Self-checkout MVP
RIACHUELO
CONTEXT
As seen in convenience stores and other retailers, self-checkout is common in several places.
In order to make Riachuelo recognized as a modern company by the market, the company's leaders wanted to implement self-checkout in stores.
But the company didn't know what customers' pain was solving or what opportunity was embracing - just considering technical feasibility and considering that it might be good for the business.
Sunmi self-checkout, device used for Riachuelo MVP
PROBLEM
The hypothesis of problem that we had was:
“our queue experience is terrible and generates a lot of dissatisfaction.”
So, a good checkout experience would reduce queues and thereby contribute to customer happiness.
PREMISSES
-
RFID use in the product scan - technology should be implemented in 80% of the portfolio in the same year
-
Prioritize card payments - in Brazil, 70% of people still prefer cash payment, but it raises fraud risk
-
Riachuelo customers can't depend on the installation of any app nor should have any product offers during the self-checkout experience
OUR GOAL
CHALLENGE
How might we give full autonomy and convenience to people when paying in stores?
MAIN FINDINGS
#1
Self-checkout can be useful for convenience and cashless payments, but it won't increase satisfaction in the queue because it's theoretically good, either because we didn't identify dissatisfied people or because it's not really a problem.
-
Apparently, there was no problem in the store queues. Data from 14 stores showed that the average waiting time in the queue was less than 2 minutes, and the average queue and payment time was 3m50s.
-
When we visited 5 Riachuelo stores, we talked to employees and watched the flow inside the establishments. There were no complaints about them either.
-
During the same visits, we observed that the store's cashier concentrated a lot of activities. Paying for purchases, paying bills in general, paying off debts, hiring cards ... All store transactions were directed there, concentrating more employees and customers in one place, even with mobile payment points.
-
In the store's satisfaction survey, the queue's time score was 4.46 out of 5. An unreliable. note, since only those who made a purchase answered, so a detractor would probably have been discarded before.
-
Studies on Amazon cite its camera payment solution as a promoter of convenience. The Chinese market also talks about ultra convenience with Scan & Go and self-checkout solutions, as in the Ali group's Hema market. The big barrier for Brazil, in this case, is that about 70% of people still prefer to use cash in purchases, something to be considered to make the product desirable.
Riachuelo stores
Purchase Journey Today
#2
Our self-checkout needs to give our customers enough autonomy.
If there's any possibility of human support obligation, it will already demonstrate that we need to improve on something.
-
Desk research indicates that robotization can generate aversion from customers and employees. It can seem like part of a robotization and can generate some aversion, both for customers and employees. However that doesn't mean that there will be fewer people working, but a reallocation of people in what makes the experience better. Walmart and Giant Eagle, for example, started working with floor cleaning robots. While their employees became free to help the customers, instead of repetitive tasks. So, instead of thinking about self-checkout as a way to automate the store, we needed to think of it as a solution to make human service more focused on listening to human problems and helping them.
-
When we visited 7 stores that already have self-checkout, most of them still had an attendant to help, and this could happen for 3 reasons. The stores:
-
needed employees to help the customer to avoid mistakes
-
thought that active customer support would be just a good idea
-
created an ineffective solution.
-
-
We identified other problems in the experiences of competitors that made self-checkout potentially more time consuming than common cashier:
-
payment machine too inclined made reading difficult for some users
-
low-quality touch screen lets call to actions selections difficult
-
plastic bags might be difficult to open because of the material
-
product tag removal might be confusing to take off
-
some common cashiers were too close to self-checkouts, in a way that hindered the shoppers' path
-
Decathlon
Renner
Decathlon
Renner
C&A
Given that the queue isn't a problem and convenience can be a great value proposition for shoppers (due to rational or irrational reasons), it's likely that the main gain with the solution would be increasing the number of shoppers and not increase satisfaction with the queue experience.
Our expectation was that more people would buy than last year and since we didn't have data to compare, our challenge was to answer: will self-checkout increase the percentage of people who bought in the store compared to the number of people who entered the store?
Defined metrics:
-
amount of people entering the store vs. amount of people entering the store and making a purchase
-
amount of people who deviate from the common queue to go to the self-checkout
-
self-checkout queue time must be less than 1min30sec
SOLUTION
I considered two trends mapped by Trendwatching, taking into account pre and post-pandemic scenarios:
AMBIENT WELLNESS - maximum health benefits with minimal effort
Reflection on how people interact with the physical space and their impact on all physical and mental aspects
A-COMMERCE - the era of automated commerce after m-commerce
Companies will use this moment to automate services but thinking about convenience. Advances in robotics enable a new generation of commerce, so we can leave humans free to do what they do best.
I also considered the challenge's premises, the brand's tone of voice, store communication, good ergonomics, and accessibility practices.
After mapping the journey using a proto persona and the store's pains, we prototyped our MVP proposal and went to the test.
MAIN FEATURES:
-
RFID product reading, without having to pass the labels one by one
-
package size suggestion
-
button to call an attendant if you need help
-
payment by gift, debit, or credit cards
-
installment payment options (just like Klarna)
Self-checkout concept
TESTS
Because of the COVID-19 pandemic, we opted to do a mediated online test with 9 people. At this time, my team conducted the research, but I validated the questions and research goals.
We observed:
-
general reaction on the concept, need on the concept and intention to use considering some tasks;
-
probability of use in specific situations.
TARGETS:
-
people shopping in physical stores and e-commerce
-
people who shop only at physical stores
-
people who shop online most of the time or always
"And the store needs to have the packaging options there, you know, and different sizes too. What if I'm buying a gift? How do I do? I think I could have those simple gift packages on the side to get ... like a deposit envelope at the bank."
Leandro, 23 y, e-commerce + store shopper
"I understood that the card's make-up will be there, right, that it doesn't even have at McDonald's, for example. I just didn't understand well how it will work to scan items, but it's worth."
Rafael 35 y, e-commerce + store shopper
"The important thing is to hide the ID, I don't want to expose this information to the whole store, you never know who's watching. But in general, it's good."
Manuela, 22 y, e-commerce shopper
"What worries me is the clothes tag... You know, what makes the alarm go off at the door. You understand that, for black people, sometimes they already have a lot of negative eyes, imagine if the alarm goes off!"
Leandro, 23 y, e-commerce + store shopper
WHAT CAME UP
-
None of the customers identified the gift package icon.
-
Only one customer identified the “ask for help” options and believed that clicking an employee would assist them.
-
Typing some information on a large screen, such as the document number, caused discomfort in 5 of the 9 respondents.
-
There was doubt about the scanning part of the process, not least because the prototype was not tested on the official devices.
-
A possible barrier to the experience is not guaranteeing that the product tag has been demagnetized, which can cause inconvenience when leaving the store. BIPOC people raised the discomfort.
The SUS score was 76.4 and we made adjustments considering:
-
the size of the help button
-
the gift package option
-
RFID was disregarded in this first version. The technology needed more tests to make sure it would work well.
Future tests in stores would be necessary to validate:
-
findability and package choices value when in the final furniture
-
easy scanning of items
-
RFID security to demagnetize items
With adaptations aimed at product scanning and payment steps, MVP was launched at the concept store in São Paulo in December 2020.
TEAM
Marina Domingues - product designer and UX Lead
Marcus Bianchini - tribe leader
Celso Arruda - tribe leader
Mario Baumgartner - group product manager
Raphael Abe - product manager
Guilherme Gaspar - tech lead
Jefferson Tavares - IT developer
Vanessa Nonato - PoS product designer
Pedro Martins - researcher
Operations Team
Store Architecture Team
MY ROLE
Responsible for the product MVP definition and validation based on the customer's problem diagnosis that Riachuelo would solve based on research. I also facilitated one co-creation session and prototyped the first version of the defined solution. Then, as a lead, I directed the research with the team's researcher and the adaptations on the screen for the viability of the MVP.