EcoTrolley: Nudging Consumers Towards More Sustainable Food Choices

Design Process

01. Empathize

Background Research

To understand the context surrounding food production and consumption, we looked into background research and existing literature. Three main areas of focus were explored during background research and literature review: 1) consumer food decisions, 2) food sustainability, and 3) methods for informing and incentivising.

For the first area, we questioned where and how consumers make decisions surrounding food. There are many factors that consumers may consider when making a food choice, from cost and convenience to food quality and country of origin. There are also many channels through which consumers purchase food. Although UK consumers continue to buy from large supermarkets, other forms of retailer, including mini supermarkets, have increased in popularity. Also, while consumers continue to shop at supermarkets, there are other increasingly important avenues through which consumers purchase food, like the online grocery market and online food delivery platforms.There has additionally been significant growth in the proportion of individuals eating out on a regular basis. All of these channels represent different opportunities for consumers to make food decisions, and thus opportunities for us to encourage sustainable decision-making.

We also investigated what makes a food item sustainable or unsustainable, as that is the precise information that we wish to communicate to consumers. Various metrics exist by which to potentially rate and rank products on how environmentally friendly they are, such as their carbon footprint, water usage, and miles travelled. There are also economic and social impacts to acknowledge when considering sustainability broadly but we decided to focus mainly on environmental impacts as these would be easier to quantify and communicate to consumers.

With regard to this communication, we finally looked into various methods and features employed to inform users of a concept and incentivize them to behave in a certain way. An increasingly popular example of this is through gamification with the use of motivational affordances, like points, leaderboards, and badges, that have been shown to produce positive effects on psychological (like motivation and engagement) and behavioural outcomes. In order for our solution to have an impact, we need to be able to influence consumers to not only think differently but act accordingly too.

User Surveys

For exploratory research, my team conducted a survey to understand consumers, focusing on demographics, existing food habits, sustainability practices, and incentives. The survey was drafted and distributed via Microsoft Forms and received a total of 36 responses.

As a team, we structured the survey to include questions on demographic, existing food habits, food sustainability knowledge and practices, and opinion of potential sustainability incentives.

02. Define

Personas

To further synthesize and understand insights of users’ experience, we created four personas based on the various features and behavioural traits represented in the questionnaire responses. Each persona represented a different supermarket shopper and their values, like Alex the 55 year-old stay-at-home parent on a budget, or Matt the 24 year old eco-conscious vegan. For each persona, we discussed in detail scenarios describing their potential experiences within a supermarket in order to contextualise these potential users and their interactions within the chosen environment.

Design Requirements

After creating personas, we finalized a list of requirements based on the research we did up to this point.

03.Ideate

Sketches

We ran a Crazy 8's brainstorming session to generate a wide range of design sketches within a 8-minute time limit, generating a total of 40 ideas. Using an initial checklist of basic criteria (like whether the design is ethical, feasible, has any clear limitations etc.) and merging similar ideas, we narrowed our initial forty ideas into 17, which we then cut to seven using a 2x2 matrix with the criteria “Persuasive” and “Innovative”.

These ideas ranged widely in execution and use of technology and so we created an evaluation matrix with the above two criteria plus “Engaging”, “Intuitive”, “Accessible”, and “Feasible”, with the criteria weighted based on priority. The design idea that scored most highly, by far, was a smart shopping cart and paired app that recognises food items and presents sustainability information/options to the user. We then created a storyboard of this idea to illustrate the design in practice and ensure we were all on the same page. As individuals with experience shopping in supermarkets ourselves, we discussed how the design might integrate into the environment and possible interactions users may have with it.

04. Prototyping & Usability Testing

Mid-Fidelity Prototype & First Round of Usability Testing

After discussing layouts and general components of the interfaces with the team, during which we developed detailed paper sketches, I worked with the Ella, another product designer, on creating mid-fidelity prototypes using Figma of the trolley screen (which I designed) and the app screen (Ella's focus). These artefacts formed the basis of our initial round of usability testing. We recruited 7 participants and provided them with a scenario and a set of tasks to complete using the interface while using a think-aloud protocol. These tasks ranged from using the app to set a goal, to looking for alternatives to an unsustainable food item, to using a voucher earned with points in the app. These tests yielded a great deal of feedback.

For the trolley, a key area of feedback was related to the process of seeing more sustainable alternatives. For example, one participant mentioned that “the pop-out asking me ’are you sure’ makes me feel uncomfortable as the language makes it feels like it is confronting me for a decision that I made unintentionally.” Four other participants reported similar feelings and suggested having a ’See Alternatives’ button on the same screen as the unsustainable product. For the app, participants had a range of suggestions, such as having clear instructions for the entire process of using the system, more visualisations of their sustainable impacts, and an explanation of the data sources used to determine sustainability, all of which were taken into consideration for the next stage of prototyping. Furthermore, participants mentioned that setting a points goal for each shopping trip could be confusing and they would prefer a monthly challenge set by the app instead.

Based on these findings, I made corresponding refinements to the in-store app design. For example, a simple change I made to improve users’ experience was to move the QR Code scanner to the right side. This was in order to make the interaction more intuitive and ergonomically friendly for the majority of the population who are right-handed. Also, an app onboarding process was created for the purpose of making the entire system clear and easy to understand. Moreover, a new and different version of the ’See Alternatives’ page was designed based on participants’ feedback so it could be compared with the original pop-out in the next usability test. In order to gain feedback on this refined design, I created high-fidelity prototypes of the screens updated with the new changes. We also created a physical prototype of the system by combining a real trolley with our screen prototypes in order to get feedback on the ergonomics and practicality of our design.

Finally, I took the initiative to build a 3D model of the trolley to further clarify the earlier sketches and give users a firmer understanding of how the fully-integrated system would look.

Second Round of Usability Testing

Using these materials, we conducted another round of usability testing with 4 participants recruited. Participants were guided through a scenario asking them to shop in a set-up supermarket environment with a modified trolley that had an iPad attached to its handlebar displaying the screen prototype.

Comments on the refinements from the mid-fidelity prototype were positive. Participants found the app onboarding process clear and helpful and additionally reported that they enjoyed the impact breakdown section on the app.

All four participants mentioned that they appreciate how the app visualised their actual impact, and all stated that the feature would definitely encourage them to shop more sustainably.

A variety of feedback was collected related to the ’See Alternatives’ process. Participants generally preferred the new prototype which had the ’See Alternatives’ options on the same page, and so the pop-out version was dropped. Additionally, participants also mentioned a need to compare the alternative and the original product at the same time to avoid having to retain that information in memory between screens.

Finally, three participants suggested having alternative options presented not only for unsustainable products but also for those considered neutral.

Several comments were also made on the clarity of the information on screen when the trolley is in use, like ensuring text elements are large enough to be read from a distance, and giving the charts on the app clearer labelling.

From this evaluation, we made several smaller improvements to the design. For instance, I added a comparison option after selecting the alternative products, allowing consumers to see both the nutrition and environmental details of the products to decide whether to replace the original item or not. Also, the language used throughout the app and the trolley screen was reviewed to ensure that it is friendly and constantly offers positive reinforcement. Finally, the visual components on screen were assessed and altered to ensure the words were clear, prominent, and easy to read from a distance, and any graphics used were consistent and followed standard heuristics in order to maintain an usable and engaging aesthetic.

06. Final Product

Walkthrough

On the first screen, the user scans the account QR code to log in. The next screen prompts the user to put the item in the cart. The first item is sustainable, and the screen shows where the product is made, how far it traveled, and the emissions. It also awards the user five points. Then the next item is put into the cart. This item is not sustainable, and takes away points, so there is an option to replace it with an alternative. The alternative comparison screen compares the original item to the replacement in terms of sustainability. After confirming, the user is prompted to pick up the item in the right aisle. This screen can be minimized if the user wants to continue shopping. After the item is collected, we can continue onto the next item. This item is neutral, but there are still alternative suggestions if the user wants something more sustainable. We can continue to check out. An activated voucher can be applied, and then the checkout is completed.

07. Reflection

Although the feedback on the EcoTrolley design has been predominantly positive, due to various constraints we have not been able to develop and test a fully-working prototype. In order to get a full and valid understanding of the system’s usability and potential, a future test in-situ at a supermarket with a comprehensive functioning version of EcoTrolley should be included in further research. A more comprehensive usability test would provide us with a sense of how interactions with the EcoTrolley would integrate into the shopping experience as a whole. A longitudinal study would also be of great benefit to see if the motivation and behavioural modifications in consumer food decisions brought about by the system last over time, or if adjustments are necessary to aid long-term change. Additionally, there are several other features that could be incorporated into the design of the EcoTrolley that we considered but chose not to prototype at this moment. For example, making the screen on the handlebar tiltable to allow users of different heights and those with wheelchairs/other mobility aids to adjust the system ergonomically. Also, a voice assistant could be integrated into the system to help those with visual impairment or information processing issues. As well as accessibility modifications, other potential features involve amplifying the gamification aspect of the design like incorporating points challenges and badges or potentially even a public leaderboard. Another extension to this design would be to create a similar product based around a shopping basket for consumers making smaller grocery shops.

Personally, this project was really exciting as the first major project completed as part of my MSc in HCI. I felt grateful to have worked with teammates from different backgrounds and with different areas of expertise—ranging from psychology, computer science, product design. Together our combined skill set and diverse perspectives contributed to a project that felt well-rounded and rigorous. I felt engaged throughout the project and drew upon design and research methods that I was experienced as well as new practices that I learned in the lectures that term. Furthermore, the design brief following the UN Sustainable Development Goals posed an interesting challenge that allowed a lot of exploration and encouraged us to create a design solution with impact.