In-time feedback which alerts users as they are drafting content that they may be breaking a forum rule and guides how they can modify their content. As developed on Reddit, each subgroup can customize the rules and feedback.

By giving in-time feedback, Post Guidance will give "good faith" users - who may be unaware they are breaking a rule - an opportunity to correct their post and reduce the chances it will be removed (manually or by an automod). This can lead to more high quality content and engagement among good faith users.

A field experiment conducted with thirty three subreddits provides convincing evidence that Content Guidance results in more rule-conforming posts and overall higher quality engagement.

In collaboration with Reddit, Ribeiro et al. (2024) recruited subreddits to participate in a randomized controlled trial to test the effects of the then new tool "Post Guidance". Each participating subreddit customized the tool to include at least two of their rules that would trigger feedback to users as they draft or are about to post content. Over a period of 35 days in each subreddit, users were assigned to either the test condition, with Post Guidance, or a control group that wouldn't see the tool. Researchers then collected data for each user for a period of 28 days.

Overall, researchers found that users who were in the Post Guidance group produced 5% more rule-conforming posts (i.e. posts that were not removed by mods or automod within three days). They also observed that posts in the Post Guidance condition received more positive engagement (upvotes and comments) than those in the control condition. (Note: all results are statistically significant unless otherwise noted.)

The researchers note that underneath those overall positive results there are likely two separate dynamics, one in which "good faith" users are successfully posting more and becoming more engaged, while in the other "bad faith" users are either dropping out or learning ways to avoid deletion by automods. That twofold dynamic may explain why two of the researchers' hypotheses didn't pan out: Post Guidance didn't result in less work for mods (possibly because they had to catch more bad faith automod evaders); it also didn't increase overall greater future engagement (possibly because bad faith users were dropping out).

Online communities frequently need to use automatic removal tools to manage the influx of comments while supporting healthy engagement, yet automods may frequently remove well-meaning posts that inadvertently broke a rule. Post Guidance tools can engage and empower good faith users to "self-correct" and improve the overall quality of content in an online forum.