A conversational UI that makes the group food ordering much easier.

Product Design, Conversational UI, Research


OverviewCompetitive AnalysisIdeationPrototyping & IterationStoryboardSynthesisFinal DesignReflections[BACK TO TOP]


People often encounter lots of problems when ordering food for large groups. Sometimes, they don't know how much to order. Other times, they are exhausted by the tedious coordinating and ordering process. How can we ease the process of group food ordering?


We designed a conversational UI, Kara, to make group food ordering easier. Kara recognizes different people’s voices and keeps multiple profiles so that it remembers users' food preferences and allergies. It helps users with multiple tasks related to selecting meals, determining portion, and getting food delivered.

Tools: Experience prototyping, Conversation modeling, Adobe After Effects
Duration: 3 weeks
Teammates: Bobbie Soques, Judy Kong
My Role: User research, Design Ideation, Video production

Competitive Analysis

To better understand the current market, we did a competitive analysis by looking at four existing conversational UI products: Google Home, Microsoft Cortana, Apple Siri, and Amazon Alexa. We created a competitive analysis matrix that compares the unique features, strengths, weaknesses, speech recognition performance, and voice friendliness of the four CUI products. More closely related to our project, we also checked whether or not the four products support food ordering, and where do they order from.

It turned out that Microsoft Cortana and Apple Siri do not take food ordering at all, and Google Home and Amazon Alexa both only support a very limited range of restaurants, such as Domino’s.


We brainstormed on different scenarios where people might be needing our CUI to order food, and wrote a short story to describe the context for each of the scenarios. We then narrowed our scope down to designing a CUI that takes group orders as we believed that the if we could handle this case well, the needs in many other scenarios will be fulfilled automatically.

We name our CUI Kara, since it sounds like a name of a friend and is very easy to pronounce. We then listed out some activities that we want Kara to support:

Recommend food/dishes
Summarize order
Schedule delivery
Account for allergies and preferences
Remind you of utensils
Recognize location (GPS)
Alternative input support (phone screen/call)
Pause and resume interaction (for group discussions)
Recognize and listen to single owner to avoid all of group ordering at once

Prototyping & Iteration

To figure out more user needs that we didn't discover, we went through experience prototyping with a few students. We asked them to act in our scenario and recorded what questions they asked and if they had any confusions. During the experience prototyping, some people were confused about the food options - “Does it only have pizza ordering, or does it have something else?“. Others asked for suggestions, like “How should I order?“ or expressed more specific concerns like “I think there are too many chickens in the group meal.“ To address these unexpected needs, we added more features to Kara:

From this process, we also recognized that Kara was a bit "boring" and acted just as a normal call center agency. The personality of Kara is part of the key components in the CUI design because after all it's not the conversations with CUI that people will remember down the road. It's how the conversations made them feel. To build a more amicable, and a bit sassy, personality, we gave Kara ability to have small talks with the users.

Another crucial part of CUI design is error prevention. A good CUI should be flexible, responsive and smart when handling errors. We brainstormed the possible errors that Kara could encounter and its responses.


To better understand the context of our CUI, we made a storyboard of when and how people would use Kara for group ordering.


After the two rounds of experience prototyping, critique, and refinements, we created the conversational flowchart of Kara, on both the general ordering procedure and the specific user flow in the party ordering scenario we worked on. The flow charts consisted of pre-attentives and attentives, each has different utterances and responses, with a few error-catching features with response.

A specific case that shows how Kara might interact with users
When Kara is activated, the screen will change from the left to the right indicating the change of state.

Final Design

We designed Kara to be facilitate users at every stage of group food ordering. It handles from calculating the amount of food, to recommending popular options, to placing an order for sometime in the future, to reminding users to order enough utensils and napkins. It also remembers what users are allergic to. It automatically knows user's locations, and sends the delivery straight to the door without user inputting an address.

Imagine a group of friends forgot to order food for a Halloween Party. Since everyone was busy preparing for the party., Alex used Kara to help them order pizza.
With the given number of people coming to the party, Kara suggested an estimate of how much pizza to order.
Kara recognized that a lot of people were talking at the same time, so she asked them to only let one person talk at a time.
Kara also helped Alex check for ingredients Alex is allergic to, and reminded her when she added something that included such ingredients to her order.
At last, Kara figured out the current location using GPS and delivered the food to their house.
Kara can handle group food ordering in many other settings, such as meeting, travel or events.


I really enjoyed working on this project because it forced me to consider how to design more intuitive interactions beyond the screens. Designing the personality for Kara particularly made me appreciate how the nuances of tones and expressions constitute the natural human interactions. I'm excited to explore more forms of natural user interfaces in the future.