How would you improve Amazon?

How would you improve Amazon?

What is Amazon - Amazon is an e-commerce platform where users can buy and sell products. Amazon also has services like AWS, Grorecy, etc 

Clarifying questions - 

1. When we say improve Amazon are we talking about Amazon e-commerce, AWS, or any other services  - Amazon Marketplace 

2. Are we focusing on mobile apps or web apps? - Mobile 

3. Is there a Goal in mind? Improve retention 

4. Are we focusing on a specific region? USA 

User groups 

Since our focus is on improving retention, I would focus on existing customers.

Furthermore, these groups can be divided in 

• Prime Customer 
• Non-Prime Customer 

Hello everyone, here is my opinion on how I would approach this process

First question: what Amazon service do we want to improve?

• Amazon.com

• Amazon Prime

• Amazon Web Services (AWS)

• Amazon Prime Video

• Amazon Prime Music

• Kindle and e-books

• Amazon Fresh

I assume that the improvement refers to Amazon.com, Amazon's original e-commerce platform both web and mobile.

Ask to clarify the scope of the improvement. What do we want to improve at Amazon.com?

• The profitability of the site.

• The sales conversion rate.

• Increase user visit rate.

• Improve user retention.

I assume that you want to improve“The sales conversion rate” of the published products. I chose this objective because I consider it to be one of the main values ​​that the platform offers, both for buyers and sellers. Increasing the sales conversion rate increases the value of the time spent on the platform.

I can tackle the problem from two major stages of the life cycle of a product.

• Stage 1: Publication of the product by the seller

• Stage 2: purchase of the product by the buyer

With this in mind, I asked the interviewer if, through analytics tools, they were able to detect where user drops occur in the user purchase flow. Possible problems

• Users do not find the products they are looking for: problem in Stage 1.

• Users fill their carts with products, but never check out: Stage 2.

I assume that the problem is identified in “Stage 2: purchase of the product”, but it is worth clarifying that we should validate this data by analyzing the different drop points of the different user segments.

I´m going to segment the users in the following way

• General consumers– Amazon.com users of all ages and demographics who purchase a variety of products.

• Amazon Prime members: users subscribed to Amazon's Prime program with the benefits of fast, free shipping, and access to digital entertainment content.

• Sellers and merchants: those who publish products to sell.

I assume that we are going to work on the segment of “General consumers, but we should validate this choice with behavioral data in the different user journeys.

Below I list the pains and needs of the “General consumers” who do not complete their purchases for any of the following reasons

1. High shipping costs.

2. Security problems of your personal and financial data.

3. Lack of payment methods enabled in your region.

4. Poor user experience: slow loading pages, complicated navigation or technical problems,

5. Lack of information or transparency: total costs, delivery times or return policy can generate distrust in buyers.

6. Price comparison: Buyers can search for the same product on other websites to compare prices and find a better deal before purchasing on Amazon.

The prioritization of these problems could be carried out through a survey carried out on those consumers who abandoned their purchase at the checkout step.

Let's assume that the problem to be solved is "Lack of payment methods enabled in your region”.

I´m going to propose the following solutions that can improve this problem

• Increase the offer of payment methods: highlight payment methods used in the competition, add more payment options, such as digital wallets or local payment methods, cryptocurrencies, or others, to adapt to customer preferences.

• Flexible payment solutions: Offer financing or installment payment options (like pay later option) to help customers purchase more expensive products without having to pay all at once.

• Improve the checkout experience: Simplify the checkout process and reduce the number of steps required to complete a purchase.

• Assistance in choosing the most convenient payment method for the user: proposing payment methods according to financial costs, discounts, promotions, financing possibilities, other benefits.

I analyze the solutions according to the impact (low, medium, high) on the user and the effort necessary to carry it out (low, medium, high)

In this way to improve the problem of “Lack of payment methods enabled in your region"I would advance first by offering"Flexible payment solutions” and for cases where it is possible according to the region, try “Increase the offer of payment methods” .

For the implementation of flexible payment solutions, I would move forward by integrating financing services offered by third parties, such as BNPL (buy now pay later) or others offered by non-banking financial institutions available in each region.

Although at this time Amazon has chosen not to offer the ability to accept money deposits directly into user accounts, it would analyze the possibility of making agreements with a financial institution that can so that it can facilitate this type of service with benefits for Your clients.

To measure these improvements I would do an A/B test comparing the conversion rate in the checkout stage of purchasing products.

It would also measure the amount and number of operations carried out through these new payment options.

• No order on an item after the trending feature no of order on an item before
• CTR on trading item
• GMV

Amazon currently has a lot of businesses. 

1. E-commerce, 
2. Cloud, 
3. Medical, 
4. Roomba, 
5. Alexa, 
6. Home bell, 
7. Prime, etc. 

It controls what we watch, what we eat, what we speak at HOME etc. 

I would like to have Amazon Homes. The building or construction part can be taken care of by someone else. But the smart software inside the house can be done by Amazon I feel. 

Amazon provides A to Z through its website or services. And this is done with the best service possible. As I spoke above, it provides all those services, so providing a smart home service makes logical sense in its roadmap. As a company, the software is its advantage and Amazon can build software for home (like how Google is building it for cars) 

Customers have a sense of trust in Amazon's products and that's shown in customers' reviews and purchases on Amazon's website for Amazon products. Examples for all home-related products: Alexa, Amazon Fresh, Whole foods, Roomba, Prime services, etc. This trust will carry across to other Amazon products.

Collaborate with construction companies to incorporate software into the buildings. Be open to accommodating software of other smart home makers (like Google) if customers prefer to not have Amazon software in it.

The housing market is a trillion-dollar industry. And with deeper internet penetration, more smart cars, smart light controllers, and the rise of smart home devices - customers would be interested to have integrated smart home systems. 

Amazon can be the OS of houses. 

It's worth placing a bet on Amazon Homes - given the customer and market trends, amazon's recent purchases or roadmap, and the TAM of the housing industry.  

Trade-offs: Privacy issues, never been into construction domain

Clarifying the situation

Amazon has a multitude of products, so let’s assume that the improvement is for Amazon marketplace. This improvement can be globally and not restricted to any geographies. The marketplace has both buyers and sellers, for the improvement I will go ahead with the buyer in mind.

Today internet penetration is at its peak and in countries like India over 10% of the population shop online.

Goal

The overall goal of amazon is to be the most customer centric product and the marketplace intends to digitize the shopping experience. The goal of the improvement is to drive further conversions on amazon.

User Persona

Three personas have been considered here thinking of individual shopping patterns

1. Explorer — Millenials and Genz’s majorly who has a vague idea on what he needs, explores the app for products, open to suggestions and may go on a shopping spree

2. Know-it-all — Middle aged people majorly who know the exact item to buy and will only purchase that and leave. Its not very easy to persuade them to purchase new items as they believe they know what’s good for them

3. Non-shopaholic — Someone who’s not into online shopping, considers offline purchasing as the way to go and relies on acquantances for inputs on purchases. 

 Based on the effort required I would like to pick the explorer segment as increased purchases in this group can also have a network effect on the know-it-alls and shopaholics

User journey & Needs

Curious —  The user is first curious about a type of item he/she heard about and wants to know more about it. The pictures may not be very convincing at times. The standard descriptions and random reviews may not be very appealing to him.

Explore —  He finds various products across price ranges and variations and is confused. He wishes he could get better discounts which would push him to make more purchases.

Buy — After consulting with his friends / family / google and comparing with other shopping sites he purchases the item. And would track the item.

Receive — If the user is excited about the item he might want to share it with his friends or post stories on the item. These maybe aspirational products and user feels a sense of ownership. They might recommend these to other people and would like to talk about them in conversations.

Solutions

Amazon Expert Blogs — Today many products have videos and reviews put out by people on youtube and other social media platforms. Amazon can introduce its on blogs with videos and reviews of product written by credible sources. Ask an expert option can connect users to experts and clarify queries they might have and show live video with the product.

Social Vote— Shopping used to be a fun and social activity for millenials and genzs. This could be introduced in the online mode. Amazon users could be added to an item in the cart where they could vote for an item in cart by their friend. More number of votes implies his friend circle approves of the purchase. 

Friends Review — Stats on how many of my friends on Amazon purchased the item could be shown incentivizing the person to make the purchase. Recommend button also can be introduced to add a friend who might also like the item.

Product comparison — Custom comparison with paramaters of choice could be introduced for products and stats on how the product fares in terms of price to same product across online and offline shops could be shown. 

Group buying — Amazon can give discounts to people to purchasing item in bulk. People can add others to their purchase and get the item delivered at lower price.

Share — People would like to talk about their purchases on social media. Share button to post custom stories/ posts can be introduced post purchase of the item. These could have fun themes with interesting captions like “Guess what I just bought ?”

Amazon VR(moonshot) — Users can shop with their friends all in various custom avatars and chat about items, all in virtual reality.

Social vote , friends review and product comparison are features I’d like to start with considering the impact on purchase they can have and the simplicity to implement them. Group buying can create huge network effect and can get other user personas as well onboard.

Metrics

1.  # of products voted for

2. Increase in individual purchase 

3. # of products compared

QUESTION: How would you improve Amazon?

BUC Method = (Business benefits + User benefits) - Cost

BUC Method = (Business benefits + User benefits) - Cost

Clarification Questions:

-          Define which business

o   Amazon Prime

o   Amazon Marketplace

o   Amazon Web Services

-          Define which department

o   Services: Prime Video, Audio, Alexa, Fire, Kindle, Audible, Grocery, Fresh and Restaurants, Home Services, Credit & Payment Services

o   Tangible Products : Like Electronics & Office, Books, Health & Beauty, Clothing-Shoes & Jewelry, Automotive & Industrial, Toys-Kids & Baby, Sports & Outdoors, Handmade

-       Define App Vs Web

Distilled Choice: Amazon Prime App users shopping tangible products

Users Segments: 

For Amazon Prime App users shipping tangible products

-          Amazon Prime

o   Users with a specific problem or need and looking for a product or solution. Shoppers tend to be one of the 5 stages in their shopping process

§  Stage 1: They know what their problem or need is

§  Stage 2: They are researching using specific keywords, using similar product recommendations from PDP pages, reading reviews, searching for brand reliability, looking for social media mentions etc

§  Stage 3: They have identified multiple products or solutions and are evaluating them side by side

§  Stage 4: They have narrowed down some choices and are in the purchase phase

§  Stage 5: They have made a purchase and are evaluating the product by using it within the free returns window

o   Users looking for deals and are open to buying any product that has a deal

My Choice: Amazon Prime App Users in Stage 4 of the Shopping process for the above mentioned departments

Pain Points 

 The assumption I am making here is that the user has thoroughly researched the product, compared it to competition and knows that he is getting the best deal on Amazon Prime. The user has added the product to cart. While in the cart he notices :

Delivery Time: The earliest that the product can arrive is in 15 days which may vary based on current COVID regulations.

Seller Reputation: Recent bad reviews for the seller’s policies (not of the product) is something that people notice when they learn about lengthy delivery times and start digging around the seller’s profile page.

Motivation to check out goes down. The items continue to linger in the cart.

Solution

In addition to sending abandoned cart emails, my proposal for when abandoned cart value greater than $100 is to build an in-app and push notification feature which would offer the user an incentive of $x in shopping credit towards the next purchase, if they commit to purchasing  within the next 24 hours.

Product Vision

“ Earn credits by committing to purchase”

The shopper is an intelligent entity who is not only looking for the best bang for the buck but also wants to close the loop of anticipation by having the product in hand within a reasonable period of time. When lengthy delivery times become the reason for users walking away from making the purchase, an added motivation to commit to the purchase will increase the chances of users purchasing the items in the cart.

Features

•  $100 threshold of cart value with items from one of the said departments to trigger an in-app and push notification for the user to claim the incentive and commit to purchase in the next 24 hours.

• Committing to purchase will result in the cart items getting purchased.

• Notification to trigger after the first 24 hours of cart abandonment.

• If the user does not commit to purchase in 24 hours after the notification triggers, then an exit survey with 1 question and pre-populated answers on why they do not want to complete the transaction gets served.

• Upon completion of the 1 question survey, an option to clear cart which would also stop the abandoned cart emails.

• The incentive value available for the user on the next purchase to be valid for 365 days.

Trade-Offs

• Requires thorough data analysis of how many users with a $100 threshold with tangible products in cart abandon their cart.

• Would Amazon lose money by offering the user the chance to check out the cart now vs letting the items linger around in the cart?

  • Needs deeper analysis on what threshold value for cart and the incentive value to commit must be for the idea to be profitable.

• There is a chance that users might take advantage of the incentive, complete the purchase and then choose to return all or partial items that they purchased. Would this become a behaviour pattern? The user would be just collecting credits and returning the items. Should an account flag be created that looks for “Commit to Purchase” credit collected in the recent X days and place limitations on when the user will be next eligible to claim such a credit again?

Clarifying questions, does this question apply to all Amazon product (AWS, ecommerce), all devices, etc? For my answer, I will assume the interviewer is referring to the ecommerce experience from a user perspective.

I would ask if there is a particular metric the interviewer would like me to drive. For my answer, I will assume the interviewer would like me to drive customer satisfaction with the Amazon shopping experience.

The structure in my response as follows:

• User Journey 
• Existing pain points 
• Prioritize pain points to solve
• Solutions for critical pain points
• Measuring success

The Amazon shopping journey broadly consists of:

1. users needing something to purchase to solve an unmet need or preference

2. navigating to amazon either through marketing channels or direct load

3. browsing or searching the site

4. purchasing a product

5. waiting for the product to be delivered

6. returning the product if it is not satisfactory.

Given how broad the question is I will exclusively focus on alleviating pain points surrounding returns.

1.  Inconvenience of bringing item back to post office to ship

2. Inconvenience of visiting one of Amazon's returns locations (Whole Foods, Kohl's etc)

3. Missing return window and not able to get refund

4. Losing packaging materials for an opened product that ultimately proved to be defective or unwanted

I will primarily focus on 1 & 2, the inconvenience of bringing items to a physical location (whatever that location may be) to return an item. It would be much more convenient if users could schedule return windows for drivers to come pick up the item at a user's home or another convenient location at a specified time. Pros of this approach is that it is more convenient for the customers, Cons is that it may overburden Amazon logististics. For this reason, it would be most effective if Amazon could coordinate return pick ups on their delivery routes. This is a strategy that has also been adopted by Walmart, one of Amazon's competitors so it has become table stakes to have a convenient return policy.

Furthermore, Amazon could set up additional drop boxes and allow customers to drop off boxes at these locations.

I would measure success primarily by looking at customer satisfaction of users who have returned using the pick up method and look for ways to improve and iterate on this process. Other metrics I would look at is the usage of this feature and how it overall increases repeat purchases for users who utlize this feature.