Amazon Online Grocery
Usability Research and Proposal of Improvements on AmazonFresh
Illustration of part of prototype
Overview
Name: AmazonFresh research and improvements
Role: Project Co-Lead, UX Designer, Researcher, Presenter
Software: Figma, Adobe XD, Illustrator, Photoshop, LucidChart
Year: 2021 Fall
Through a semester of teamwork, we worked for the manager of Amazon online grocery to smooth its user experience. We identified the existing problems, challenges, and opportunities, and concluded five design decisions that facilitate the online shopping experience on AmazonFresh.
Process Map
Click the icon to jump to the step directly
Conceptual Model:
-
Current Setting
Problems & Goals:
-
Problems
-
Redesign goals
.png)
User Research:
-
Literature Review
-
Survey
-
Interview
-
Persona
Task Analysis & Use Case:
-
User Course
-
User Scenario
-
HTA Chart
Design Decision:
-
Decision 1-5 with 2 alternatives for each decision
1.0|Conceptual Model
1.1 Current setting
Screenshot of AmazonFresh
Note: The AmazonFresh top navigation bar and submenu of categorical options are nearly identical to Amazon.com. The major distinguishing features are a slightly different color palette and the AmazonFresh logo.
AmazonFresh is accessible to a broad demographic. AmazonFrsh sits within Amazon's main UI. As a result, it has similar features and functionality, as the above screenshot showed. The interface has not changed its appearance or usability. The Amazon platform ignores many modern UX/UI best practices. In spite of this, users are accustomed to the navigation and layout.
AmazonFresh and other grocery retailers struggle with fresh food. AmazonFresh’s interface does not allow shoppers to visualize or customize produce. Online grocery shoppers are often disappointed with their produce (Clark, 2020). Many customers have complained about not receiving the produce they wanted from AmazonFresh (Sitejabber, 2021).
AmazonFresh’s current interface has four serious design flaws as it relates to fresh food.
2.0|Problems & Goals
2.1 Problem I
Unappealing and uninformative images
Fresh food images are unappealing and uninformative. Produce is not placed in context (Figure 1). Images are cropped from backgrounds. In grocery stores, customers see produce items situated with other food. Avocados are not suspended in a white background in brick-and-mortar stores. They are organized by size and placed with complementary foods.
Customers do not have adequate visualizations of size on AmazonFresh (Figure 2). For example, avocados classified as medium appear to be bigger than ones classified as large. This lack of context for size can lead to incorrect purchases. Different produce items have similar or identical images (Figure 3). For example, the images of non-organic and organic strawberries are identical. This can lead to confusion for shoppers. The similarity of images makes locating desired items more difficult.
Figure 1: the screenshot of different strawberries products
Note: The same image of a bowl of strawberries is used for two separate products, non-organic strawberries and organic strawberries, on AmazonFresh. These images were accessed on September 9, 2021.
Figure 2: the screenshot of different avocado products
Note: The image on the left is supposed to represent a medium-sized and the image on the right is supposed to represent a large-sized avocado. However, the avocado on the left is pictured as larger than the one on the right. They also represent different stages of ripeness even though customers currently have no choice in this aspect of produce. These images could confuse customers.
Figure 3: the screenshot of flower bouquets on Good Eggs
Note: Having people hold the bouquets for sale provides context about size for the customer and situates the item in a real-world setting. The images are more visually appealing and interesting than a white background.
2.2 Problem II
Inability to customize selection
AmazonFresh does not provide a way for customers to fine-tune their selections. There are no images of the possible ripeness of produce items. For example, the images of bananas do not have spots. They are entirely yellow (Figure 4). Dairy items do not display expiration dates. Customers have no way of selecting the freshness of their produce or dairy (Figure 5).
Figure 4: the screenshot of bananas
Note: People have strong preferences about the ripeness of their bananas and yet no way to customize or even visualize options.
Figure 5: the screenshot of milk
Note: There is not an expiration date on images of milk where one would typically expect to find them.
2.3 Problem III
Lack of Communication
AmazonFresh shoppers cannot communicate with employees to make comments or ask questions. It is important for customers to be able to send instructions about items in their orders. Online grocery shoppers cannot stand in the aisle and select produce by sight, touch, or smell. They depend on an employee to do that. The inability to add instructions or chat eliminates the possibility of re-creating an in-store shopping experience. AmazonFresh competitors have implemented this technology with success (Figure 6).
Figure 6: Instructions for shoppers on the Instacart app
Note: Instacart customers can provide written instructions and images for their shoppers. These features allow customers to have more control over their grocery orders.
2.4 Problem IV
Generic shopping experience
AmazonFresh customers do not have a tailored shopping experience. Customers expect more from online grocery retailers than they did in the past (Hall, 2019). AmazonFresh does not inspire shoppers with complementary food items. They do not offer suggested ways to use food. Online shoppers now appreciate the smart marketing techniques of most online retailers.
2.5 Redesign Goals
This design project aims to recreate an in-person shopping experience on AmazonFresh. We limited the design project to four areas of the AmazonFresh user interface.
Risks‼️
Challenges💪
✅ Customize fresh food selection.
✅ Increase appeal of fresh food images.
✅ Provide a tailored shopping experience.
✅ Enhance communication between shoppers and employees.
The main challenge in this redesign is adequately recreating the experience of in-person grocery shopping. It is impossible for customers to feel or smell produce from their computers or phones. Customers overwhelmingly report wanting to pick out their produce (FATbit Blog, 2017).
There is a risk in implementing the redesigned interface. AmazonFresh has to overcome customers’ negative experiences with purchasing produce. The proposed redesign will require time and labor. Customers might not be willing to try the redesigned interface. The redesign would not have a high return on investment. There is also the risk that most new customers will return to stores after the pandemic.
3.0|User Research
3.1 Literature Review
Current state and Future of Online Grocery Shopping
COVID-19 has increased the number of online grocery shoppers. According to Brick Meets Click (2020) online grocery sales increased by 233% in early 2020. The number of online shoppers grew by 146% from their previous survey in August 2019. There has been a slight decrease in online shopping since the early days of COVID-19. Industry experts believe grocery e-commerce will remain popular after the pandemic. (Packaged Facts, 2021).
Demographics of Online Grocery Shoppers
Market reports tend to refer to age groups by generation.
-
Millennials: 25-39
-
Generation X: 40-55
-
Baby Boomers: 55-75
-
Generation Z: 18-24
Millennials and Generation X are the target age groups. They are comfortable with online shopping and value convenience. Millennials are more willing to pay delivery fees (Kodali & Cyr, 2020).
In spite of fees, Baby Boomers have started shopping online. This is a direct result of the pandemic. Many Baby Boomers report they will shop online post-pandemic (Packaged Facts, 2021). Baby Boomers are important to consider in grocery e-commerce. They tend to be more loyal shoppers and have more disposable income. Grocery delivery alleviates mobility issues (Bhattarai, 2021). Generation Zers do not yet have much disposable income. Online grocers still need to market to them. At some point, they will become more established in their careers. They are also already savvy with technology. They will then turn into online grocery shopping (Packaged Facts, 2021).
Gender is not a predictor of online grocery shopping (Packaged Facts, 2021). However, men are far more likely than women to use a single app. They do not shop around on different apps like women (Sprinkle, 2019).
Income and education level are influential factors. Households bringing in more than $50,000 are the grocery e-commerce target. Those earning more than $150,000 are even more likely to be customers. Income correlates with education. Shoppers with higher education degrees are the core users (Packaged Facts, 2021)
Diet predicts the likelihood of online grocery shopping (Packaged Facts). Plant-forward consumers participate in grocery e-commerce more often. Nutrition label readers are also online grocery shoppers (Owen, 2020).
Households with children shop online more frequently (Packaged Facts, 2021). However, this could just be a result of age. More Millennials and Generation Xers have children at home.
Attitudes Towards Online Grocery Shopping
A third of shoppers find online grocery shopping frustrating (Kodali & Cyr, 2020). Consumers are often disappointed with their produce. There are also concerns with data security. They appreciate the time saved with grocery e-commerce. Online purchase histories are also rated positively (Statista, 2020).
Customization is important to online shoppers (Demeritt 2021). Consumers expect products and services tailored to their needs (MarketLine, 2020).
3.2 Survey
We wanted to investigate potential users' grocery shopping habits. We conducted a simple eleven-question survey using Google Forms. Our questions focused on user demographics and behaviors. Our anonymous survey recorded 71 responses.
Survey results revealed that most of our respondents are within the first age group, 18-24 (98.6%). The vast majority reside in North America (84.5%) (Figure 7). They have used online grocery platforms at least once (77.5%). These results might not indicate the majority age demographic. We sent the survey to direct connections such as friends and family.
The majority of respondents have grocery shopped online. We do not have an explanation for the 25% who have not. Further research should address their motivations.
We asked about online platforms and physical store preferences. There is variation among preferences for shopping platforms (Figures 8 and 9). Amazon and Whole Foods have the market share for online shoppers. Approximately a fifth of our respondents do not have a preference.
Our survey also asked about grocery shopping habits. Many respondents (43.7%) reported rarely shopping online (Figure 10). A third shop online every week or two weeks. The majority prefer in-store shopping, even during the pandemic (Figure 11).
Our survey asked about the incentives for shopping online (Figure 12). Most users cite the convenience of home delivery. Some respondents reported COVID-19 safety concerns. These results align with industry reports. Young adults are willing to pay for convenience. The pandemic has increased online shopping behavior.
The majority of respondents say they are competent with technology (Figure 13). This is most likely the result of the age of respondents.
Survey Conclusion
We can conclude that young adults are grocery shopping online. They do not prefer online grocery shopping, even during the pandemic. Young adults are tech-savvy and value convenience. They are also concerned about COVID-19 safety. These results align with current industry reports. We cannot make conclusions about other age groups. The limitation of this survey is lack of age diversity.
3.3 Interview
We conducted structured interviews in-person and online. We used a standard set of interview questions. Respondents said they had used Amazon Whole Foods, AmazonFresh, Walmart Fresh, and Instacart. Most of them prefer using a mobile device for online grocery shopping. Some respondents said they preferred the mobile interface. Others appreciate the convenience of shopping on mobile devices. Those who shop on computers do so because the images are larger.
Saving time is the greatest motivator for online grocery shopping. They appreciate the convenience grocery e-commerce affords. They also like avoiding heavy lifting. One interviewee enjoyed being able to shop at multiple stores. Online shopping allows him to find the best produce and prices quickly.
They reported pain points in online grocery shopping experiences. Pictures are cited as the most important information about an item. Interviewees stated pictures often do not provide enough information. They cited a lack of detail on produce items about ripeness. However, they need to be improved to be more informative and helpful. Interviewees reported inconsistent access to nutrition labels.
Figure 14: Illustration of the interview from Freepik
3.4 Personas
Micheal(Primary persona)
Michael is AmazonFresh’s target customer. At the age of 40, he is a cusp Generation Xer. People in these age groups are the most likely to shop online for groceries. They report being willing to pay delivery convenience fees. Michael’s annual household income is over $150,000. He has a postgraduate degree. These features make him a likely customer of AmazonFresh. As a male, he is just as likely to buy groceries online as a female. However he is more likely to only use one service. He does not shop around for the best price. This means he will continue to use the most convenient platform.
Figure 15: Persona 1 Michael
Zoey(Secondary persona I)
Zoey represents the AmazonFresh target age group Generation Z. She uses her mobile phone for nearly everything. She prefers the mobile app interface for online grocery shopping. The annual income of Zoey is $60,000, and she hasn’t bought her own car. Right now she is on a tight budget. Soon she will earn more and thus spend more. Online grocery shopping helps her avoid carrying heavy items on buses. She has used technology since she was a young child. Zoey prefers to conduct business online.
Figure 16: Persona 2 Zoey
Judy(Secondary persona II)
At the age of 70, Judy represents the Baby Boomer generation. She is not in the current target age group. She is important to consider based on the current pandemic. Judy has a fixed income of $70,000 per year. This puts her in the target income group. Baby Boomers are typically not savvy with technology. They need platforms that are straightforward. Judy tends to grocery shop at a particular store for a long time. She has brand loyalty. As COVID-19 continues, Baby Boomers can potentially be a larger demographic for Amazon fresh.
Figure 17: Persona 3 Judy
4.0|Task Analysis & Use Case
4.1.1 Course
-
The system displays Amazon Fresh page
-
The user enters “avocado” in the search bar
-
The system displays the search result of “avocado” in Amazon Fresh
-
The user browses and chooses the avocado
-
The system displays the detailed page of avocados chosen by the use
-
The user reads the description
-
The user determines how many items or how many pounds she/he wants
-
The system prompts the choice made by the users, including the name of the item and the amount of selection
-
The user makes confirmation
-
The systems add this customized selection into the shopping cart
Figure 18: Illustration from Freepik
4.1.2 Alternative Courses
1. If the user does not go to Amazon Fresh directly
1.1 The user goes to Amazon and clicks Amazon Fresh tab
1.2 The system redirects to Amazon Fresh page
2. If the user does not search for the product in the search bar
2.1 The user uses the navigation bar and locates the produce category
2.2 The system will prompt the produce page
2.3 The user browses under the “Fresh Fruit” or “Fresh Vegetables”
7. If the user wants to buy the avocados he/she bought before
7.1 The user goes to purchase records
7.2 The system prompts the purchases records page for the user
7.3 The user finds the avocados
7.4 The systems redirects to the avocado pages
7.5 The user determines how many items or how many pounds she/he wants
4.2 User Scenarios
Scenario 1
Figure 19: Pic of Michael's Persona
After finishing the last zoom meeting on Wednesday, Michael opened the browser and clicked the bookmark ‘Amazon Fresh.’ Due to COVID-19, he has been working from home since last year, and he got into the habit of shopping for groceries online. Wednesday night is the ‘grocery night’, and he planned to buy next week’s groceries for his family. He searched for ‘avocado’ and clicked the first product page. The picture on the left side of the page shows an iPhone 12 and three different-size avocados. Then he scrolled down to see more information like the nutrition table and related recipes. After checking the detailed information, he came back to the top of the web page. The page contains different customization factors to choose from: size, ripeness, and amount. Michael chose 2 large medium-ripeness avocados and 3 small high-ripeness ones, then put them into his shopping cart. After that, he came back to the search page and continued to buy other groceries.
Scenario 2
On the bus back home, Zoey opened the browser on her smartphone and typed amazonfresh.com in the URL bar to access Amazon Fresh homepage. Then she clicked on ‘Past Purchases’ in the top navigation menu and identified the avocado purchase two weeks ago. She selected 3 large slight-ripeness avocados and put them into the shopping cart.
Figure 20: Pic of Zoey's Persona
Scenario 3
As the number of COVID-19 infections continued to rise, Judy decided to continue ordering groceries from Amazon Fresh this week. She opened her desktop computer and googled ‘Amazon Fresh’ to get to the front page. She looked through the navigation menu and clicked produce. Judy wanted to make avocado sandwiches for her granddaughter, so she needed some fresh avocados. She found the ‘Fresh Fruit’ catalog and scrolled down to search for avocados. Suddenly, she saw her favorite avocado brand and clicked it straightforwardly. After viewing detailed information, she chose 5 medium-ripeness avocados and added them to the cart.
Figure 21: Pic of Judy's Persona
4.3 Hierarchical Task Analysis (HTA) Chart
We designed a hierarchical task analysis for our system to display two distinct actions that users will complete when purchasing produce on Amazon fresh. The tasks are searching for produce (via the search bar) and menu navigation. These two actions are independent of each other. Within the task analysis, we highlight alternative courses for example: if the user directly goes to the Amazon Fresh website as opposed to the main Amazon page.
Figure 22: HTA of Searching
Figure 23: HTA of Menu Navigation
Figure 22: HTA of Searching
5.0|Design Decision
5.1 Decision 1: Visualization for Customization
.png)
Figure 24: Choice 1 of Decision 1.
Figure 25: Choice 2 of Decision 1
We designed a hierarchical task analysis for our system to display two distinct actions that users will complete when purchasing produce on Amazon fresh. The tasks are searching for produce (via the search bar) and menu navigation. These two actions are independent of each other. Within the task analysis, we highlight alternative courses for example: if the user directly goes to the Amazon Fresh website as opposed to the main Amazon page.
5.0|Design Decision
5.1 Decision 1: Visualization for Customization
Description: Shoppers desire control over their online shopping experience (Demeritt, 2021). This is especially true of fresh foods. Selecting produce is the second most cited reason for shopping in stores. Customers have reported receiving subpar produce (Packaged Facts, 2021). Providing options for the ripeness and size of produce would help this problem. We will provide additional information and options for produce. Customers will choose from a spectrum of ripeness. Each option will include information about the timeline for ripening. There will also be more information about size options.
Alternative 1--Provide details about ripeness options and size in text descriptions. Include size measurements in inches and centimeters.
Figure 24: Alternative 1 of Decision 1.
Alternative 2--Provide images of the stages of ripeness. Supply an image of the food next to a common item, such as an iPhone. This will give context about size. Provide details in descriptions for those with visual impairments
Figure 25: Alternative 2 of Decision 1
Rationale: Alternative 2 is chosen because we want to enhance the shopping experience with more choices. Customers will select the ripeness and size of produce. We will provide images to customers instead of just text. Customers select produce by touching items in stores. Images can recreate this experience better than text. Industry experts agree pictures are essential (Packaged Facts, 2021). Our method of giving size context is a result of one of our interviews. The interviewee stated they would appreciate seeing a product next to an iPhone for context. Size comparisons provide benchmarks for customers (Lidwell et al., 2010). Detailed descriptions will accompany images. This will assist shoppers with visual impairments.
5.2 Decision 2: Aesthetically Appealing and Contextualized Photographs
Description: Images are essential for online grocery shopping websites. Shoppers tend to notice pictures before the plain text. Images in context help customers imagine consuming the product (Fairhurst, 2020). AmazonFresh currently uses sterile images (just contain produce itself with a white background). AmazonFresh images of food do not have backgrounds. We will make products more appealing by placing them in context.
Alternative 1--Add images of the food in the context of a recipe.
Figure 26: Alternative 1 of Decision 2.
Alternative 2--Include images of the food against white background and offer no context.
Figure 27: Alternative 2 of Decision 2
Rationale: Alternative 1 is chosen for this design decision. We will improve the experience with visually appealing images. Beautiful photos of products generate a halo effect for customers (Lidwell et. al., 2010). Aesthetically pleasing visuals make products seem more usable (Shokurova, 2020). Placing the food in the context of a dish increases the desire for the product. Two of our interviewees stressed the importance of images in search results. Another interviewee said beautiful pictures would be an improvement. An industry report supports our interviewees' claims. Fifty-eight percent of consumers have left e-commerce sites when images are not present (Stackla Report, 2021).
5.3 Decision 3: Opt-In Cross-Merchandising and Suggested Recipes
Description: Customers expect more from online grocery retailers than they did in the past (Hall, 2019). We will enhance the shopping experience by offering customers ideas about how to use a product. Customers will see other items that pair well with the food. They can learn about the best cooking methods. They will be able to discover recipes that include the food. Customers can further customize their experience by specifying their likes, dislikes, allergies, and special diets.
Alternative 1--Include images of the food in context with a link to the recipe.
Figure 28: Alternative 1 of Decision 3.
Alternative 2--Search for the name of the dish and the ingredients you need in the results.
Figure 29: Alternative 2 of Decision 3
Rationale: Alternative 1 is chosen for this design decision. We will include images of dishes with that ingredient. Customers can opt to click on the dish for more information. Providing images of the dish encourages customers to shop more. Searching by recipe or dish does not encourage the same level of exploration. We also will provide images of related foods. This type of personalization is appealing to online shoppers (MarketLine, 2020).
5.4 Decision 4: Live Chat with Shopper
Description: AmazonFresh does not have a way for customers to note preferences. Customers may need to make special requests or provide details. AmazonFresh employees cannot ask customers about replacements. Employees omit out-of-stock items from the order. Customers feel frustrated when that happens. They may have needed that ingredient in the near future.
Alternative 1--Implement customer notes without live chat.
Figure 30: Alternative 1 of Decision 4.
Alternative 2--Implement customer notes and live chat.
Figure 31: Alternative 2 of Decision 4
Rationale: Alternative 2 is chosen for this design decision. We will implement product notes and live chat. A 2020 study concluded customers benefit from talking to an employee (Singh & Söderlund). Customers in stores can access help in real-time. Online shoppers should have similar access. Live chat increases customer satisfaction (Redman, 2018). One of our interviewees stated his preference for Instacart. He can modify his order in real-time. He appreciates direct access to an employee. Some customers do not want a live chat. Implementing the notes option is important for them. This decision may bring potential cost and operational changes to the employees.
5.5 Decision 5: Expiration Date Selection with Dynamic Pricing
Description: Online grocery shoppers express frustration about expiration dates. They report receiving expired food in their orders (Cleamons, 2021). Brick and mortar grocery stores offer discounts on perishable foods. In-store customers have the ability to select the date and price. Some customers do not mind food that is close to expiring. They like the discount. We will provide shoppers with expiration date options. We will employ dynamic pricing with this feature. With dynamic pricing, the cost will decrease as the expiration date nears (Packaged Facts, 2021).
Alternative 1--Provide a dropdown menu of expiration dates. The price will change based on the selection.
Figure 32: Alternative 1 of Decision 5.
Alternative 2--Provide an expiration date in the photo. The price will not update.
Figure 33: Alternative 2 of Decision 5
Rationale: Alternative 1 is chosen for this design decision. We will implement dropdown menus with available expiration dates. The price will decrease for food with older expiration dates. Embedding the expiration date would make dynamic pricing more difficult.