Go or no-go on a new Facebook reaction, and the metrics that judge it after launch

You are the product manager who owns reactions on Facebook: the set of one-tap responses beyond Like that people can leave on posts and comments. You need to decide whether a new reaction should join the set, and, if it does, how to judge whether it worked. ### Clarifying Questions - Has a specific reaction already been proposed (by user research, a partner or leadership), or am I deciding whether any new reaction is worth adding? - Would the addition be permanent and global, or could it be temporary, tied to an event, or limited to some surfaces such as comments? - What does leadership want reactions to achieve: richer self-expression, better feedback for people who post, or a clearer signal for feed ranking? ### Part 1 — Deciding whether a new reaction belongs Describe how you would reach a go or no-go decision: what problem a new reaction would have to solve, what evidence you would gather, which costs and risks you would weigh, and what result would make you stop. ```hint Start where people go quiet Look for moments when people clearly want to respond to a post but none of today's reactions fits, and ask how you could size that from behavior you can already observe. Then ask what one more option does to a tray that everyone uses. ``` #### What This Part Should Cover - Evidence of an unmet need to express a feeling, and how big and how global it is - The costs of another option: a more crowded picker, overlap with existing reactions, and meanings that differ across cultures - Effects on systems that consume reactions, such as feed ranking, notifications and insights for Pages - An experiment and pre-committed criteria that settle the decision ### Part 2 — Measuring success after launch Assume the new reaction ships. Define the metric that tells you it worked, the supporting and guardrail metrics around it, and how you would separate lasting value from a burst of curiosity. ```hint Raw taps flatter the launch Anything new in the tray gets tapped in its first week. Ask what you would compare against to show that people are expressing something new rather than moving taps over from reactions they already use. ``` #### What This Part Should Cover - A primary metric that captures incremental expression rather than raw usage of the new reaction - How much the new reaction draws from existing reactions and from comments - Guardrails for misuse, conversation quality and feed quality - A time horizon that separates novelty from sustained use ### What a Strong Answer Covers - One explicit goal for reactions that drives both the decision and the metrics - Treating reactions as a shared signal used by ranking, notifications and the people who post, not only as an interface element - An experiment design that accounts for reactions being social, since one person's reaction appears on another person's post - Attention to cultural legibility, accessibility and potential for harassment - Clear outcomes for each result: ship, iterate on the icon or placement, or drop ### Follow-up Questions - The test shows heavy use of the new reaction, but most of it replaces Love and Like and total reactions are flat. Do you ship? - How would you handle a request for a temporary reaction tied to a global event? - A group of users starts using the new reaction sarcastically to pile onto certain posts. What do you do?