Streamlining McDonald’s customer service with enhanced chatbot capabilities: faster responses, seamless ordering, and exclusive coupons.
Project Overview
An AI Chatbot for McDonald's
Joindeapp specializes in opening new sales channels for clients by integrating AI chatbots on social media platforms such as Facebook Messenger and WhatsApp. These chatbots allow users to complete their entire customer journey, from shopping to checkout, directly within the chatbot. In collaboration with McDonald’s, Joindeapp created an AI chatbot to promote their new campaign featuring the Travis Scott meal. This chatbot aims to enhance customer service response times, facilitate easy ordering, and offer exclusive marketing campaign coupons.
Role + Team
End-to-end Lead designer working with Joindeapp dev team
Timeline
Mar 2022 – May 2022
Focus
AI- mobile application – UX/UI
Problem
The existing turnaround time for McDonald’s chatbot was between 2 to 6 minutes, with some queries receiving no response. Additionally, the current chatbot lacked essential functionalities such as ordering through chat and accessing customer service.
Solution
The Joindeapp team successfully implemented new features for the McDonald’s chatbot campaign, significantly enhancing the user experience. These improvements enabled customers to complete their tasks through the chatbot more efficiently.
Approach
Given the challenges, my approach involved analyzing other B2C chatbot competitors on Facebook, such as Levi’s and Kayak. This analysis informed the ideation, user flow, and user testing phases to craft the optimal design solution.
1
Competitive Analysis
Examined successful B2C chatbots to identify effective features and best practices.
2
User Flow
Brainstormed potential features and mapped out user flows with the chatbot.
3
Design Ideation
Created multiple iterations of key feature prototypes to test with users.
4
Usertesting
Conducted testing sessions to gather feedback and refine the chatbot design.
Competitive Analysis
Utilizing analytical insights gleaned from our proficient analytics team regarding the usage of bottom navigation in our production environment, I transitioned my focus to explore cart placement and overall cart experience across various ecommerce platforms.
1
_
KAYAK Unique Features:
Real-time travel updates
Interactive travel tool
Gamification and rewards
2
_
Levi’s Unique Features:
Personalized shopping experience
Community and social integration
Interactive product information
Sketches
With an understanding of how other B2C competitors’ unique chatbot features operate, I decided to list acceptable features for the McDonald’s chatbot experience, including user stories and responses the chatbot would be able to offer.
1
_
Flows
Browsing the menu and ingredients
Making a purchase
Exclusive coupons
Signing up for the email list
Store locator
Asking FAQs
Logging into accounts for deals
1
_
Userflow
Userflow Visualization
With sketches complete over the different flows and features, I created an enhanced version of my sketches to better collaborate with team members like developers and the PM on the team.
Previous
Next
Design Ideation
With previous research completed, I started to design iterations that covered the entire flow of the McDonald’s chatbot experience.
1
Iteration
A user signing up for McDonald’s and then performing tasks within the conversational flow. Users receive quick and accurate responses to their needs.
Users logging into an account.
2
Iteration
The signup process was removed based on consistency with Facebook Messenger.
3
Additional iteration
I created an additional iteration that was separated from the previous two iterations to have it exclusively for obtaining unique coupons through easy gamification quizzes pertaining to the chatbot. This feature not only enhances the overall objective but also attracts customers to make purchases.
Usertesting
Upon completion of the ideation phase, I closely collaborated with the other designer on the team to conduct user testing on our prototype. With a diverse pool of over 30 participants through usertesting.com, our tests focused on time on task and new features for the McDonald’s chatbot.
Key Takeaways
1
Quick Replies:
Users often clicked on quick replies before typing anything. Thus, unique quick replies replaced previously used ones to minimize the need for typing new inquiries.
2
Sign Up:
The sign-up option was removed; no sign-up flow was needed as logging into Facebook already uniquely identifies the user, allowing them to start using the chatbot immediately.
3
Terminology Confusion:
The term “Log in” was confusing since users are technically already logged in if using the chatbot. This was replaced with an option to link an existing account.
Final Designs
Based on our comprehensive research findings, the Joindeapp team successfully implemented the new design changes for the McDonald’s chatbot campaign.
Outcome
The enhanced AI chatbot for McDonald’s, developed by Joindeapp, transformed customer interaction by significantly reducing response times, simplifying the ordering process, and providing quick access to customer service and exclusive deals. This project improved customer satisfaction and drove increased engagement and sales through innovative AI technology.