In early 2023, a webtoon artist — a creator of the serialized digital comics that dominate Korea's online publishing scene — posted a short note on social media. After six years of publishing a new episode every week, a single overhaul of the platform's recommendation algorithm had erased nearly 70% of the work's monthly income. The artwork was as strong as ever. Reader ratings held steady. The only thing that changed was that the platform had decided to stop connecting that artist's fans to that artist.
For a long time, quality was the condition for survival. If you made things better than your competitors, you could at least expect to stay in the market. As generative AI lowers the barriers to content production, that expectation is expiring fast.
Polish Is Getting Close to Free
Since generative AI took hold in earnest in 2023, the content market has shifted in visible ways. Tasks that once belonged exclusively to skilled professionals — image generation, video editing, copywriting — can now be handled quickly by anyone. Adobe Firefly generated more than 12 billion images within a year of launch, and a substantial share of them came from ordinary users, not professional designers.
When supply rises, prices fall. The pattern shows up clearly in the recent results of stock-image platforms. Submissions of AI-generated images grew by hundreds of percent, while per-image earnings shrank sharply over the same period. The price collapse had nothing to do with declining quality. With supply surging, polish alone could no longer hold up the price.
The same dynamic is playing out simultaneously in video, music, and text. Every minute, 500 hours of video are uploaded to YouTube. In that environment, the channels that survive are mostly not "places that post well-made videos" but "places people have a reason to come to." Industry observers in Korea and abroad argue that the axis of competition in the content business is moving from production quality to ecosystem ownership. Fandom, data, community, platform — whoever directly owns even one of these, the thinking goes, keeps their bargaining power.
Owning Your Fans vs. Renting Them From a Platform
The problem is that most content creators don't own their fans — they lease them from platforms. Fifty thousand Instagram followers are people Instagram has temporarily granted you access to. If the platform changes its algorithm, rewrites its policies, or shuts down one day, the question worth asking is how many of those 50,000 could find their way back to you directly.
Netflix understands this structure better than anyone. As of 2024, Netflix's monthly active users surpassed 300 million. But those 300 million are Netflix's customers, not fans of any particular director or writer. When a star producer defects to a rival platform, most viewers don't follow. The fans are bound to the platform, not to the creator.
Cases of directly owned fandom tell a different story. For her 2023 Eras Tour, Taylor Swift distributed concert tickets primarily through direct channels that didn't depend on any platform, and engineered a surrounding economy worth more than a trillion won — roughly $750 million — on her own terms. That same year, Starbucks counted 33 million active rewards members in the United States. Starbucks doesn't serve the world's finest coffee, yet that member data is an asset no café can replicate on short notice.
For a One-Person Business, "Design an Ecosystem" Is Still Mostly a Slogan
At this point, an objection deserves an honest hearing. "You need to design an ecosystem" is hard to take as practical advice if you're a solo creator or a small operator. Running a newsletter, maintaining a paid community, and analyzing reader data all demand time, money, and sustained execution. Asking a one-person operation to build an ecosystem while also making the content is, in effect, assigning two people's jobs to one.
The numbers back up how hard this is. On the newsletter platform Substack, fewer than 5% of publishers sustain paid subscription revenue above $500 a month. Plenty of people start an email list only to abandon it, worn down by the time and cost of upkeep. Knowing the value of an owned channel and actually keeping one running are two different things.
The objection is entirely fair. But the fact that building an ecosystem is hard is not evidence that it can safely be deferred. One reason most solo entrepreneurs push it down the priority list is the belief that, for now, content quality alone will carry them. How fast generative AI is eating away at that belief is where this essay began.
Is What You're Building an Asset — or a Lease?
If directly owning your fandom, data, and community now matters this much, how can a solo entrepreneur or content director take stock of where they actually stand?
One approach is to look past the surface of a product or service to the depth of the relationships it leaves behind. Veterans of marketing and design observe again and again that consumers respond less to the product itself than to the experience and sense of belonging it creates. Buying something, in this view, goes beyond simple ownership; it's closer to participating in the worldview the brand expresses. Seen this way, content quality does the work of bringing people in the first time, and the ecosystem provides the structure that makes them stay.
A workable self-audit starts with simple questions. First: do you have a channel that connects you directly to your readers and subscribers? That means routes that bypass platform algorithms entirely — an email list, a KakaoTalk Channel (the broadcast feature of Korea's dominant messaging app), a Telegram group. Your social media follower count is irrelevant to this question.
Next, consider this: if you stopped publishing today, would anyone still be searching for your name three months from now? With a steadily accumulated body of work and recognition within a community, people will come looking for you even if the platform disappears. That's the moment your relationship with your fans starts shifting from a lease to ownership.
It's also worth checking whether the content you make is accumulating data. If you can track which topics get a response, and which formats get read at which hours, your content production grows more precise over time. Raising polish without response data is like insisting on the same direction with no feedback.
Building an ecosystem doesn't mean building a major platform yourself. An email list of 200 people you manage directly can be a more stable asset than 20,000 followers who belong to a platform. Start small — but secure at least one connection that doesn't depend on anyone else's algorithm. That's the starting point.
The ability to make great content still matters. But in a market where it guarantees nothing on its own, it's worth asking, at least once, whether what you're building right now is a body of work — or a set of relationships.
