The pitch is reasonable on its face: a group of creators agree to like and comment on each other's posts, everyone gets an early engagement boost, everyone wins. It does not work, and the reason is instructive.
The signal is the wrong shape
Recommendation systems do not count engagement, they model it. What they are trying to predict is whether a given user will respond well to a given post. Engagement from the same forty accounts on every post, regardless of subject, teaches the model that those forty people like everything you make — which is information about them, not about the post.
It contaminates your audience data
Worse, it makes your own analytics useless. You cannot tell which posts genuinely landed because every post carries the same artificial floor. You lose the feedback loop that would have told you what to make more of.
The audience mismatch
Pod members are usually creators in adjacent niches, not your actual audience. Signals from them push your content toward more people like them — other creators — rather than toward the people who would follow you.
What to do instead
Genuine community participation looks similar from the outside and works completely differently. Show up consistently in two or three community tags, reply to people whose work you actually find interesting, and let the reciprocity be real. Slower, and it compounds instead of contaminating.