Online ordering through chatbot interactions
Mobile app design
Fancy Fancy is an imagined high-end restaurant located in downtown LA aimed at providing welcoming food complimented by a great ambience. Guests can treat themselves to their favorite greens, steak, or pasta and finish off with a guilty-pleasure of choice.
My role
UX / UI Design
UX Research
- Individual project for Google UX Design Certificate
Design tool
Figma
Timeline
12 weeks
The Problem
Users thinks chatbots cannot help them solve the problems
Users normally don’t like chatbots because in their past experience with chatbots, they tend to ask to talk to a representative instead since chatbots cannot effectively help solve their problems.
The Solution
Chatbot initiates and takes customer’s orders by laying out available options.
Research - User Interviews
Users assume that chatbots are not capable of fixing complicated issues.
I interviewed 5 users who had experience placing orders online or through phone calls.
Here are some of their thoughts on the experience with Chatbots
4 out of 5 interviewees stated that they personally don’t like chatbots.
They prefer talking to a representative directly when they think they have a complicated issue.
However, there’s one interviewee like chatbots. “It’s a good option when I don’t feel like talking on the phone.Or I can’t talk on the phone.”
Research - Competitive Analysis
The Main Insights
Users prefer placing orders with restaurants to save money.
Users can try trusting Chatbots if tasks are not too complex.
Food ordering/delivery platforms provide adequate information about menu items and orders for users to have more control over their decisions made online. For example, clear and detailed menu item ingredients, order status tracking, past order history, and more, and these features are highly liked by users.
How might We
How might we simplify the interactions with a chatbot and let users do the least work to complete an order.
Design + Evaluation + Improvements
I recruited 10 people to test lo-fi and hi-fi design prototypes and iterated my designs based on the users’ feedback. In the 2 rounds of usability testing, there were 3 major usability improvements:
The Final Screens
If I did launch the app
If I did launch the app, I would be looking at conversion rate and in-chat back-button usage. Conversation rate can show the data of the percentage of successfully placing orders via our app. Looking at the data of in-chat back button usage can reveal the percentage of users who cannot place orders through the chat successfully.
I would also be interested in doing an in-app questionnaire or a quick interview with users who pick up orders at the restaurant.Asking users these two questions would be helpful to determine if the product is successful.
Question 1: How easy is it to use the app to place orders? (Very Easy / Easy / Okay / Difficult / Very Difficult )
Question 2: Why do you think it’s difficult to use? What difficulties did you face while using the app?
Conclusion + Lessons learned
What I’d do differently next time.
1.Watch users using competitors’ products
During the primary research, I only asked users about their online ordering experience but I didn’t see it myself. If I could have seen how users placed orders online with competitors’ products, I would have had better ideas about the user flows they followed or pain points users had when placing orders online.
What information they looked for first, what frustrated them or what implements they particularly loved and why would be revealed if I had chances to see users interact with products directly.
Seeing how users use products will be different from listening to their articulated words about the experience. What users say may not always be the same as how they actually think and do.
2. Interview real restaurant servers
Interviewing restaurant servers and acquiring their knowledge about how to take orders and what common questions customers would ask is something I wish I could have done. Iterating the content from the perspective of servers is a point that I missed in the first place as designing the product.
Customer may only be clear about his/her own needs as a customer, but restaurant servers have a collective knowledge about most customer’s needs from a different perspective.
3. Chatbot or not chatbot?
This question lingered in my head for a while before I started the project. Even though this is just a practice project, I tried gathering user insights and doing research to determine if making a chatbot feature for online ordering is a good option in terms of desirability, efficiency and cost.
One thing I would do differently next time with chatbot design is making good use of AI support behind it. Suggesting new menu items based on their past order history or making combo suggestions based on items that are already added into the order.
One more thing I’d like to improve the experience by giving users choices to either order with a chatbot or go with a regular mobile app ordering experience as in browse all the menu items and add one by one manually. Because….some users may drop out instantly if they find out interacting with a bot online is their only choice.
