When a platform pulls the boost from its favored creators, how much of their popularity was ever really their own?

This explores what happens when a platform pulls back the extra reach or other perks it once gave selected creators. The collection has no study that measures what happens to a creator's audience after a boost ends, so it can't settle that part of the question. What it does show is why boosts get handed out, why they get taken away, and why the popularity built on them may be thinner than it looks.

Start with how the boost works. Forbes reported that TikTok used an internal "heating tool" to push videos from selected influencers and brands into millions of feeds, and that the choice was tied to winning business partners rather than to how audiences responded Does TikTok use special boosts to inflate partner videos?. That matters for the question: if a creator's view count was partly made by the platform, losing the boost means losing reach the creator may never have earned with their audience. Neither the creator nor their followers can easily tell where the earned part stops and the manufactured part starts.

Cory Doctorow's "enshittification" lifecycle helps explain why platforms take these perks back Do platforms inevitably decline through value extraction cycles?. In his account, a platform first gives value away to attract users, then to attract business customers, and then pulls that value back for shareholders. Seen this way, a boost is a recruiting expense, like a subsidy. Withdrawing it is not a malfunction. It is the platform moving on to the next phase once the boosted creators have done their job of drawing in audiences and advertisers. One caveat: Doctorow's evidence is illustrative examples, not a systematic sample, so treat this as a strong framework rather than a measured result.

The same lever can push in the other direction. On one platform, the algorithm gave AI-generated content less exposure than matched human-made posts across about 179,000 pairs. That suggests platforms already adjust reach by category, not just account by account Can algorithmic distribution prevent AI content from overwhelming creator diversity?. If the platform's priorities change, human creators who benefit from that tilt could lose it. And AI content is already competing for the engagement influencers depend on, while building no lasting reputation for any real person Does AI content displace human influencers on social media?.

Here is the part you might not expect: a strong track record may not protect you when conditions shift. A study of the freelance platform Upwork found that freelancers' past performance did not shield them from the drop in work that followed ChatGPT. Top freelancers may even have been hit harder Does a strong track record protect freelancers from AI?. That study is about AI competition, not boost withdrawal, but the parallel is worth taking seriously. Success on a platform often depends on conditions the platform controls, so a creator who looks secure may be among the most exposed when those conditions change.