Design an experience for getting book recommendations on Facebook. How would you measure success?
0 votes
in Product Design by (1.3k points) | 189 views

2 Answers

+1 vote

Interviewee: To clarify the scope, the requirement is to design a book recommendation feature for a FB user, correct?

Interviewer: Yes

Interviewee: Is the goal of the recommendation feature to provide a discovery channel that connects users with books that align with their preferences and interests?

Interviewer: Yes

Interviewee: Does FB want to introduce this experience to improve their engagement metrics and potentially partner with online book distributers for a revenue opportunity? Is that the objective?

Interviewer: Yes, FB is trying to grow strategic relationships with publishing houses and exploring avenues to discover deeper user interests in a creative way.

Interviewee: First I would think about what type of user would be interested in such a feature and whether there is an opportunity to discover or a need for the user. 

Let’s explore the possible user profiles a little bit to understand the user needs

  • Reading enthusiasts (young adults and adults)

  • Parents of children who like to read

  • People interested in pursuing a writing career

  • Libraries looking for new material to add to their collection


Given that the FB user base has a majority share of males and females in the age group of 24- 34, followed by the 18 to 24 age group. The feature can be targeted to appeal to this group versus the young adult age group.


Target user profile summary


Bookworm Percy

Age: 28 

Location: New Jersey, USA

Education Software engineer


About her: Prefers to use a grocery shopping app to do her weekly groceries, she buys clothes online and uses Kindle to read.



  • Discovering new books/stories to read

  • Cataloging her book collection


Pain points: 

  • Keeping track of books she is reading

  • Finding books beyond the best-seller lists

  • Discovering unique stories from lesser-known publishing sources


Reading habits:

  • Loves to re-read books

  • Enjoys e-books and the flexibility it provides

  • Fast-paced reader

  • Member of multiple book share communities


Problem statement: As a user, I want to find new books without having to search for them, so that my to-be-read list is always full.




Feature goal: Create a book discovery experience for a fb user






Success metrics:

  • Strong engagement indicators will help determine whether 

    • Tracking mouse hovering, which serves as a proxy for attention on the feature

  • Performance 

    • Number of impressions or number of times the recommendation feature was in-view of the user for more than 2 seconds ATF

    • Clicks - User clicked through to the destination URL

    • Click-through rate


Positioning the feature


The feature can be designed in the left or right rail of the feed or directly into the feed. Depending on user testing results, we can determine the position of the recommendation feature. 


Although the feature is part of FB layout, it still an ad of sorts so the principles of native ad formats apply to the recommendation as well. Sponsored ads typically show up within the feed in the same format as a FB user post accompanied with an image, this technique helps keep the user engaged and continue scrolling without breaking expected format and visual flow. However, options to position the recommendation as a menu item on the left side-rail under the ‘Explore’ section with a simple alert might help keep the in-feed ads from being cannibalised, especially because it is a revenue stream for FB.

by (15 points)
0 votes

Clarifying the scope:

When you say books - it can be a book in any format? ebook, audio books or printed 

What's the goal here - is it new user acquisition or engagement or driving monetization, I'm going to assume it's increasing engagement 

How do we measure success? - I will assume time spent as the north star metric

Which platform? - App, mobile/desktop web? Assume both with a focus on app 

Other health metrics to keep an eye on :

  • What % of users discover the feature
  • What % of users try to use the feature
  • What % of users actually use the feature
  • what % give recommendations/how many recommendation/ user per week/month 

Who are the customers?

Since the goal is here to drive engagement we'll focus on facebook user sets - let's do a deep dive of the behaviours and identify the potential cohorts we'd like to target with this exercise

  • Cohorts
  • Readers - those giving recommendation
    • Why would they recommend: Gives them status, they appreciate positive feedback from people 
    • When do they recommend: People might recommend AFTER reading a book or during
    • Audience and Privacy: People want to ONLY share recommendtations with the world and not the rest of their profile 
    • Format: People can read books in a variety of formats - it could be ebooks, audio books or printed books,they might recommend a particular format of a particular book 
    • Target audience:  They might recommend a book for a particular audience of an author or another book or a genre
  • Recommendation seekers - those looking to start reading 
    • Reading habits: Users typically read by genre or author or both
    • Active or passive: People might recommend books or seek recommendations
    • Where do they seek: Could be on their own page or in a group 
    • Seeking for others: Some might seek suggestions for others - for example parents seeking to get a book for a child or people looking to gift books and seeking recommendation
  • What are there needs
    • On readers side:
      • Auto suggest: We want to ensure that both recommendation seekers are referring to same frame of reference so auto suggest to suggest the book titles will help ensure there isn't proliferation of the same book's name into other names 
      • Privacy setting: To encourage recommendation we need to reduce friction, creating own page just for recommendation will hamper discovery for those seeking recommendaiton and increase friction for the person giving recommendation hence there should be a setting
      • Status tracking: Imagine a recommendation giver has a number of followers then the followers may want to know what he's reading and look forward to the same 
      • Format: The recommendation might be specific to a format - say audiobook vs text book
      • Quotes:  The users may want to refer to excerpts from the book for the recommendation, we should allow the user to do so
      • Target audience: The book recommendation might be applicable to a specific audience, this will help the relevant recommendation seekers find the right recommendation/recommmendation giver
      • Feedback from readers:  The recommendation seeker would love to hear feedback from the user such as they read the book after his recommnedation or enjoyed the recommendation or a star rating
    • Recommendation seeker
      • Seek/search recommendation:  The user should be able to search for recommendtation/request if needed
      • Seeking for others: The user shoud be able to seek recommednation for other people 
      • Seeking recommendation: The users should be able to seek recommendation in group or via profile 
      • Auto recommendation: the user may find it useful to have book recommendation suggested to him/her basis their author/genre/book preference 
      • Offers : The user might find it useful to find offers on books they are interested in reading 
  • What do we want to focus on 
    • Recommendation giver
      • Auto suggest: table stakes - must have for recommendations 
      • Privacy setting: table stakes - users are beginning to take their privacy very seriously, this will help drive adoption
      • Status tracking:  Must have, users often start reading a book but don't finish, this will give enough feedback to users to ensure more people finish reading the book and hence you get more recommendations
      • Format: post MVP not a must have 
      • Quotes:  post MVP not a must have
      • Target audience: Must have, it'll help ensure that recommendation seekers can search for recommendations and also help drive recommendations 
      • Feedback from readers:  Must have, it creates a powerful feedback loop that will help drive retention from recommendation givers
    • Recommendaiton seeker
      • Seek/search recommendation:  Must have, relying only on people asking drastically narrows the funnel hence adoption rate 
      • Seeking for others: post MVP, not sure if this is even a large enough use case right now 
      • Seeking recommendation: Must have, table stakes 
      • Auto recommendation: high complexity but must have since it'll help drive discovery without spamming the user (assuming we'll be able to use data from pages/user's profile to ensure this works)
      • Offers : high complexity , low impact  Post MVP, only makes sense if this features gather traction and besides monetisation is not the goal here 
Overall : Impact on timespent of users exposed to the feature those who didn't get exposed
by (62 points)
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
To avoid this verification in future, please log in or register.
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
To avoid this verification in future, please log in or register.

Post your resume

Post homework assignment

About this site

Product Management Exercises is the best place to get help preparing for product manager interview questions. Any member of Product Management Exercises can post product manager interview questions, submit answers to the questions, and give feedback to other members' answers.

Resume Review

Post your product manager resume for feedback and give feedback to other members’ resumes.

Follow PM Exercises