15% off membership for Easter! Learn more. Close

How do you define the goals and metrics for Facebook Newsfeed?

Asked at Meta (Facebook)
11.7k views
Answers (3)
crownAccess expert answers by becoming a member

You'll get access to over 3,000 product manager interview questions and answers

badge Gold PM

Defining goals and metrics for the Facebook Newsfeed. 

First, let's understand what the goals of the newsfeed. Newfeed is Facebook's landing page. When users log-in they see the Newsfeed as their first stop within Facebook. I know from my experience, you'll see your networks posts, event activity, news from pages you follow, ads and other relevant content. The goal of the Newsfeed is to help you keep updated on the recent (or not recent) activity that is happening in your network. Your network can consist of friends, places you follow, pages you follow (music artists, communities). 

When I think about metrics, I'll hone in on Facebook's motto which is connecting people from around the world to each other. Key metrics I feel like we should track when it comes to the Newsfeed are: 

1) Daily Active Users & Weekly Active users & Monthly active users 

2) Time spent on the newfeed from login to navigating away 

3) On average within a session on Facebook, what % of time do users return to the newsfeed 

4) Retention metrics such as: 7 day retention of users, 28 day retention of users 

5) Engagement in the newsfeed this would include: reactions, comments and shares in newfeed content. Reactions would include: likes, hearts, sad face, angry face etc. 

6) Click through rate for ads; this would help us understand whether ads are relevant 

7) Number of connections made through the newsfeed page; friend suggestions are displayed on the landing page. I would want to understand if those are helpful to users 

Access expert answers by becoming a member
4 likes   |  
Get unlimited access for $12/month
Get access to 2,346 pm interview questions and answers to give yourself a strong edge against other candidates that are interviewing for the same position
Get access to over 238 hours of video material containing an interview prep course, recorded mock interviews by expert PMs, group practice sessions, and QAs with expert PMs
Boost your confidence in PM interviews by attending peer to peer mock interview practices, group practices, and QA sessions with expert PMs
badge PM

Description of the Product:

The Newsfeed is a key element for Facebook, as it is the page a user is taken to as soon as they log in. It has a scroll-through style that shows friend activity, relevant and recommended content, as well as advertisements. This is where the most action takes place, and was implemented in order that the user would not have to navigate to a specific friend's timeline to see what they have been up to. Rather, they can get a variety of information aggregated in one place.

Goals:

As with most FB products/features the goal is to promote user engagement.

Customer Journey:

The journey is simple as mentioned above: once the use logs on, he is immediately taken to the Newsfeed. Hence, we can neglect awareness, acquisition, and activation (since we can readily assume the user has already registered an account) to focus on engagement and retention.

Metrics:

  • For engagement, I think the main thing we can measure the average number of interactions a user has per visit to the Newsfeed. That is, how many times they like, comment, share, or post content per visit. We can perhaps break this down further to see which type of interaction is most prevalent. This is useful to gauge the how the presented content captivates the user - it could be a sign of how well the algorithm that determines relevance/recommendations is performing.
  • We can also measure the average time users spend on the Newsfeed. A limitation to this is measure is if the user is idle on the page (maybe left it open while eating lunch in the other room). We can simply dismiss these cases by excluding those who have not scrolled for longer than 5 minutes (which I think is possible to extract).
  • For retention, it would not be wise to measure visits per week, as users are automatically taken to the page; they can sign in and be brought to the page without engaging with it. So instead, we could measure the number average of interactions (interactions defined the same as the first bullet) per user per week.

I don't want to get into the evaluation and ranking of these metrics because I think that is out of the scope of my focus. Hopefully this is an adequate answer but please give me some feedback since I am still practicing and working on getting better!

Access expert answers by becoming a member
2 likes   |  
1 Feedback
badge Platinum PM
Good answer ; but for these type of questions you need to prioritize your metrics It also shows your prioritization skills as a PM.

Prioritized Metrics

1. avg dwell time of the user (how much time a user is spending scrolling on newsfeed)

2. # of interactions / user - # of likes / comments / shares (would help measure the quality of the content generated in the newsfeed)

3. DAU & MAU accessing newsfeed

4. Churn Rate

5. # of escalations / bugs received per week in regards to newsfeed

6. avg revenue generated / user (the more time spend by user, the revenue genrate will increase *assuming its CPM)
6
Get unlimited access for $12/month
Get access to 2,346 pm interview questions and answers to give yourself a strong edge against other candidates that are interviewing for the same position
Get access to over 238 hours of video material containing an interview prep course, recorded mock interviews by expert PMs, group practice sessions, and QAs with expert PMs
Boost your confidence in PM interviews by attending peer to peer mock interview practices, group practices, and QA sessions with expert PMs
badge Bronze PM

  1. Clarification of the product feature where the user sees updates from friends, pages, groups the user follows Goal. Ads are being shown with various widgets in between like people you may know etc.
  2. Goal - Say the goal is to increase the engagement (time spent) on feed by a Facebook user (as adoption and retention is not really in the picture as a user lands on newsfeed after signing in/opening the platform)
  3. Success Metrics
    1. Avg time spent/user/day
    2. # of like/comment/shares per user per day
    3. # of new groups joined or pages liked
    4. # of new friend requests sent from "People you may know widget" in the newsfeed
    5. # of new posts created/user/week as users will get encouraged looking at friends posting and they would also like to post
    6. % of total FB revenue generated by clicking on ads in the newsfeed
    7. Counter metric: DAU as people might get overwhelmed/unhappy looking at so many happy/proud status updates from friends
  4. I would prioritize #1 and #2 ad the key engagement metrics and the rest as secondary.
  5. Say a follow up question comes asking -- How would you design an experiment to increase the time spent on the newsfeed
    1. Idea: Change the order of the posts that show up at the top -- Instead of showing the latest unseen update from any of the friends show the unseen update from a friend with maximum # of mutual friends (This might be incorporated in the algo already. But, let us assume it is not for the purpose of designing an experiment)
    2. Metric to AB test on: Time spent on the FB feed
    3. Data set to see:
      1. Random sample of users in the experiment group and with the update in the feed algorithm
      2. Calculate the time spent before and after the change
      3. Run the experiment for sample size and duration determined by the Power analysis tool (formula with significance value, uplift expected, and volume of users that we have) that gives us results to determine whether it is statistically and practically significant
      4. Analyze the results whether it is conclusive or inconclusive
        1. Plan the dial-up or dial down
      5. Critical thinking about the analysis:
        1. # of mutual friends might not mean 2 people are close. 2 classmates can have majority % of the friends as mutual friends but might not be close friends. On the contrary, they may not even like each other at all. Showing up a post from this person might in fact reduce the time on the feed as the person might abandon the app as soon as the first is from the other person
      6. Hypothesis:
        1. Best case scenario: If the time, # of posts, DAU increase - Dial-up
        2. Worst case:  If the time, # of posts, DAU decrease -  Dial down
        3. Inconclusive:
          1.  If the time, # of posts increase but DAU decrease
            1. Short term decrease but network effect might increase in long term
            2. If the cost is not too much, dial-up
          2.  If the time, # of posts decrease but DAU increase
            1. Dial down as success metrics dip and DAU increase could be because of something else.
    Access expert answers by becoming a member
    0 likes   |  
    Get unlimited access for $12/month
    Get access to 2,346 pm interview questions and answers to give yourself a strong edge against other candidates that are interviewing for the same position
    Get access to over 238 hours of video material containing an interview prep course, recorded mock interviews by expert PMs, group practice sessions, and QAs with expert PMs
    Boost your confidence in PM interviews by attending peer to peer mock interview practices, group practices, and QA sessions with expert PMs
    Get unlimited access for $12/month
    Get access to 2,346 pm interview questions and answers to give yourself a strong edge against other candidates that are interviewing for the same position
    Get access to over 238 hours of video material containing an interview prep course, recorded mock interviews by expert PMs, group practice sessions, and QAs with expert PMs
    Boost your confidence in PM interviews by attending peer to peer mock interview practices, group practices, and QA sessions with expert PMs