How would you increase the basket size of different products for users?

Question

How would you increase the basket size of different products for users?

Users currently come and purchase one type/specific product and check-out.

How can we increase the differentiated products purchased by users?

The solution could perhaps be a combination of various levers of upsell/crossell,etc through the various stages of pre-purchase, purchase, post purchase journey.

What will be the rollout strategy?

How will you measure success?

Asked at Amazon over a year ago

1.4k views

Asked at

1.4k views

Top Product Growth interview questions

What would be your strategy to bring partners onto the Facebook platform for Black Friday?4 answers | 2.2k views
How would you increase the revenue of Google Maps?3 answers | 5k views
How would you grow the number of Instagram users?2 answers | 3.1k views
See Product Growth PM Interview Questions

Invite members
Invite by email

Add another

Tarun Kumar · Answer 1 · 2023-10-23T10:15:41+0000

Clarification needed:

1. What product are we talking about? (Answer: Assume it's Amazon)
2. Are we looking at a particular platform mobile or web? (Answer: Ideally we would want to move these metrics for both)
3. What is the shopping behavior, where does more checkout happen? Is it on mobile or web? (Answer: For the sake of this discussion let's assume this to be Mobile)
4. Are we looking at a geography? (Answer: Your call, for this discussion let's think of India)
5. What is the timeline for the same? (Answer: We need to improve these numbers quickly, let's target this for the current quarter or 3 months)
6. What is the objective? (Answer: You decide)
My take: Since we are talking about increasing the basket or cart size per user, I'm assuming we want to have monetization and conversion as our goals)

7. Are we looking at a user cohort or across all users (Answer: You decide)

Going ahead

Define metrics: I'd like to see the following metrics improve

Avg. number of products in cart at checkout

User cohort:

1. Power buyers: These are my core users, they are pretty much online. They like to shop while looking for unique experiences, they care more about reviews from their friends and family. They typically shop with Amazon >= 1 in 2 weeks. These users typically tend to have a high-end smartphone and live in Tier 1 or Tier 2 cities.
2. Occasional buyers: These are users who shop >= 1 in 1 month. These are users who generally live in Tier 2 or Tier 3 cities.
3. Dormant buyers: These are users who generally shop once in 3. They are the once who rely on offline stores for their needs, they are not so tech-savvy or enabled to purchase on E-commerce, and they rely on someone else to shop for them.

From the above user group, I'd like to pick power buyers as they have more propensity to buy things from the platform. Here I can test my solutions quickly on them.

Check with the interviewer, moving ahead I'd like to think of some pain points for these users.

User needs:

1. I want to discover and search for items that are trending.

2. I feel shopping online and discovery is not as delightful an experience as in an offline store.

3. I want to get suggestions on the items that I buy early on. Social angle in shopping.

4. I want to understand a bit about the story of a brand before making a purchase

Going ahead I'd like to prioritize a few needs,

Users Gen-Z and Millenial are typically living virtually, so the appeal to them is different. They want to be more authentic and look for unique experiences. I'd want to double down on this user psyche, as well as the need to discover items that are trending and making shopping feel more social kind of experience. Why? I feel the need to get some social validation and search/discover for items are the biggest pain points.

Check with the interviewer about this, they give you a nod. Going ahead, I'd like to brainstorm some solutions

1. Search and Discover items that are trending

The trending Items page is designed especially for millennials and Gen-Z based on the shopping data accumulated so far and also doing some trend analysis around their shopping nature
Cross-sell items based on user behavior as soon as the user adds an item to the cart and goes to the cart
Add tags to your items added in the cart that re-confirms user's belief in the product

2. Social experience in shopping

Share AR pictures of you wearing a shirt, glasses, sneakers, and so on with your social circle.
Create a poll and share it with your friends list on WhatsApp or Facebook, and ask them to rate it
A small and quick gamified quiz around asking friends what sort of experience/product would fit me today basically a way to get recommendations on products.
Create avatar and as you shop the avatar gets tags from community, you could then be influencer and guide other Gen-Z and Millenials and seek your shopping advices.

Now to go ahead and think of prioritize solutions.

ItemsImpactEffort

A trending page for Gen-Z and millennials	High	High
Cross-sell items	High	Medium
Add tags to items	Medium	Medium
Bring AR to the shopping experience	High	High
Create a poll and share it with friends	Medium	Medium
Create and share fun engaging quiz	Low	Medium
Create an avatar and level up in the shopping game	High	Medium

Based on the above table I'd pick two items to deliver in this quarter:

1. Cross-sell items once the user has added a product or spends enough time looking at a product

2. Create an avatar and ladder up in Gen-Z or Millenia board to suggest items to other users here

Metrics:

1. Average no. of items in cart to increase

1.2. No. of items added from cross-sell

1.3 No. of items added that are recommended by leaders

1.4 Guardrail: Don't want too many returns on the items

1.5 Guardrail: Don't want too many abandoned cart

For starters, I'd like to understand user behavior and create model specifically for a particular user base for item recommendation.

For creating an avatar, will look at the data and give some users early access to create avatars while shopping. To solve for adoption, I'd first create some celebrity avatars that shop on Amazon and they would nudge users to join the avatar list and make shopping a fun and engaging experience. The avatar chosen by users can also recommend or upsell items as they are shopping. This could be like Duo-lingo's animated character that speaks to you while you shop. You can turn them off but if they become too intrusive.
I'd launch this based on user profiles and get some traction for users who are in tier 1 and tier 2 cities. I'll have some marketing efforts to advertise these features on social media, budget can be decided and once the MVP is launched the budget will be increased slightly.

Summarize this entire scenario around GTM and appraoch

Check with the interviewer.

Looking for the feedback here.

amrita_n · Answer 2 · 2022-04-10T01:58:55+0000

Clarfications:

1. Since this question was asked at Amazon, I believe its specifically targeting the users of Amazon's ecommerce platform. Let's target this specifically for US geography

2. Also, I interpret the problem statement as two-fold: a) increasing the size of the shopping cart b) increase the variety of products in the shopping cart

3. Amazon already has a huge user base and user awareness/adoption does not seem to be a goal at this point. We can focus on the problem statement for the existing customers

Possible user groups:

1. Power users: typically amazon prime members, customers who rely heavily on online shopping and place typically >5 orders per month on amazon

2. Casual users: These are customers who leverage online plus in-store shopping experience and place around 1-2 orders per month on amazon

3. Occasional users: do not rely on online experience except for situations like pandemic etc. or for certain items that are not available in-store. I would guess these users typically use amazon once in a couple of months

I believe the general sentiment around user groups #2 & #3 is that they prefer to look and feel the items in person before buying them. I would categorize them as 'risk-averse', so the opportunity lies within group #1. We can use them as a target for increasing the basket size. (Also amazon has already optimized in refund policies and ease of returns, hence not much scope here to address groups 2, 3)

Let's look at a basic user journey and try to leverage the opportunity areas:

1. User has specific item(s) in mind, and logs on to amazon.com

2. Searches for specific item(s). If item not found, exits- opportunity area

Solution:

>> The user will receive an email on similar items/substitute items

>>The email will bundle up suggestions on possible related items (e.g. nike shorts will bundle up suggestions for running shoes)

>>If the item becomes next available, users will receive an email notification

3. If the item is found, adds to the cart but forgets to check out and exits- opportunity area

Solution:

>>The user will receive an email reminder - 'did you forget something in the cart?'

>>The email will bundle up suggestions on possible related items

4. If the item is found, adds to the cart and goes to checkout. Enters payment details, shipping address and completes the order

>>The user will be given suggestions to add possible related items to the cart before checking out

Opportunity areas outside of user journey:

1. Buy more, save more: The users can be given deals on certain items frequently bought by them, i.e. buy 1 get 1 50%. This will encourage the increase in the cart size.

2. Periodic emails on discovery items based on demographics: The users will receive emails (say monthly) on 'undiscovered' items that are purchased by the customers with similar demographics (location, gender, possible age-range etc.)

3. Periodic emails on discovery of 'top selling' items: A weekly email on top-selling items across all categories on amazon US for the past week.

4. Referral: Ability to refer a product to a friend via email notification, by click of a button next to product description

Prioritization:

Solution	Impact	Effort
Solutions to specific items not found (user journey #2)	M	L
Solutions to items added to cart w/o checkout (user journey #3)	M	L
prompts to add related items during checkout (user journey #4)	M	L
buy more, save more	H (there is a natural user tendency to spend more to avail more discounts)	L
Periodic emails on discovery items based on demographics	H	M (this will require some AI to identify the correct user personas based on the browsing history)
Periodic emails on discovery of 'top selling' items	H	L
Referral	H	M (this will require some development efforts on the UI)

H=high, M=medium, L=low

Roll-out strategy:

All items with 'L' effort will be prirotized for phase-1 in the US region.

Success metrics:

1. avg number of distinct items per user per month across all orders

2. avg size of each order per user per month

3. avg distinct category purchases (i.e. apparel, books, etc,)

Counter metrics:

1. Impact to the purchase frequency of power users (i.e. # of orders per month)

2. No. of returns for the purchases of 'discovery items' or 'related items'

3. No. of bad reviews (<3 stars) for the purchases of 'discovery items' or 'related items'