If you are the Product Manager at Vistaprint in charge of Recommendations for the My Account section, how would you approach it?

Question

If you are the Product Manager at Vistaprint in charge of Recommendations for the My Account section, how would you approach it?

Asked at Flipkart in the past year

502 views

Product Design category

Asked at

502 views

How to answer Product Design questions

Interview Guide

Top Product Design interview questions

How would you design a bicycle renting app for tourists?62 answers | 82.5k views
Build a product to buy and sell antiques.54 answers | 66.8k views
How would you design a web search engine for children below 14 years old?36 answers | 42.9k views
See Product Design PM Interview Questions

Invite members
Invite by email

Add another

Kumar Vansh Moondra · Answer 1 · 2023-11-05T07:48:32+0000

Problem Statement:
Need to launch a Recommendations engine for the My Account section at Vistaprint

Product Description:

Vistaprint is an E-commerce website that specialises in personalised merchandise like pens, business cards, bags, and more. Within this platform, My Account is an essential feature on Vistaprint that allows users to manage their orders, update their personal details and track the status of their shipments.

Clarifications/Assumptions:

1: Even though our business might seem like B2B, it actually operates more like a B2C model because most of our customers are individuals who need things for themselves

2: We've already got a recommendation engine (looks like a content driven)in place at multiple spaces after checking UX, so we'll use those same engines to make this work at My Account as well.

3: For now, we'll treat the use cases and needs the same for all countries (since we deal in multiple locales), unless the data shows us something different, which isn't the case at this early stage.

4: When I said "recent" in the "User needs/Solutions/Tradeoffs/Feature" column, I meant orders placed in the last 7 days. And when I said "past," I meant orders placed more than 7 days ago.

5: I have considered 7 days as a period as having an understanding that delivery happens within 7 days after the order gets placed. This is based on the general delivery estimate of 2-3 days mentioned on the website. {{Assumption}

6: Assuming that it’s the initial stage of this feature, we'll start by testing it with the most important use cases in the 1st iteration. Plus, we believe that most of the people who use My Account have ordered before because My account serves purpose for returning customers

Objective:

We're categorizing users based on AAAERRR (Awareness, Acquisition, Activation, etc.).

In this case, I will take relevant objectives that come with product adoption an important key as we're talking about launching the feature, and the product's success is primarily determined by how many people become aware of it, decide to start using it, how they get activated or engaged with it, and make it all the way through the user journey.

Before committing to any specific objective within Product adoption, let’s go through the use case and will select later.

User Groups:

1. Small business owners (returning and new user)

2. Event planners (returning and new user)

3. Entities such as Digital Marketing/Brand Marketing agencies(Returning and new user)
4. Procurement Managers from large scale companies (Returning user)

To start, we'll focus on Small Business owners who order for themselves (not other businesses). They bring in the most money and there are quite a few of them, based on what we've learned from the initial data and our understanding of the UX

Imp Use needs/Use case:

Here, I'll begin by focusing on the user's need to use My Account. Since our task is to introduce recommendations in My Account, it's crucial to first understand what users want from My Account.

Returning user:

Use Case	Priority	Why at P0/P1?
As a returning user, I log in to My Account to manage my profile information, update my address details, etc (recent)	P1	1: Lack of intent - It's not a good idea (from a customer point of view) to recommend new things to the user if they've just bought something in the last 7 days. If they needed something, they would have already bought it when they made their recent order 2: Avoiding cannibalisation: A customer might log in the day after ordering and change their mind, deciding to cancel the order for any strange reason such as better product/discount/ratings (order is not moved to printing yet assuming), that’s why keeping 7 days in buffer to avoid the same
As a returning user, I log in to My Account to see a summary of all my orders just to get a quick overview (recent)
As a returning user, I log in to My Account to check the specific information about the order I just placed (recent)
As a returning user, I log in to My Account to check the specific information about the order I placed (past)	P0	Pros: 1: Intent to buy must be better 2: Less work for the user & seamless experience (less fussy experience)
As a returning user, I log in to My Account to manage my profile information, update my address details, etc (past)
As a returning user, I log in to account to receive personalised product suggestions based on my profiling, past history details, etc so I don't have to waste time searching for items from Home, etc (past)

New user:

Use Case	Priority	Why at P0/P1?
As a new user, I sign up for My Account to get special offers for my first-time shopping.	P1	1: Sign up - Use case not relevant to our problem statement 2: As we've mentioned before, most of the users who visit My Account have been here before, so we won't focus on this use case for now.
As a new user, I create an account to receive personalised product suggestions based on my browsing interests, so I don't have to waste time searching for items at PLP	P1

In summary, I'm giving the highest priority to customers who have ordered from us over a week ago and are coming back (P0)

Solutions:

Use case	Solutions	Priority
As a returning user, I log in to My Account to check the specific information about the order I place/or I want to do reorder (past)	Selling more products by using cross-selling and upselling techniques in 3 different ways: -Content driven recommendations -Contextual driven recommendations -Hybrid recommendations	Content driven: P0 Context & hybrid: P1 Reason: Since we're working on the MVP, especially in the My Account section, I suggest we start with content-based recommendations initially to roll with minimal efforts. 2: Later, we can expand to contextual recommendations based on impact as there are many factors to consider like seasonal, timing, the device used, browsing history, etc., that will influence contextual & hybrid recommendations. (in short, lot of use cases)
As a returning user, I log in to My Account to manage my profile information, update my address details, etc (past)
As a returning user, I log in to account to receive personalised product suggestions based on my profiling, past history details, etc so I don't have to waste time searching for items from Home, etc (past)

Feature:

Solution

Logics

Feature Representation

Priority

Content driven
algorithm

1: If I ordered something from 7 to 30 days ago, suggest related products to go with what I bought (Cross selling).

2: If I ordered something over 30 days ago, recommend upselling related to my recent purchase

Quick note:

Products bought in the last 30 days usually last about a month.

That's why it's a good idea to suggest related products (cross selling) to customers within 7-30 days because they're still using what they bought recently.

1: Making recommendations visible with messages like "You might like these" in carousel/ progressive disclosure manner on My Account Landing & Relevant pages in My account (both cross selling & upselling)

2: Interactive Mode:
For eg of cross selling If we're displaying recommendations for business cards, we can show a video or image of someone buying the cards, then a meeting where they use a notebook to take notes, and finally using a pen to write. This sequence shows the steps or experience users can have with the suggested products, hence cross selling in this example

3: Contextual Imagery: When implementing visual storytelling, we can use images and visuals that are not just product shots but also convey a context or lifestyle. For example, if we are recommending pens, we would include images of writers who like customised pens to write, giving gifts to their employees/clients. These pictures set a mood and make users feel a certain way, making their experience better and helping us revenue

1: Carousel/Progressive Disclosure Manner: P0

2: Interactive Mode: P1

3: Contextual Imagery: P2

To decide on the order to work on these, I'll use a prioritisation method like RICE or T-shirt sizing. As I see it, implementing a carousel or progressive display should be straightforward because our website already uses similar setups in other areas, so it won't require much effort.

So, to sum it up, I'll start with prioritising the Carousel/Progressive disclosure approach.

Tradeoff:

Thinking about the fact that some users can fit into both situations.For these users, we'll see how many of their recent orders match the same category within the last 6 months.

If more than half (50%) of their orders are in the same category, we'll focus on suggesting products related to their previous purchases (upsell). If less than 50% their orders match the same category, we'll recommend different products that complement their previous purchases (cross sell)

Success Metrics:

I'll be splitting the measurements into three: primary, partially primary & secondary:

Metric	Metric category
Number of impressions	Secondary
Number of clicks	Secondary
Unique impressions per user	Partial Primary
Unique clicks per user	Partial Primary
CTR (avg)	Secondary
CTR (unique)	Partial Primary
Number of users added to the cart	Partial Primary
Number of users have proceed from cart to checkout	Secondary
Number of conversions	Primary