Project developed in a week as part of a design challenge for a fictional international supermarket chain. In my view, we don’t ever have to reinvent the wheel, but there are millions of ways to use a wheel. That’s how I approached this project. I got a very specific brief requirement and a market that already has multiple solutions to the problems individually. It was less about creating something new and more about how to use existing solutions better.
Supermarket checkout automation can lead to job losses and create accessibility hurdles.
Neomarket is a fictional international chain that has arrived in Brazil. One of the company’s main pillars is accessibility. There’s also a focus o positive social impact.
For this design challenge, I was alone. No developer, no design partner, no manager. I had to design in isolation, which goes against one of my design principles. But with control out of my hands, I had to adapt.
- Implement automated checkouts at the supermarket
- Keep accessibility at the core of the project
- Create positive social impact
- Maintain the chain’s ethos of technology an innovation
- Avoid common accessibility problems
- Automation without job loss
- Current market technology limitations
- Designing in isolation
A simple, accessible and integrated system comprised of Automated checkout, phone app, self shopping scanner, delivery. Implement without negative job impact and with accessibility at its heart.
The only person in the project. Being a design challenge there was no possibility for collaboration and multiple factors were down to assumptions.
What it is
Bespoke self-checkout built with accessibility in mind. Physical buttons, an audio option, were two of the main motivator to steer away from the industry standard of touch screen only.
A system implemented via App and specific hardware. Connected to the automated checkout, providing convenience, flexibility and more agility at checkout. Shopping is tranfered to the automated checkout for payment.
Team training to assist people with impairments, support in the use of automated checkout. Product selection in-store for online shopping/deliveries and internal navigation training to support customers.
Phone App that uses internal location systems and helps users to guide themselves via AR, audio or map. The function can be used to identify product shelves, aisles, paths to be followed within the unit.
Delivery system implementation with stores as distribution centres for fast deliveries, job retainment and generation.
Create and edit shopping lists that can be used for delivery orders and in-unit purchases. For in-store use the list can be used in conjunction with the scanner and built-in guide functions, plotting the best way to purchase.
Highly iterative and fast process. Time constraints led to a lot of what was done. Research time was prioritised and was the task with the largest amount of time allocated.
How I worked
Systems thinking was paramount for this project. One of my first steps was to understand the supermarket model, market and trends in the sector. Crossing those pieces of information with innovations around the world gave me an initial view of the system around supermarkets. It was challenging to learn about the business, its clients and technologies in such a short time. So I tried the create a more high-level view of the whole picture.
Since this was an exercise with little initial information give I made some assumptions to create a more complete picture.
- Large budget for technological implementation
- First stores deployed in larged, more developed, cities
- Custom made hardware is possible
I created a wiki inside notion similar to . It includes
- Project management
- Project diary: ideas, decisions, questions, task log
- Personal diary: Daily planning, reflections and reviews
- Service and project documentation
- Research database
- Project review
What do I need to know? That was my first question, so naturally one of my first steps was to do market research. I included trends, growth, job creation, main chains, reach, business models, consumers habits. Looked into local and international statistics, how they compared. This way I could get a view of the wider context, how do supermarkets work? What are their margins? What’s the forecast for the sector? There was a lot a didn't know, so I needed to fill in the gaps.
I looked into de delivery market for groceries specifically. This was due to an early insight that delivery could, or not, have a role in the final solution. Supermarkets own delivery systems vs. third party deliveries. Pay scales and differences, advantages, costs. This could be a way to deliver a positive social impact.
Since automation was an integral part of any proposed solution I dedicated some time to understanding automation options, job impact, cost, manufacturers, interfaces. Essential information for any decision making regarding the project. Were there solutions that didn't disrupt jobs? Is automation a good alternative for every customer?
I watched videos of people with different types and levels of impairment inside supermarkets. 100% via youtube due to 7 days time constraints, and pandemic, constraints. That way I could get as first-hand information as possible in the circumstances I was in. I took a lot of time to organize common complaints and stark differences between them. My idea was to maximize what I could get from the information provided. How to address the overlapping pain points and the diverging ones with little to no confusion?
Based on the limitations of the research further assumptions had to be made regarding user behaviours and wider preference. For this project, I decided to take the words of the people in the YouTube videos as the wider truth in terms of accessibility. I couldn't find conclusive data on partial automation impact on jobs inside supermarkets, so one of the assumptions I made was that there is an impact and my solution should account for that.
Through research, it became obvious that I needed to accommodate very different use cases. But how to do it? Initially, a drew the best scenario I could imagine for all the identified needs, mainly using solutions that already existed in the market. I took three solutions that were already used inside supermarkets - automated checkout, shop & scan, delivery - and made them work in conjunction.
Here is where I identified redundancies in the solution ideas. Flow creation for all contact points, keeping them aligning between different platforms for a seamless experience when switching between modes. A visual representation of all the ideas gave a clearer vision of what could be combined, refined or abandoned.
- Automated checkout, as required by the initial briefing
- The connected phone app and shop & scan, leveraging the accessibility feature built-in devices customers are already used to.
- Keeping a legacy checkout area
- Implementing a self managed delivery system for wider control and better pay
Flow creation for all contact points, keeping them aligning between different platforms for a seamless experience when switching between modes.
My reasoning was to simplify and speed the process. Also, given the conditions of the project.
- Short project timeline, 7 days. Cutting any recruiting, actual testing, would cut time.
- Pandemic made it harder to observe users.
- Larger user pool, being to observe shopping in other countries.
- Even with cultural differences, supermarkets have a lot in common.
In some cases, nothing is more accessible than speaking with another human. This way I could also maintain jobs and give an option for clients who might prefer traditional checkouts.
This way I could keep accessibility as my top priority by creating a tailor-made solution. The higher cost is concentrated in the initial phase of implementation and potential gains in business, brand profile, should offset that in the long term.
Better and more comprehensive research. Larger focus on the local market.
Live testing solutions were proposed to acquire more data to discover flaws and further necessary improvements.
Is it viable to implement custom hardware as proposed?
In many phases of the process, I felt the need to have a different perspective, counterpoints to mine, new ideas. Not being able to talk to any developer left me resorting to assumptions about implementation feasibility.
Again resorting to assumptions to fill in gaps that couldn't be filled due to the nature of the design challenge. That gave a clearer vision of how much teamwork contributes to better results.
Exercise vs. Real project
In a real project, time would be much greater than seven days, it would be possible to have a roadmap, time allocated to different phases, test runs.
A team of designers for starters. Developers, PM, PO, data scientists. In a real project, a real team could have many different forms depending on the budget, company. But for certain I wouldn't be the only person.
Time is dedicated for research to come up with solutions, to study the market, its consumers, to test out solutions. Actual data about an actual supermarket chain, finances, target audience, current technologies used.
The proposed solution and products have expected results. They were crafted to deliver more value and better the experience.
An individual jack of all trades. Through the phone app is possible to shop, navigate the store, shop and scan, have audio guidance.
Dedicated hardware to Shop & Scan.
Devices that meet accessibility standards and work in parallel with traditional checkout.
Different options to fulfil different shopping preferences and needs. A more flexible, efficient and inclusive supermarket that becomes the preferred option due to its high level of service.
Active effort to retain, and create jobs, in other areas of the business to compensate for possible automation losses. Training former cashiers to work in other positions. Increased hiring keeping up with possible growth in deliveries demand.