Clarifying questions:

• Reactions mean ? – 5 types of likes – laugh, love, thumbs up etc.
• Fb mission – help people across the globe built communities
• Significance of reaction:
  • It indicates how people perceive certain posts to the content creator
  • It helps build correct recommendation engine for the user, showcasing relevant post & ads
  • All in all increases engagement & ad revenue

What problem are we solving – It’s expanding the ways as to how people can react to posts/videos/listings etc. beyond the 5 predefined reactions. It might help define better content showcased as well as convey the creators more meaningful insights.

Why are we solving it –

• The distribution across 5 current reactions – will help me get an idea of what are people using and how often
  • Get data cuts across geography, age, profession etc.
• Secondary research – existing reviews on playstore and other channels will tell us if there is any signal from current users
• Primary research – are people even looking for options ? If so, what other options would they want ?
• Competitive analysis – are other competitors providing the option to add any other reaction etc.

Why solve it now ? Business goals of instagram – To increase engagement across the platform for both content consumers & creators

• DAU & WAU (active user is one who engages on posts – only reactions)
• Ad revenue

Impact – Medium to high, although might want to consider the downside of adding too many new reactions. Need to strike a balance. Effort – Low, as we already have the reaction feature in place.

Risks/Trade-off:

• Too many options – might slow down decision making for users. We need evaluate the point beyond which we shouldn’t add more.
• Does adding new reactions impacts other metrics like comments/share/time spent/content created etc.

GTM strategy:

• Tool tip when the user uses reaction or the first post after launch
• Utilize video ad spots to introduce new feature launch

CQ:

• My understanding of reactions are emoji like stickers/responses that allow users express themselves beyond “likes” without having to speak or type a word. These are meant to be quick and convenient to promote interaction between people. Correct? [Sure]
• Based on my usage/understanding of Instagram (IG), I believe there are two types of reactions on IG: 1) Reactions in Stories (limited to 6 without ability for users to customize), 2) Reactions in DMs (unlimited/customizable via a + by double tapping the message). So are we intending to launch additional reactions in the Stories or DMs? Do we know how many additional we want to launch or is it up to me to decide? Did I miss any other use cases?
  • Interviewer: Assume we need to launch additional reactions for Stories. You decide on the number.
  • Ok, great.
• Provide Framework: Ok now that I have a fair understanding of the product functionalities/scope of the ask, I want to quickly touch on the Goals of this feature in the context of the product and the overall ecosystem and how it benefits the users and all parties in the ecosystem. From there I would like to build hypothesis on how enabling additional reactions can impact our product goal, target users, etc. I will also touch on the core metrics, user related metrics, and recommend solutions with some trade-offs in mind.
• Business Goal:
  • Instagram overall mission is to capture and share world’s moments, which aligns with the overall Meta vision on empowering people to build community and bring the world closer together.
• Product/Feature Goal:
  • Reactions provides a convenient way to enable interaction (and therefore engagement) between parties; making it easier and faster for people to interact and engage with the content they love which then spins the content engagement flywheel. Easier reaction > prompt more user to engage> more engagement with creator content > more creators are encouraged to create.
• Hypothesis:
  • By adding additional reactions we will positively impact engagement because we will provide more options to people to express their feelings beyond current offerings
  • Potential Risk: While this might help user quick interact with Stories, it might negatively impact deeper engagement with stories (i.e., comment, and back and forth messaging that is prompted by viewing the stories in the absence of wider selection of quick reactions. I will park this for now and revisit while thinking about counter metrics.
• User Segments:
  • Content Creators
    • Goals/Why reactions matter to them? Get more engagement with their content, get a sense of what their viewers like/dislike
  • Content Consumer/Viewers
    • Goals/Why reactions matter to them?: Explore/Discover more relevant content, make it easier to connect with people they love and content they enjoy
• Metrics (topline):
  • # of active users with >1 engagement with reactions in stories weekly
    • This metric matters because it shows a healthy balance between content creator and content consumer
    • I chose weekly because every day user may not necessarily post content daily or be active on Instagram on a daily basis.
• Breakdown of Metrics by Users and How I would prioritize what
• (Build a table)
  • User
  • Metrics
  • Why It matters

Content Creator

1. # active creators per week/weekly active people
   1. Tells me what % of people creating content, helps us ensure healthy supply
2. Avg # of stories posted/creator/week
   1. Tells me how often creators are posting since we expect this to increase if engagement is increasing
3. # active creators who received >1 reactions
   1. This metric is important because it will help me understand the impact to their overall engagement

Content Consumer

1. # of stories watched
   1. Input to the following metric
2. # of stories with >1 reactions per week
   1. More videos watched means more opportunities to engage with the content

Platform (Insta)

1. Avg time spent on IG/Weekly active users
   1. Tells us how long people are spending time on IG and we improve certain products the time should go up
2. Weekly Active Users (WAU)
   1. If we make the product better we would like see this metrics go up

In the next section, I am going to identify and list Hypotheses Impact on the Key Metrics discussed above. This will allow me to define success criteria before running an A/B test. Hypothesis Impact on Key Metrics I will narrow down metrics from above and prioritizing the one that are important to track in A/B testing (Build a table) Metric Impact Why?

1. # of stories with >1 reactions per week
   1. Increase
   2. Since we add additional reactions, people have more options to quickly react to the stories; therefore the engagement will go up
2. # of active users with >1 engagement with reactions in stories weekly
   1. Increase
   2. More activity since people have more options to choose from
3. Weekly Active Users (WAU)
   1. Increase
   2. Overall activity can go up since more choices for reactions are available
4. # of deep engagement (impact to conversations initiated from stories)
   1. ? Decrease
   2. People may choose quick reactions over conversation if they find a relevant reaction emoji (ex: baby angel face emoji is faster to send than typing “he/she is an angel” in response to baby story content.

Variations:

• Based on the above criteria, I will design an option where creator can choose the reactions between 6 to 9** but default to one pending test results

** Caveat: I will work with VUI designer and data scientists to determine the right number for additional reactions. I could also do a usability to see what numbers can become overwhelming to users. For now, I chose 9 based on the aspect ratio of stories , I don’t want that the emjois to completely block the screen or become too crowded/confusing to choose from. A/B tests:

• Control: Keep reactions limited to 6 with no option to customize or change quantity
• Treatment 1: Stories will have 9 reactions to stories
• Treatment 2: Stories will have 9 reactions with option to change to 6 reactions

Ship, No Ship, Retest Decision Ship:

• Whichever option optimizes the four metrics mentioned above

No Ship:

• If any or all of the key 4 metrics impacted negatively
• If there are major regressions

Retest:

• If no statistical significance. Inconclusive

Overall, I will continue to monitor the performance on primary and secondary metrics as well as counter metrics to ensure we are not cannibalizing other areas of the product. Additionally, I will work with GTM team to propose a stepwise roll out plans across devices, geographies to minimize the risk of disengagement.

A high level framework

Clarifications:

Reactions already exist in Instagram?

Any specific region or user group we try to target for the launch?

…

Product description:

IG is …

Reactions are Emoji alike icons users can use to reply to posts/photos/comments

FB’s mission:

Help people stay connected and build communities

IG’s mission/goal:

Allow users to capture world’s moments and share with friends

Reactions Goal:

To provide users an easier way to react to the content and to better express themselves beyond “like”. Promote interactions between users, which fits into both IG and fb’s missions.

User personas this change might impact:

Content creator

Content viewer

Success metrics for Reactions:

Content creator

# of reactions received per creator per day

# of reactions received per post per day

# of different reactions per post per day (e.g 2000 of Love, 1500 of Haha)

Content viewer

# of reactions used per viewer per day

# of different reactions used per viewer per day

Success metrics for IG:

Since the feature eventually impacts the overall health of the product, we need to keep track of the product success also. Some critical metics:

Content creator

# of posts per creator per day

Content viewer

# of post impressions per viewer per day

# of post shared per viewer per day

Counter metric:

Growth of reactions usage might cannibalise other features, e.g. Comments

# of comments per viewer per day

# of comments per post per day

North star metrics:

Based on the fact that: 1) Reactions’ goal is to promote user interaction 2) we want to test whether we should launch new reactions

NSM:

# of different reactions used per viewer per day

also keep an eye on # of post impressions per viewer per day

A/B test with the north star metric in mind

Summarise