Overview

Bento is an app-supported meal delivery service which I created as a school project while attending UC Berkeley Extension’s UX Certificate Program. It specializes in work-week lunch delivery that can be scheduled days in advance. Its differentiators are its modular and customizable meal arrangements and an easy to use group ordering feature. The project was initiated in late 2017 and completed in early 2018 over a period of 10 weeks.

Platform/Format: Mobile application

Key Interactions: Customizing orders, Inviting friends

Inspiration

During my time living and working in Japan, it was common for my work colleagues and I to place a collective order for bento boxes (Japanese packed lunches). I loved the opportunity to bond over food with my co-workers, however we would often encounter one of two pain points.

1 - One person efficiently placing the order meant that not everyone would be happy with what they got.

2 - Each person could decide individually what to order, but the process was time consuming and splitting the bill required extensive accounting skills.

Both scenarios could get the job done, but no one was ever completely satisfied with the outcome. With a meal format as modular as a bento box, there’s no reason anyone should have to get a food item they don’t like, nor should ordering or paying for food require project management skills or complex math. These are the issues Bento seeks to address.

Discovery & Research

Survey Findings

To understand the prospective users of this food delivery service, I designed and conducted a survey about people’s lunchtime habits during the work week. The questions focused on respondents’ decision-making when making choices about food, and reported habits around group dining. From the results it emerged that roughly half of respondents plan their next day’s lunch at least one day ahead. The second most frequent approach to lunch, however, was to have no plan with 26% preferring a spontaneous approach. The app must be made to function equally well for those who want to order ahead and those who need to order on the fly.

While it has become more and more common for modern workers to eat alone (nytimes), 30% of respondents to my survey reported having lunch with colleagues at least once a week. Of that subset, over half said they ate with colleagues nearly every day. What this told me was that though social trends might be moving more towards solo lunches, there is still a significant number of people who want to make lunch a more social occasion.

Primary Persona

With the new features I was hoping to bring to the ordering experience (group orders and modular meal assembly), I decided to make my primary persona someone who would actually use those features. As a nod to our secondary user, someone ordering lunch for one, I knew it would be important to make it easy to bypass any of the additional social features that our primary persona would be using.

Comparative Analysis

I conducted a thorough analysis of direct and indirect competitors for work week lunch consumers, looking at the following brands:

There are a multitude of models of app-enabled ordering available with different solutions to making the experience faster, more enjoyable or more customizable. For Bento, I wanted to create a service that doesn’t sacrifice speed or convenience, but still allows for deep customization. The group order feature was also mostly lacking in existing apps (though some have added it since this project was undertaken) and represented an area of great opportunity.

Design Goals

Key Interaction: Inviting Friends to an Order

Sketches

To make sure that users could access all aspects of their order (items, location, time, group order options) in one place, I created an order home screen. With so many possible actions (most of which are required inputs), it was important that there be no extraneous information cluttering this screen and that the areas for user input be clearly indicated. Early iterations were heavy on buttons with icons. Due to poor information scent in early reviews, these were revised to a more standardized, form field format.

For the invite interaction itself, I wanted to make it as easy as possible for users to find their friends and invite them. Allowing for name, email or phone number lookup while also providing a list of recent invitees will help users find whom they’re looking for. An additional feature added at this stage was the ability to combine or split the bill as desired.

Key Interaction: Building a Bento

Sketches

My goal when designing the Bento building interaction was to keep the focus on what the customer is creating while still providing access to an extensive menu. I sketched and ran paper prototype sessions with two types of menu interaction, one that slides in from the side and overlays the bento in progress and another that slides up underneath. While the first arrangement allowed for more menu items to be displayed at a time; testers stated their preference that the menu not obscure the bento they were building.

Further input was gathered from users in the form of a card sort of menu items to produce optimal grouping of the different food items being shown.

Evolution to Final Screens

In the final mockups, I focused on simplifying language and labels, and defining a visual style for the app. I decided on a deep red color to represent the brand and used it in instances where I wanted to focus users’ attention. While the use of the static bamboo background reduced available screen space by about 10%, the aesthetic benefit felt like a big improvement. Additionally, it’s inclusion only negated the utility of the very edges of the screen which is a poor location for interactive features. In order to polish the aesthetic further, I incorporated some aspects from Material Design such as cards and their form field style.

Final Takeaways

I found the biggest challenges in the project to be:

• Designing a compact, easily navigable food menu that still had a large and varied selection of items.

• Thinking through the communication and information flow for the group order system.

• Developing a visual style that was additive to the desired feel of the app experience without contributing visual clutter.

Though I was happy with the outcomes, I would like to re-examine some pieces of the app and possibly conduct further user-testing on interactions that weren’t included in the current version (such as receiving an order invite).

In the time since this project was completed, many food delivery apps have added some group functionality. However, the ability to pay separately while ordering as group has, to my knowledge, not yet been implemented in any of the existing major apps and remains an area of opportunity.