An engagement metric is any quantitative indicator used by platforms, publishers, and researchers to measure the extent to which audiences interact with a piece of content rather than simply view it. Common engagement metrics include likes, reactions, shares or retweets, comments, replies, saves, click-through rate (CTR), average watch time, completion rate, and dwell time. Aggregate measures such as "engagement rate" typically express interactions as a share of impressions, followers, or reach.
For political researchers and MUN delegates, engagement metrics matter because they shape what information audiences encounter. Algorithmic ranking systems on platforms such as Facebook, YouTube, TikTok, and X weight content partly by predicted engagement, which can amplify emotionally charged, polarising, or sensational material. Internal documents disclosed by whistleblower Frances Haugen to the U.S. Securities and Exchange Commission and the Wall Street Journal in 2021 indicated that Facebook's 2018 shift to a "Meaningful Social Interactions" ranking signal — itself an engagement-weighted formula — increased the visibility of divisive political content.
Engagement metrics are central to debates over platform accountability. The EU's Digital Services Act (DSA), which entered into application for Very Large Online Platforms in August 2023, requires designated platforms to assess and mitigate systemic risks arising from the design of their recommender systems, including engagement-based ranking. Researchers also distinguish between:
- Active engagement (comments, shares) — signals endorsement or contestation.
- Passive engagement (views, watch time) — signals attention but not intent.
- Negative engagement (angry reactions, reports) — sometimes still boosts reach.
Critics, including former Google design ethicist Tristan Harris, argue that optimising for engagement conflates attention with value. Counter-proposals include "bridging-based ranking" (used in X's Community Notes) and time-well-spent metrics, which reward content that diverse users find useful rather than merely reactive.
Example
In 2021, whistleblower Frances Haugen disclosed internal Facebook research showing that the platform's engagement-weighted ranking change in 2018 amplified divisive political content across European elections.
Frequently asked questions
Because platforms rank content partly by predicted engagement, material that provokes strong reactions — including misinformation and outrage — can spread faster than measured, accurate reporting, with consequences for elections and public health.
Keep learning