The food industry largely looks to the restaurants to understand trends. They use these trends to design what sort of Chinese dish will end up in the frozen aisle in 6 months or what flavour of humus will be next. 

Food and data with FoodGenius

working with a startup on a tool for people in the food industry powered by big data

I worked with a small startup team on their venture to create a tool for people in the food industry, powered by big data. FoodGenius, after being startup-in-residents for a couple of months, moved into their own office and ran two design and prototyping rounds independently. They had gotten mixed feedback on their two directions and reached out to IDEO for further help. See their two prototypes below.

The food industry looks to the restaurant scene for inspiration as they make products for the grocery stores but before FoodGenius, had to rely on manually visiting or calling restaurants and enquiring about the menu or buying the same reports as everyone else in the industry. They reports similarly rely on a research company calling a sample of restaurants to get data.

The needs and 'entry points' for the target users varied greatly. While some are working on briefs to create new flavours for humus, others are working on deciding price points or figuring out who to target for a new snack. Yet others are working on how to best advertise their new creations through tantalising recipes.

The general idea of FoodGenius is that you can search for an ingredient. The system then takes data from millions of menu items on the web and shows you an ordered list of which ingredients are most often (or least often) paired with it. You can build a "dish" by adding ingredients together. The list re-organises to reflect the next ingredient most commonly paired with the "dish."

The needs and 'entry points' for the target users varies greatly. While some are working on briefs to create new flavours for humus, others are working on deciding price points or figuring out who to target for a new snack. Yet others are working on how to best advertise their new creations through tantalising recipes.

The biggest insight and departure from the two initial directions, was to consider all variables as potential inputs and outputs. 

FoodGenius initially treated the metadata about your "dish" as a filter. Only show me fast food and fast casual. And only starters and entrees.

We designed a more flexible system where every variable can be a lever to pull or an output depending on need. This meant that the user can ask questions like: Is this combination more common in a fast food or fast casual locale? Is it more common as a starter, main, side or dessert? 

As you can see on the right, all of the parameters listed can be treated as filters or as outputs (in %).

Brief

• Design a platform for professionals in the food industry to tap into big data from the restaurant industry

Time and team

• I joined the FoodGenius crew as a designer-in-resident and worked with the CEO, CTO and COO
• The final week we had support from a visual designer from IDEO
• 4 weeks

Reflections

This project, albeit on the shorter side was a huge learning opportunity for me. I'm super grateful to Justin, Dave and Eli for showing me a new way to approach design. Quick decision, more hustle, more soul, more passion and more food. More "on the line" in that, the "clients" have a lot riding on the work both professionally and personally. This compared to IDEO's process which is perhaps more thorough but also more indulgent.