Redesigning Amazon India’s Shopping App
Now, I know what you’re thinking.
“Pfffft!! Another novice designer trying to ride the wave of a big company by redesigning their app to make up for a flimsy portfolio”
But, I promise that you won’t feel that way once you reach the end of this
Why would you ever do this!!? 😱
I wanted to see how far I can push myself out of my comfort zone, creatively— by redesigning something I use regularly.
I chose Amazon because I've been a dedicated user of Amazon, (other than on the Big Billion Day Sales 😛) since it was launched in India. I started using it for the great prices and discounts it offered on books and I stayed for the great Customer Experience it gave me.
Side note: I do not work for Amazon, and the views from this case study is strictly my own and is not reflective of Amazon. As a relatively new designer, I’m acknowledging that my vision might be overly ambitious. In an ideal environment, I would be working at Amazon where I would have access to the data and metrics to help influence the ideas and concepts generated. Until then, this case study is not meant to be comprehensive or exhaustive in any way. In any case, I am certainly not suggesting that I know more about how Amazon should look than the amazing team of designers working there.
The Big Picture
What I wanted to do in this redesign:
- Create something that satisfies the business’ and user’s objectives
- To design a user interface which reduces the cognitive load on the user in any possible way
- To design keeping User Empathy at the forefront (Human Centered Design)
What I wanted to achieve through this redesign for myself:
- Learn how to conduct and analyze user research, create flow charts, wireframes, polish my UI and prototyping skills
- Complete my first design project from start to finish while assuming multiple roles — User Researcher, Data Analyst & UI/UX Designer
Research 📰, Research 📱 and some more Research 💻
I started with understanding Amazon as a company
Over its nearly 22 years, Amazon has moved into one sector after another and owned it. They started in 2013, but don’t let this fool you into thinking that they aren't outshining the market. They are already well on their way to becoming the undisputed leader in the e-commerce market in India.
Amazon is a leader in UX. What might come across as a little old UI to a passerby or some novice designer (like yours truly) is actually amazing when it comes to usability.
Sometimes when it comes to bad UI, the saying -“If it isn’t broken— don’t fix it” lies true.
After understanding Amazon as a company, I quickly did some market research.
The fact that the shopping experience of Amazon is in fact much superior than any of the competitors became immediately clear. Because of its simple UI elements, Amazon’s load time was shorter than any of its competitors. Going by the customer reviews, the customer satisfaction with Amazon was much better than any other competitor (responsive customer service).
Amazon thoroughly taps into the psychology of its users, especially by using Herd Mentality and the Bandwagon Effect with their secret sauce : Their Recommendation Algorithm
Understanding the Users
Contextual Inquiry
After going through lots of user reviews and talking to people who shop online, one thing became clear :
Designing for e-commerce is an unforgiving task. Consumers — especially those in India — are inherently price-conscious. From televisions to books, the cheapest listing wins.
While talking to them I mainly focused on-
- Understanding the motivation and use cases of users who shop online
- How memorable their experience was
- Problems they faced/continue to face as they use the app
- How they shopped in real life (stores/malls) since I was curious to see if I could draw any parallels to it.
However, I was unaware of the fact that I was being affected by confirmation bias. Since I was only selectively listening to the issues faced by me as well while using the app.
I reminded myself that goal of user research or usability testing is to learn more about users instead being right about users. So I paid attention to data points that break my hypothesis and carefully evaluated them.
To eliminate the factor of my own personal preferences and behaviors when it comes to users. I decided to take into account a bigger user group and conduct a guerrilla usability test of the app.
Usability Testing
To cover my bases throughly (as thoroughly as you can through guerrilla usability tests) , I decided to interview 5 users each from men and women followed by a usability test.
Based on these I found that some users had repeating patterns or motivations to use the app. So I decided to bucket these behavioral models and create explicit archetypes. The two most prominent ones I found were :
- Intent Centric — These users already have a general idea of what they want to buy before they open the app and primarily use the search and filters options. ex- A user wants to buy tennis shoes.
- Interest/Event Centric — These users have a vague idea of what they want to get and take some time going through various options in different categories before they decide on what they want. ex- Some friend’s birthday or an anniversary gift
Making sense of the data 👨🔬
Based on the insights I gained from the usability tests and the contextual inquiries, I decided to do a quick affinity mapping to figure out the major issues I would have to solve.
Problem Areas ⛳️
I identified 3 key problem areas after doing the usability tests and analyzing the notes from the contextual inquiries-
Problem #1 : Interface
Now I know that I mentioned that having a simple UI is something that has hugely benefited Amazon (so all of you saying ‘AHA! stupid noob contradicting himself!’ can rest assured. I’m not adding any complex UI 😛)
A lot of the input I got from the participants, mentioned that the screen was way too crammed with details. Some of them even mentioned that is why they prefer desktop over mobile. But this could very well be a platform bias.
In any case, there was a problem of information overload in the current app.
So, to tackle this issue, I decided to sort through the existing interface and compile an interface inventory. Because I wanted to identify all the possible elements and cases. Other than tackling the recurring problem of too much information at once, I also noticed that there were some consistency gaps here and there.
The top left of a phone screen is generally reserved for meta information (name of device, wifi indicator, back navigational button, apps running in the background, battery life, search, etc.). It’s not a core section of the screen like the center (where your eyes naturally fall) or at the bottom (which are the easiest spots to reach with your fingers). Naturally, the implied message is that things at the top of the screen are to be glanced at, not clicked on.
Top-level actions should be the easiest to discover, but I still wanted people not to lose their access to “everything else” in the app.
I chose to call the primary actions in the tab bar and put the ones which weren’t immediately important or vital to the experience in the hamburger menu.
Problem #2 : Comparing Products
Through the insights I received in my interviews about shopping habits in physical locations… One thing that stood out, was that people love to compare. I mean who doesn’t like getting the most value out of their money? Especially in a price-sensitive market like India.
This could also arguably (only a hypothesis, no data yet) be why so many users found the website to be more convenient. Every time you click on a new product, the website opens it in a new tab. Allowing you to browse through a couple of products and then compare the ones which you like in adjacent tabs.
A clear need for this was observed in the task-centric archetype. They knew exactly what kind of product they wanted. However, they had to keep information about the material, prices, volume etc when comparing more than 2 products.
In one usability test, a user actually brought out a notepad and jotted down these things since she wanted to compare the products before buying them.
Because of this, the Compare feature was added. This allowed users to add more than one product to compare. The app can then compare multiple items on the common information provided by the seller/available to Amazon. Eg- Material, Discounts, Reviews etc.
Problem #3 : Filters
In both the task-centric and the event-centric archetype… The users struggled with applying filters in an intuitive way. They would either add it in their search queries (eg. Stainless Steel Bowl under Rs. 2000) or they would completely overlook the filters section (since it would get lost in all the information being thrown at them)
Not only were people not using the filters as intended, but some users were also dissatisfied with the order in which they were present. Which got me thinking, what if we could crowdsource the data on certain query keywords and arrange the filters in relevance to the search query?
In things like clothes, users might give material more preference than, say material in the case of a bag.
Another issue in filters was the load time. Currently, there is a small load time after every click in which all the search results change in real time.
I proposed the ability to select all filters and then apply at once.
Some might argue that the current method only involves one click while my method of selecting multiple filters and then applying them involves two.
But then again in 90% of the cases which used the filters, the users chose more than one filter. And when people are interested in choosing more than one filters… when there is a small load of time after every selection, it gets annoying.
Camera Search
I haven’t made a change in the camera search feature(basically Amazon’s version of Google Lens, where it allows you to search by image) simply because, a large share of the users of Amazon India Shopping app don’t own phones which support this feature.
That being said, since a lot of people in my user base did own such phones, I came upon a couple of problems which I hope I can get back to sometime in the future.
The biggest problem was the fact that the discoverability of the feature was non-existent. None of the users who had access to this feature on their phones had used it before.
Another issue which came up was dealing with how socially acceptable randomly taking a picture of something in public would be. Say you were out chilling with your friends and you see someone wearing some really rad clothes. Would you be able to take out your phone and snap a clear picture of them without making things awkward? What are the use cases where this is really easy to execute? How can the app give the user suggestive prompts to use this feature in such moments?
All interesting questions to explore into with the correct user group. Since there’s a clear want for it (as found from the interviews)
Takeaways 🎉
Being the first project I had ever undertaken as a growing product designer, this project helped me realize the intricacies of all that goes into making product decisions especially with an established product with a huge user base.
There are a gazillion factors one must keep in mind!
The biggest problem I was faced with was the lack of data to base my design decisions on. Since I was in no way privy to the data based on which the current design is made… I had no way of knowing if the designers had already considered what I was suggesting! (I’m sure they have, Amazon has some amazing designers)
I realized this after I finished and have said this to anyone else who had asked me what their first “UX Project” should be — Never redesign an established product unless you have access to their usage data.
Despite the fact that I reached out to designers from Amazon while designing this concept, there was only so much they were allowed to tell me.
At the same time, I’m glad I got to do this. Because even though it took wayyy longer due to the constraints I set for myself in my approach… I ended up with a much better product understanding of how changes in an established product might come to be!
LATER ON
I also felt that all the UI elements, while simple, weren’t consistent across the breadth of the app. So I decided to make an interface inventory of the app.
Many are familiar with the concept of a content inventory. Content audits are usually performed in the early stages of a website redesign process to take stock of all a site’s content. It’s a tedious process involving spreadsheets and caffeine, but the hard work pays off. You end up all your content laid out on the table, which gives you a new perspective on what your content is and how to tackle it.
An interface inventory is similar to a content inventory, only instead of sifting through and categorizing content, you’re taking stock and categorizing the components making up your website, app etc.
