How would you improve Facebook Reactions?

You should start to answer this product improvement question by asking clarifying questions:

Clarifying questions:

  Why are we looking to improve -  get more adoption & engagement

  Any specific geographic regions - world

  Any specific platform/device - no. all platforms/devices

  Is this just for the main Facebook app or Messenger and other products. Just Facebook.    

  Any particular gaols ? engagement 

  Any particular user types ? No you pick

   First Let me explain FB Reactions and how it fits with FB missions.

   FB Reactions is a set of 6 new emojis that allow you to react to a post in addition to the traditional like. FB's mission is to bring the world closer together and provide tools to build communities. By adding reactions FB helps people to express themselves more and thus bring them closer together 

    Approach:

   I will list down some user personnas, then pick a user personna, list some pain points. Then prioritize and pick some pain points.

   Then I will list down some solutions, prioritize these solutions. Provide some metrics to measure the new feature

   and finally summarize my recommendation

   User personna:

   There are 2 main user types here - content creators - people who create post and would like to view reactions and content viewers - people who view and react to posts.

   I will select content viewers as the vast majority of users are content viewers.

   Pain points:

   1. What reaction should i give for the post                                 -  High value, Medium complexity  

   2. Is the reaction I give appropriate - High value, Medium complexity

   3. I want to see posts with a certain type of reaction                      - Medium value, Low complexity

   4. Where can i see trending post for a certain reaction  - Medium value, Low complexity

   5. I do not want to miss any posts by my best friend. Can I automatically react to all her post.  - Medium value, Medium complexity

   6. This is a great post, I want to recommend similar post to my friends     - Medium value, high complexity

 I would pick the first 2 pain points as they give high value also both pain points will have similar solutions.  

Solution:

   Reaction recommender - Using AI and Machine learning, for each post this feature would recommend a reaction. User can click and the reaction gets added to the post. 

   # of people per reaction - This feature will show the total number of people with each of the reactions. Users can then decide what reactions to give.

   Automated reaction - User selects few people to received automated reactions. Whenever a user's friends has a new post, within 10 mins this feature will automatically react to the post. Automated reaction feature will be smart enough to understand the user's post and post an appropriate reaction. 

    Impact Effort

1. Reaction recommender   H M

2. # of people per reaction               M L

3. Automated reaction                  M M

Based on the prioritazion, it looks like Reaction recommender would have the highest impact. Also looking back to our goal of engagement, I believe adding this feature will increase engagement. As users see different types of reactions on a post they would engage more.  Also, now there would be a reaction recommended for each post, the user does not have to think about what reaction to give, FB automatically recommends you the reaction. This is much easier and will increase engagement. 

Some metrics I would use to measure success of this feature: 

1. %of reactions clicked through Reaction recommder. 

2. # of user that have used reaction recommder at least once

3. # of user using reaction recommeder every day/week/month

Summary:

My task was to improve FB reactions. The goal of improvement is to increase engagement for FB reactions. 

I narrowed down user type to be content viewer and suggested 3 solutions. Based on my prioritization I selected Reaction recommender as my solution. 

Finally I suggested few metrics to measure the success of reaction recommender.

Scope: There are 7 reactions. Like, love, care, haha, Wow, Sad, Angry. 

Customer journey:

User sees a post. Holds thumb on like, the reactions expand, and clicks one of them. 

What is the goal for the team?

Overall goal - lead to increase in engagement. DAU, DAU/MAU etc. 

Reactions specific metrics: 

1/ Are 7 reactions used? 

2/ Whats the msot popular 

3/ How many users actually use teh reactions? 

4/ How many users have used like, but never used any other reaction? FOr e.g, many may not even be aware. 

What is the product vision for reactions? What was the vision when it was launched?

When you see content - you want to express what you feel about that content. 

What are all the possible reactions -- many are missing - frightened. Are we reflecting our emotions? Many emotions are missing - fear, anxiety, peace, disgust are missing. I am sure there are fundamental emotions missing. 

What was the reason these 7 were chosen? 

Others include - endorse. 

There are 3 buckets 

1/ Existing reactions → how to increase engagement 

2/ WHich new reactions can we add?

3/ How to apply reactions to content. 

FOr e.g, Long form content like TV show. You can have several emotions through out the show. 

Lets look at customer journey 

Without understanding the product vision, its difficult to improve. I will go with enhancing the reaction with audio or video clips