Aubrium Blog
Research, evidence and original analysis on how social platforms work — no listicles, no recycled advice.
What you cannot do when you don't post from the app
Publishing through Instagram's API is not penalized. It is limited. Shopping tags, story stickers, most of the music library and mid-flight collaborator changes do not work. Here is the documented list, taken off Meta's own pages.
The watermark penalty: recycling your TikToks to Reels
Instagram has said three times, in public, that it will not recommend reels carrying a visible watermark. TikTok's version sits in a developer page almost nobody reads. Neither platform will tell you how big the penalty is.
The setting between allowed and removed that Meta admits exists
Meta says shadowbanning is not real. It also publishes a framework for reducing the distribution of content that stays up. Both are true, and the second one is written down.
Does the "Made with AI" label cost you reach?
Meta and TikTok both label AI content, and creators assume the label buries the post. TikTok states in writing that labeling does not affect views. Meta's policies describe the label and never describe a penalty.
Instagram's war on reposts, in its own words
Meta promised to rank originals above copies, label reposts, and cut serial aggregators out of recommendations. Here is what the company actually committed to in writing, and what it left undefined.
The tactics the gurus sell are the ones Meta says it demotes
Comment YES. Tag three friends. Meta has an explicit policy against engagement bait and says it demotes the posts. The policy is written for Facebook, and no Instagram ranking document mentions it.
Comment X and I'll DM you: does the automation work, and is it allowed?
The comment-triggered auto-DM runs on Meta's own API and stays inside two documented windows. The post that asks for the comment is the part that may count as engagement bait, and Meta shows engagement bait to fewer people.
The shadowban you can actually check
Instagram publishes a recommendation eligibility status you can see right now. The folklore version of the shadowban has no evidence behind it.
