score and impacted the brand reputation.
Increase the completion rate
The online request form had a high percentage of requests that were not completed. A major overhaul was long overdue.
Reduce the cost spent on answering ERS calls
This is the goal that was discovered through the discovery workshop and SME interview. The organization currently loose money by offering roadside assistance service, however it’s the core benefits many members value. The way the business can reduce the spending on ERS is to reduce the call volume to the call center.
Define.
Journey map of users, truck drivers, agent, and business
Based on the data we gathered, we have created a journey map that showcases the different hand-overs. We identified the opportunity areas from the journey that users are taking.
Roadmap
Voice and Tone guideline
After the visit to the call center, I led the creation of voice and tone guideline for the chatbot. I leveraged the existing voice and tone guideline the content team follow, modified the principles so that the guideline can be used as a tool to align the chatbot dialogue to our agent conversation.