Designing for Positive Behavior Change

March - June 2017

Curbing Food Waste

For a course on designing for behavior change, I worked with 2 teammates to apply theories from Psychology to promote positive behaviors in curbing food waste. As we are a group who enjoy food in all its forms and are aware of the global food imbalance, the idea of controlling how much gets thrown away was particularly relevant to our interests. For 3 months, we combined these theories into the user centered design (UCD) process - from problem definition, research, and design ideation to finally, an interactive prototype. 

Project Overview

Under the guidance of Professor Gary Hsieh, we developed an app that positively enforces behavior to limit how much is bought so as to prevent how much gets thrown away. Based on the lifestyle and values of our target user base (i.e. millenials), we focused on features that help guide and plan their interactions with food. Our solution is an app-based design that provides relevant information to users and gently nudges them toward the desired behavior for more awareness and impact. It is a shopping planner that is geared toward buying only the amount you need not what you think you need.

 

Design features were based on the following Psychological models of behavior change:

  • Trans-Theoretical Model (TTM)

  • Theory of Planned Behavior (TPB)

  • Health Belief Model (HBM)

  • Social Cognitive Theory (SCT)

  • Self Determination Theory (SDT)

  • Incentive Design (ID)

  • Goal Theory (GT)

My Role

Lead Project Researcher

  • Developed a research plan that guided the team through the initial research phase

  • Created the report-out of our findings 

  • Created and developed personas based on findings from interviews

  • Led analysis that prioritized user needs and funneled findings to concrete features during ideation

  • Planned and conducted evaluations of prototypes and reported findings quantitatively

Lead Interaction Designer

  • Developed Information Architecture (IA) to organize features we wanted to pursue

  • Used Sketch and InVision to develop mock-ups

Problem Statement

Every year, approximately 50% of all produce in the United States is thrown away. This equates to about 60 million tons - or $160 billion worth - of produce, about “one third of all foodstuffs.” Producing all this food wastes water, land, and other resources. According to the Environmental Protection Agency, wasted food is the single biggest component in American landfills, which are a rising source of methane gases with negative climate consequences. In spite of all this waste, an estimated 5-10% of the population is still going hungry or are malnourished.


Approximately half of the food waste occurs at the consumer level, where an average US household wastes 1,160 pounds of food annually. Major factors include poor planning, overbuying, and improper storage. According to a Harvard Law School study, confusion over the meaning of date labels is a major contribution to unnecessary food waste. Yet, the International Food Information Council (IFIC) Foundation found that many Americans are “shockingly unaware” of how much food waste they produce. And for those Americans who do recognize that their food waste is a problem, they do not know where or how to start reducing it.

Goals

Our aim is to empower people to reduce food waste by providing a shopping companion app that:

  • Helps users plan ahead for shopping trips and buy only what they need: Based on a quick intake questionnaire, users can specify how many people for whom they cook during the week, how many meals they want to cook, and browse suggested recipes. The app custom generates a grocery list for exactly the food they will need for the week.

  • Creates an efficient and customized shopping route through the store: It saves users time they might spend otherwise trying to figure out what to buy, and how to piece it together into meals.

  • Teaches users how to store food optimally and use it before it spoils: Users have several different points of interaction to learn about food storage and usage tips/tricks:

    1. A database where they can search to look up how food should be stored

    2. A dedicated flow after completing a grocery shopping trip that walks them through storing that food at home

    3. Notifications and alerts when food in their pantry is getting close to expiration

    4. Tips and tricks for alternative ways to use up food should they find themselves unable to cook the meal they had planned.

Research Methods and Findings

Methods

  • Literature review of consumer shopping habits and food waste

  • Competitive Analysis of apps in the market

  • Interviews with prospective users

Findings

  • A shopping plan is the most effective in terms of reducing food waste

  • Saving time and money are what millennials value when it comes to grocery shopping

  • Millennials are unsure about what to do to curb waste and would appreciate clear instructions from credible sources

  • There are many apps that promise to help reduce food waste, including those that help users shop effectively, donate unwanted food, use up expiring food, and store food optimally. However, none are based on a foundation of proven research.

Personas

Findings from semi-structured interviews and surveys corresponded with findings from our initial research, that poor planning, overbuying, improper storage contribute to food waste. We developed two personas to help guide our design decisions: a primary persona and a supplementary persona. Michelle, our primary persona, is a millennial and reflects the significant design points we hope to address as our target user. We also concluded that the participation of grocery stores is crucial for maximum effectiveness in curbing food waste. We identified Whole Foods as our supplementary persona, as their goals of a sustainable future align with ours regarding food waste.

Ideation and Feature Definition

Ideation and feature definition were based on the user story of our persona, Michelle. By envisioning her flow as she currently interacts with food, we identified points where she can be nudged to make a different choice. These changes in behavior applied from our research findings.

User Flow

Currently, Michelle shops for groceries once a week without planning ahead. Upon entering the store, she picks up a shopping cart and takes her usual route but finds that she needs to revisit aisles multiple times, which wastes time. Michelle ends up purchasing too much food since she decides what to buy on the spot. When she returns home, she stores food inefficiently, which shortens the shelf-life of certain items. A week later, she finds herself tossing away a good amount of produce she previously purchased.

 

The ideal flow encourages Michelle to plan for her grocery trip. When she arrives at the store, she takes an optimized route to save time. She only grabs what she is able to finish, and she learns to optimally store food to increase the shelf-life.

Legend

text

Where nudges should occur for behavior change

Where current behavior should be blocked for behavior change to occur

Ideal behavior

Interactive Prototype

The following outlines the top four scenarios designed in our interactive prototype. Pertinent screens are annotated with supporting behavior change theories. 

Scenario 1: First time Using the app and creating grocery list

We are applying facets of the TPB and HBM theories to get our persona started with the app. The onboarding introduction reinforces the idea that this is a positive behavior that the user has control over to make a change (TPB). It additionally raises awareness of the severity of the food waste problem (HBM).

Michelle hates to plan ahead for shopping or spend time figuring out what to cook. The “Plan Your Shopping Trip” feature takes the effort out of making a grocery list by removing barriers (HBM) of planning a shopping list and wasting time wandering the grocery store without a plan. She simply inputs relevant details, and the app then auto-generates recipe suggestions.

When prompted to pick her meals, the app lets her know if she is being too ambitious by trying to cook more meals than she probably would. Playing on the food pun trend throughout the app, a “slow your roll” popup appears, warning Michelle to only plan for what she will be able to cook for the week and keeping her in check.

Scenario 2: Using the app while grocery shopping

The app generates the most optimized route on a map to purchase the grocery items. Since Michelle has a busy schedule, providing an optimized route to follow with the time and number of items to purchase helps her save time. This removes the barrier of figuring out where to go to find produce items (HBM), and she no longer needs to rush since she knows the amount of time needed at the store.

When Michelle checks off an item from her grocery list, a pop-up displays information on that item. From tracking what foods she previously did and did not finish, it gives her actionable tips on how to finish the item. This forces her to think about her choices and wastefulness. If she previously used up that grocery item, the card will thank her for consuming the amount she buys, which increases her self-efficacy that she can reduce waste (SCT).

Finally, this screen confirms that she’s completed her list and again reminds her of the importance of not wasting food. If she used the app to meal prep and create a grocery list, she can share  what she plans to cook for the week along with her contribution to reduce food waste. By sharing a goal publicly, Michelle is more likely to complete it and remember to use her groceries (SCT).

Scenario 3: User completes shopping and comes home to store food

When Michelle arrives home, the app's enabled location services sends her a push notification to learn how to store food optimally. By reminding her as soon as she arrives home, she is prevented from putting away groceries as per her usual routine. It is a nudge that tells her there is a correct way to store food for maximum freshness and that the app can give her that information if she decides to follow through on it (cue to action from HBM).

An alternative way to enter the Pantry section is to go through the app from the phone home screen. The app icon shows a passive alert to Michelle that there is an important notification. This is in case she decides to turn off push notification in her phone settings. Participants in our user study indicated that too many notifications can be annoying so there is a chance that users may turn it off. The passive notification provides a backup nudge to check the app (cue to action).

The home screen alerts Michelle to store recently bought items. The storage info is on top of the screen because it should take priority at this point in the scenario. Prioritizing storage tips gives Michelle a nudge to store items properly (cue to action) and an easy way to perceive which actions she should do next (lowers barriers, HBM).

When she goes to the Pantry, she is faced with a screen that shows all the items she has just bought. She can tap on any item to see storage instructions and shelf life information of each food item as popups.

The information displayed for each food item gives succinct and easy to follow instructions, which removes barriers of not knowing how to store food properly (HBM). It addresses confusion on how long to keep items and when they are considered unsafe to eat. Seeing the results of storing food properly increases perceived self-efficacy because she eventually finds less spoiled food due to improper storage. 

Scenario 4: User has been using the app for extended period

Michelle has been using the app for a period of time now. It has helped her move from a Contemplation phase to Action, with the promise of progressing further towards Maintenance (TTM). She received notifications from the app to keep her routine of planning ahead for grocery shopping. She furthermore gets alerts when food in her pantry is close to expiring so that she can take action. Again, if she turns off push notifications, she can still see alerts on the app icon to signify there’s an action that needs to be completed.

On the home page, the progress bar displays how much of the purchased grocery items are being consumed. Showing that users are saving more food over time increases self-efficacy, and displaying how close they are to the goal of using all their produce provides the necessary feedback to help them adjust their effort toward minimizing food waste. Since there are 4 items expiring soon, an alert at the top prompts Michelle to go to her pantry to indicate which items have been finished (cue to action). 

When Michelle selects “Go to Pantry,” or taps on the push notification warning her that a certain food item is about to go bad, she is brought to the pantry screen. Here, she sees expiring foods displayed at the top in the order of urgency. The use of the red caution icons and red text are further cues to action and tells her she needs to either finish her food or update her pantry to indicate whether or not she used them.

Selecting an expiring food item leads to this pop-up where Michelle can read tips on how to use her ripe tomatoes or select whether she used or tossed the item. By tracking her produce, the app displays her progress on the percentage of food she finishes and also helps her through her future grocery planning and shopping trips.

Evaluation of Effectiveness

We used 3 surveys before and after prototype evaluation to measure effectiveness of our design in promoting behavior change. We found that in general, our participants (N=10) perceived the design as effective for them.

  • Attitude Survey​​

    • 40% increase in how easy they perceived planning ahead

    • 40% increase in how easy they perceived using all their groceries.

  • Intention Survey

    • 60% increase in likelihood that participants will follow food storage guidelines

    • 50% increase in likelihood that they will use up foods before they expire.

  • SUS Survey

    • Average SUS of 80​

Reflections

As I learned about and applied theories of behavior change, I became increasingly concerned about the amount of power I wielded as a designer. (Who am I to force people to an ideal behavior? What is ideal behavior anyway?) These ethical considerations have been reinforced by recent news coverage of the actions of certain social media sites. Design for behavior change sounds great in theory, but without deep reflections on its ethical implications, I am hesitant to apply it too broadly. Nevertheless, I am grateful for the opportunity to learn about these design techniques.

Another thing I learned was applying academic research to industry design. Academic papers tend to focus on the abstract which makes it difficult to apply concepts concretely. My experience in this class and with designing the above project gave me a mechanism to understand academic sources and apply them effectively.