SkipTheDishes x BrainStation #DeliveringDigital Hackathon

Food Mood

Food Mood uses image recognition to help satisfy all those cravings you get from scrolling through your friends’ food snaps on social media.

Project Overview

In September 2020, I participated in the #DeliveringDigital Hackathon presented by BrainStation and SkipTheDishes. Working across timezones and disciplines with my peers in Vancouver and Toronto, we created Food Mood – a mobile solution that helps users find the food delivery they’re craving with ease.

ROLE
  • Product Designer, Asset Collector, Stitcher
PROJECT TYPE
  • Hackathon, Academic
TIMELINE
24 Hours
TOOLS

Pen & Paper, Figma, InVision

DESIGNERS
  • Kayley Cheung
  • Sharon Fung
  • Judy Chen
DATA SCIENTISTS
  • Luai Ghazi
  • Mengfei Liu
WEB DEVELOPERS
  • Emma Sweetnam
  • Reza Haleem

Problem Space

“As the restaurant industry is experiencing large disruption due to COVID-19, offering delivery can help restaurants mitigate costs, while providing consumers with a safe and reliable way to support local businesses. With many consumers adopting these services, this is the perfect opportunity to think of ideas for the future of the food delivery service Industry. What digital solutions can we create to support the food delivery industry?

Project Goals

As a team, we wanted to create a digital product that would help users easily find and support local restaurants. This included:

  • Helping users find restaurants quickly and easily
  • Streamlining the process of discovering delivery food options
  • Making meals easier to order, thereby increasing the number of completed purchases on meal delivery apps

Discover & Define

Research & Problem Framing

Secondary Research

To effectively tackle the problem, we needed to learn more about the usage of food delivery services in Canada.

41

Increase in food delivery service users in 2020

42

of restaurants added delivery services as a result of COVID-19-related closures

4

Average 30-day user retention rate on food delivery apps in Canada

Online food delivery is a $3.25 million industry in Canada, with expected annual growth of about 9% by 2024. Users, particularly those in the 18-29 age bracket, are willing to pay more for the convenience of having food delivered rather than cooking themselves. So why are food delivery apps experiencing such high rates of user attrition?

Hypothesis

We believed that delivery meals often do not meet users’ expectations of what they think they will be receiving when they place their orders, leading to great disappointment and making them less likely to order again. Based on our research and previous experiences with online food delivery, we moved forward with the following insights:

  • Social media users trust their friends’ opinions and experience food envy when they see photos of meals posted online
  • Users see food that appeals to them on social media but don’t know where to get it or what search terms to enter; they then abandon their search or settle for something they don’t really want
  • Users are in a hurry to get food delivered so they will settle for something that sounds “close enough” to what they want and are often disappointed when food arrives and unlikely to order again

Proto-Persona

We developed a proto-persona based on our secondary research, hypothesis and assumptions about our target user.

Proto-Persona

Opportunities for Intervention

  • Making it easier to find the dishes seen on social media (or the dishes that are the most simliar)
  • Gamification of food ordering – making the process more enjoyable and increasing likelihood of user retention

Design Question

How might we simplify the search for take out meals so that users can easily find food that appeals to them?

  • Making it easier to find the dishes seen on social media (or the dishes that are the most simliar)
  • Gamification of food ordering – making the process more enjoyable and increasing likelihood of user retention

Ideate

Exploring solutions

Task Selection

We worked with our data science and development team to ideate on possible solutions we could build within our short timeframe.

Our data scientists proposed that they build a model that used machine learning to match any uploaded photo to similar photos from a provided data set of Yelp! restaurant reviews, while the developers agreed that they could build a function that would take an image input from a social media feed (e.g. Instagram) and return suggested meals based on their similarity to the image in our new mobile app, which users could then use to order any of the meals for delivery.

For us as designers, this meant we needed to design an interface that allowed the users to complete the following tasks:

  1. Select a photo of a meal that appeals to them from their Instagram feed
  2. View suggested meals and their similarity that are available from nearby restaurants
  3. Select and modify the meal if necessary
  4. Check out and view order progress

User Flow

Sketches

We performed a “Crazy 8s” version of sketching independently to put out as many ideas as possible quickly. We then shared them between the three designers and discussed and voted on which ones would best serve the project goals and purpose of the product we were creating.

Wireframes

After selecting our most viable sketches, we worked to translate them into low-fidelity wireframes and handed them off to our developers so that they could begin to build the structure of the app while we developed a visual identity for the product.

UI Design

Incorporating branding & visual identity

Brand Inspiration & Moodboard

We opted for bright colours, reminiscent of retro video games, to appeal to the sense of nostalgia that draws in millennials and make the process of ordering food feel fun and delightful. We decided on the name “Food Mood” for the product, because it reflects the aim of satisfying cravings with what the user is “in the mood” for, while also being short and memorable.

UI Inspiration

For the UI, we opted for large, bold elements that were both accessible and made the process easy to navigate for the user.

High Fidelity Demo

Conclusion

Future thinking & key learnings

Next Steps

While we were able to design and prototype a minimum viable product within our 24-hour limit, we also considered what our next steps might be were we to continue to work on Food Mood, which include:

  • Create an image upload feature so that users can upload photos from any source
  • Implement an order rating system so that users could re-order meals they enjoyed and the data model could continue to learn and provide more accurate recommendations

Key Learnings

Hackathons are a true test of one’s creative confidence – with a 24-hour timeline on which to deliver, we had to go through the design process at an accelerated pace, and make decisions quickly. The most important lesson I took from this was how to work on an interdisciplinary, cross-functional team and collaborate remotely while still meeting all of our project goals on time.

Thanks for reading!

If you'd like to learn more about my project and design process, feel free to get in touch:

E-mail Me