Sprout - Food Inventory Management for Small Restaurants


Overview

This in-class project started from our initial research on food waste, and we were thinking if it was possible to re-allocate this food surplus to help small restaurants, an essential part in the F&B industry. Our goal was to help small restaurants to manage their resources better by automating day-to-day purchasing, prepping, tracking and encouraging restaurant owners to make data-driven decisions.

Project Timeline
Aug - Nov 2020
15 weeks

Collaborating Department(s)
Georgia Institute of Technology

Supervisor
Prof Herb Velazquez

Role
UX Designer

Special Thanks to
Teammate - Sunny Choi

THE CHALLENGE

How do we help small independent restaurants manage their resources better?

THE SOLUTION

By automating daily tasks and showing the right data for future preparations, we incentivize smarter ingredients purchasing and usage habits.

Easy Inventory Tracking

Running Low
Show low in stock items so restaurants know when to restock without having to constantly check their pantry

Expiring Soon
Notify when items are close to expiry so they can utilize ingredients to the fullest

Scan Receipts to Populate Inventory
To save time, scan receipts to update the inventory (OCR), users can double check and edit any missed items.

Auto-generated Labels
Auto-generate labels based on purchases so they will never forget to label ingredients they purchased, plus the date purchased and opened

Reliable Bookkeeping

Small restaurants can now have all their expenses, sales and forecasted sales all in one place.

Users can easily view their past purchases as well as weekly sales, filtered out by external factors.

Informed Decision Making
Consolidate historical sales data with things that are happening around the restaurant to predict sales and help make decisions

For example, you may not want to stock up as much or you can try to increase delivery rates

Relevant News
Collect news that are relevant to the restaurant, for example, an increase in meat prices during the pandemic.

This will impact how much you spend on raw chicken, and maybe menu pricing needs adjustment

Design Process

THE PROBLEM

By the first year...

26% restaurants fail

By the third year...

60% independent restaurants fail

These independent operators make up 90% of all F&B businesses and are essential to the industry.

Yet these same restaurants generate

22 - 33B lbs food waste each year


Ecosystem Map

We wanted to know if there was a correlation between these two statistics. So to get a better understanding about food waste, we created a rough mind map of all stakeholders in the food ecosystem and types of food waste generated.

THE DREAM

We explored how minimizing waste can enable businesses to be more profitable and sustainable, and came across this concept of business sustainability.

Sustain environment   ↔   Sustain business
In business sustainability, a business uses social and environmental pillars in addition to finance, to help construct an environment in which it can thrive

THE HYPOTHESIS

Is there a possibility to reallocate these wasted resources to instead make profit for small independent restaurants?


Target Users

We interviewed 7 small, independent restaurant owners all of which had 4 or less employees except for one restaurant that had 30 employees. We were also able to work closely with one wings restaurant based in Atlanta.

During earlier interviews, we asked broad, open-ended questions related to:

Persona  →  age, restaurant type, num of employees, goals
Business process  →  purchasing ingredients, day-to-day prepping, customer base
Apps & Technology  →  e.g. Ubereats, Doordash, POS systems
Marketing / Finance  →  profit tracking, bookkeeping, target revenue


User Persona & Journey

After consolidating data from our interviews, we were able to map out a high-level view of how running a restaurant looks like on a day-do-day business.

We were able to see areas of high frustration and where waste is generated, which were heavily saturated in the process of purchasing, storing and preparing ingredients.

Narrowing down to Inventory

We followed up on our interviewees to schedule 30-minute calls to ask more questions around inventory. We also got to shadow one restaurant for a day.

Shadowing

Key Takeaways

Constantly checking storage spaces
A big pain is checking if there's enough left for production and keep going back and forth into (inaccessible) storage spaces. Items become underutilized, forgotten, and later discarded.

Multiple sources of supply
Restaurants purchase ingredients from many different suppliers, which makes it difficult to keep track of supplies coming in. Planning purchase trips when restocking require extra effort.

Overestimating daily sales
Restaurants prep daily and it's difficult to estimate how much they'll need. They tend to overestimate sales which results in food surplus that can no longer be served to customers.

Everything done manually
Processes within restaurants involve lots of manual labor, such as mental notes on what ingredients to buy, handwritten labels, etc. which are often times unreliable and lack accountability.


Ideation

After wrapping up our initial research phase and establising a direction, my teammate and I had a (remote) thumbnailing session to quickly generate ideas.

We used these to produce mid-fi concepts before showing them to a restaurant to get feedback.

MID-FI CONCEPTS

Showing the Value of Each Inventory Item

Our Idea
Calculate value of each ingredient by taking into account item cost and annual consumption (ABC inventory)

Value Proposition
By showing the value of ingredients, restaurant owners may be able to cut down on those that aren't as utilized

User Feedback
It's not necessary to show this info on a screen, as restaurants already know which items are most popular

Updating Everything Based on Orders Received

Our Idea
Generate a shopping list from a cloud-based inventory, updated real-time by orders taken from the POS

Value Proposition
Eliminate the need to keep checking the pantry, reducing time, steps, and effort taken to plan purchase trips

User Feedback
Valued ability to easily track inventory but concerned with accuracy of item count and units (e.g. wings are bought in lbs, and sold in pieces)

Taking the Physical Environment into Account

Our Idea
Auto-generate labels containing relevant information for each ingredient based on purchase

Value Proposition
This idea offers a systematic way to label ingredients to prevent items from being forgotten

User Feedback
Liked ease of integration into current processes, only concern is items moved to different containers

Predicting Future Sales Based on Restaurant Trends

Our Idea
Give insights to restaurant owners on factors that may affect sales such as time of year and location

Value Proposition
Tackle the issue of overpurchasing by providing an estimate of sales based on data

User Feedback
For users, this idea is valuable as it encourages them to prepare better, only concern is lack of historical data

Allowing Customers to Participate in Reducing Waste

Our Idea
We wanted to explore the customer side by offering the convenience of adjusting price based on portion sizes

Value Proposition
Ethical solution that encourages more interaction with customers and collective efforts to minimize food waste

User Feedback
Concerned that this concepts would reduce sales instead of increasing customer loyalty


Key Concepts

We narrowed down our value propositions once again to the three selected concepts and merged these features into a single, cohesive application.

Track Inventory
Make relevant info available at all times and save time on daily tasks

Record Purchases
Give more structure in workflow by automating manual tasks like labelling

Forecast Sales
Encourage better preparation and data-driven decision making

HOW IT WORKS

Acceptable Minimum Capacity

We took inspiration from the grocery store model for our inventory module where we use an acceptable minimum capacity as a threshold

1)   Item(s) gets checked out
→   Order being placed at the restaurant and reducing count of ingredients needed for that order

2)   Item(s) fall below minimum capacity
→   Ingredients being low in stock

1)   Item gets checked out
→   Urge restaurant owners to replenish supplies

Since it's not necessary to know the exact count of items every time, we resolve the concerns about accuracy

Prepped ingredients can be made from multiple ingredients, e.g. salad mix

Menu items can be made of a mixture of ingredients and prepped ingredients

One supplier sells many ingredients, one ingredient can be sold by many suppliers


Storyboard

To visualize how users would interact with our solution, we storyboarded how our app would integrate in and help with restaurant owners' daily routines.


Wireframing

We started drafting low-fidelity screens for our app to bring it all together, while keeping in mind the user interactions we touched on the storyboard.

MOVING INTO HI-FI

Incorporating Visual Design

We wanted to convey an inviting, natural and organic feel throughout our product to reflect our fictional brand's mission to help small restaurant businesses while contributing to a sustainable environment

Home Page

Inventory Items

Scanning Receipts

Sales and Forecasts

Design System

User Testing

Information to action
From testing, we found that the homepage was not proactive enough to encourage action from our users.

Inevitable Long Onboarding Process
Our app relies on the user to correctly link all their menu to ingredients for the app to bring value to them.

Mobile App vs Desktop Use Cases
Mobile is convenient for purchasing but for viewing trends and historical data, they doubt a mobile UI can show everything.

Design Ideas Applied

Connect to POS Interactions
To save time, restaurant owners can load menu data from a POS they already own.

Bulk Actions to Connect Menu and Ingredients
We applied drag-drop and bulk interactions to shorten the process of connecting menu to ingredients, enabling sub-recipes

More Actionable
We applied cards with more call-to-action, for example, we turned running low snapshots into a shopping list. This way, users know what to do, further shortening the steps to derive an action plan for purchasing.

Final Product