<aside> 💡 Collaboration with Anahita Gharai, Dawit Barrows, Emma Drapp, & Frannie Ello

</aside>

Our second to last assignment in the class marked the practice of a new form of prototyping that many HCDE students were familiar with, but had not been able to put into practice: Behavioral Prototyping. Similar to the movie Wizard of Oz, this form of prototyping aims to test emotional touchpoints, user behaviors, and other key interactions in a low-resource format. For this project, our team had prototype one of three experiences: a voice-operated assistant, a gesture recognition platform, or a chatbot/text messaging app.

Out of these three ideas, our team decided to create a chatbot that needed to meet the following requirements:

• How will users know which commands are available to them?
• How will they understand how much progress has been made through their task and when they are finished?
• How will your system handle cases where the user makes a typo or types a command that is not understood?
• How much flexibility will your chatbot tolerate in terms of natural language vs. requiring users to be strict on their syntax?

With these requirements in mind, our group decided to create a chatbot prototype, reframing the McDonald’s ordering system into a text message format that utilizes flows similar to a binary tree.

Brainstorm

Before diving into creating our chatbot flows, we first needed to outline what tasks we wanted our user to perform with the chatbot. The aim of this brainstorming was to create tasks that felt natural to the flow of ordering a meal, as well as uncover additional use cases that we might have to account for when it came to prototyping.

We narrowed down our tasks into the following steps:

1. Order a Big Mac
2. Clarify if the item was a “Meal” or just the item itself
3. If the item is a Meal, clarify what size
4. Allow the user to explore drink options
5. Confirm & send their meal to purchase

Once our team clearly defined our required tasks, we needed to familiarize ourselves with jargon, flow, and customization options to utilize within our chatbot. Our group utilized the following resources to understand what steps take place within the ordering process:

1. Look through McDonald’s menu on their website
2. How to order a Big Mac from McDonald’s on Doordash
3. Looking through “drive through” ordering scripts

The three resources that were used to help frame our prototype messaging: McDonald’s website, McDonald’s storefront on Doordash, & Drive Thru Ordering Scripts

Prototyping

To test out our Chatbot, we used an interface that was simple and accessible by both our team and participants: Apple iMessage. Our group chose iMessage as a medium for the following reasons:

• Using a familiar interface could aid in adding a sense of realism to our users who will be testing out the prototype
• Companies such as Delta, Best Buy, and others use iMessage as a customer support medium
• “Wizarding” this prototype is simple, effective, and does not require excessive equipment to develop and test

A mockup of what we aimed the chatbot to look like.

Moving forward, our group created an outline of the expectations between our team members and the participant.

1. Have our participant receive a text from “McDonald’s” to begin their order.
2. The Wizard will be using their iPhone device to text the user as if they themselves were the McDonald’s Ordering System.