If you were the PM for Facebook Watch, how would you decide between releasing an entire 10-episode series every week, or releasing one episode a week for 10 weeks?

Clarifying question:

Can you confirm how much is the avg time per episode & what is the average watch time of viewers on watch?

This is for Facebook originals. Right?

The watching behavior is correlated to the genre of the content. If it is a reality show, I wouldn’t prefer watching it all at once. However, for thriller content I would want to get to the climax quickly. For this discussion, is it ok if I focus on one genre of content?

If I think about the role of watch in fb portfolio, I think it bring people back to fb app and discover new videos

What is the product?

Facebook watch allows user to discover and watch videos — short form and long form. It’s a place where people can follow video creators they care about, start conversations about videos with friends, and build communities of fans who share their interests.

Mission: Entertain the world with engaging stories and bring world closer together

Goal: Watch product is known more for short form of entertaining videos. Given the question is more about series & our mission is to entertain the world with engaging stories I will focus on number of interactions as my primary goal. The number of interactions will tell me whether people find the stories and episodes engaging or not.

Users: Creators (individual creators, uploaders, FB original series producers), Viewers (Fans, Binge watchers), Moderators

At a high level, the funnel would look like

1. Viewers discover the video
2. They watch it with their friends & families
3. They react/comment/share the video
4. They come back to watch

To evaluate the tradeoff, I would look at

1. #Interactions
2. Daily viewers
3. Returning visitors to all up Facebook app
4. Watch time
5. Does it help FB all up expressed as incremental DAU for FB due to Watch
6. Quality of content expressed as: Completion rate of the video, #interactions (Reactions, Likes, Shares)

Pros & Cons

To understand whether to release entire series every week or 1 episode a week for 10 weeks I will try to understand the pros and cons of doing 1 episode a week first

Pros

• Curiosity: Releasing 1 episode a week builds curiosity among fans
• Returning visitors: Acts as a trigger for users to come back — Increase in returning visitors to FB
• Limited time: Users have limited time so this strategy might align well with busy professional or busy consumers
• #Interactions: People will engage in more conversations if the curiosity is high

Cons

• Churn risk: Risk of churning binge watcher cohort

Launching entire series

Pros

• Watch time: Helps increase watch time on daily basis
• Binge watchers: Helps us acquire and retain binge watchers
• Consumer choice: Consumer needs things on-demand. They are less patient and need instant gratification. Giving the decision-making choice to users will result in less dissatisfaction.

Cons

• #Returning visitors: Getting people back to the platform becomes more difficult as there is no hook
• No unique advantage: It might be a metoo approach and customers might compare it to typical subscription like Netflix prime

Looking at the pros and cons of both side and the overall analysis, my primary preference is launching one episode per week as it optimizes for #interactions and #returning visitors(that is important for all up fb). My intuition is people will engage in more conversations for the one week and try to anticipate whats coming up next . To validate that, I would first run a test with thriller genre.

Hypothesis: If we launch one episode per week, we will see higher number of interactions

• Variant: Launch one per week
• Control: Existing Content
• Cohort: Thriller
• Primary metric: #interactions
• Counter metric: As we are launching once per week I want to make sure the daily time spent doesnt drop. To ensure that doesnt happen I will monitor Time spent on daily basis
• Secondary metrics: completion rate of videos

If we don’t see stat sig improvement in #interactions, I will consider running a second test of launching all episodes at once.

Hypothesis: If we launch all episodes , we will see higher number of interactions

• Variant: Launch all episodes
• Control: Existing Content
• Cohort: People who watch thriller genre
• Primary Metric: #interactions
• If we launch 1 episode/week, the risk is not everyone will come back daily. To ensure that doesn’t happen my counter metric is
• Counter Metric: Daily viewers

Once I understand both perspectives, I will ensure the overall FB metrics are also up and I don’t cannibalize the FB DAU. If that is positive, I will go ahead with that specific strategy for the thriller genre.