Setting a launch success target for Facebook Reactions per post, overall and per reaction

You are the product manager responsible for Reactions on Facebook, a new way to respond to a post that adds Love, Haha, Wow, Sad and Angry alongside Like. Define what a successful launch looks like, expressed as the number of non-Like reactions a post receives, and explain how you would arrive at that number. Then say whether the expected number should be the same for each of the five new reactions. ### Clarifying Questions - What window does "at launch" cover: the first day, the first few weeks, or the point where usage levels off? - Does Reactions reach every user and country at once, or roll out in stages? - Should the target be one average across all posts, or set separately for posts from friends, Pages and groups? - Which historical data can I assume I have, such as Likes per post, comments per post and the tone of comments? ### Part 1 — Set the launch target Propose a number of non-Like reactions per post that would count as success at launch, and show how you derive it. ```hint Find a baseline you could measure Pick a starting figure you could pull from data that exists before launch, such as the feedback people already give on posts, and lay out each step between that figure and a non-Like reaction within the launch window. ``` #### What This Part Should Cover - A precise metric definition: what counts, which posts and what window - A baseline taken from feedback behavior that exists before launch - Labeled assumptions for each step from the baseline to the target, with the arithmetic shown - A target stated as a range, with a sanity check ### Part 2 — Does the number differ by reaction? Say whether each of the five reactions should be expected to reach the same number per post, and explain why or why not. ```hint Picture the posts behind each reaction For each of the five reactions, think about which posts would call for it and what using it signals to the person who posted, then decide whether one shared number makes sense. ``` #### What This Part Should Cover - A clear position on whether the numbers should differ, with the reasoning - The factors that could make the reactions behave differently or alike - What success would mean for each reaction, and how it would be checked ### What a Strong Answer Covers - A launch target tied to the goal Reactions serves, with its derivation, rather than a number chosen by feel - Arithmetic that is correct and consistent with the stated assumptions - How the distribution of reactions across posts and post types affects the measure chosen - Guardrails for what the new reactions could displace, and how each would be monitored - How the target would be calibrated or revised as real usage data arrives ### Follow-up Questions - If non-Like reactions per post beat your target but comments per post fell, would you call the launch a success? What would you check first? - How would you tell whether a rise in Angry reactions means the feature is working or that the content people see has become more upsetting? - Would you set one global target or one per country, and why?