Key concerns are:
Comment Visibility: Comments are not inherently visible in user feeds - only the main post is. As such, assigning impressions based solely on subreddit subscriber count for mentions within comments may not accurately reflect real visibility.
Suggested Approach: They recommend exploring a more nuanced method for calculating impressions on Reddit, especially for comments. This could include factoring in engagement metrics (upvotes, comment count, etc.) from both the post and the comment itself, possibly applying a weighted proportion of the subreddit audience size rather than the full count.
At the moment Reddit impressions are skewing data in reports as they are very unrealistic. I propose we review the current methodology and explore whether a more engagement-driven model can be implemented for Reddit impressions, especially in comment threads.
