Measuring whether in-meeting emoji reactions on Zoom work: metrics and causal read

Zoom has added emoji reactions that participants can send during a meeting. As the product manager, how would you determine whether the feature is effective? Explain what the feature should achieve, which metrics you would track, and how you would tell whether changes in those metrics are caused by the feature. ```hint Effective at what? A count of emoji sent says little on its own. Decide what problem reactions are supposed to solve in a video meeting and for whom, then ask what should change if the feature works and what might get worse. ``` ### Clarifying Questions - What goal did the team set for reactions: more participation from quiet attendees, live feedback for the presenter, or fewer interruptions? - Was the feature released to everyone at once, or is there a holdout group or staged rollout I can compare against? - Which surfaces carry it: desktop, mobile and web clients, regular meetings only, or webinars as well? - Can hosts or account administrators turn reactions off? ### What a Strong Answer Covers - A clear goal for reactions grounded in what video meetings lack, and the participants it serves - Metrics across adoption, repeat use, the outcome the feature exists for, and guardrails, each precisely defined - Segment cuts that matter for meetings: size, host versus attendee, muted or camera-off participants, platform - A causal method that fits how the feature was rolled out and handles the fact that everyone in a meeting sees each reaction - How each pattern of results would translate into a decision, including whether reactions displace chat or hand-raising ### Follow-up Questions - Adoption is high, but chat messages per meeting have dropped. Is that good or bad, and how would you tell? - Presenters in large meetings say floating reactions distract them. What would you measure, and what might you change? - How would you decide whether to add more reaction types?