Step 1: Find a type of website experience that sucks for users. Example: Blogging, image sharing.
Step 2: It's so simple! Just do the same, improve on the UX, don't plaster ads, have a clean site and give it a cool name. Example: Medium, Imgur.
Step 3: As people flock to your site because it's cleaner than the alternatives, have the lightbulb moment: "Oh wow, I could monetize all these users!". So you plaster the site with ads, marketing gimmicks and what not and you don't suffer for it. It's not that it works. It's just that you have a userbase that trusts you and is already used to you. They don't know anyone else.
What are your users gonna do, build their own site that doesn't suck?
It really is a depressing, wasteful cycle.
I do agree with this, but I believe we can make it at least not-as-bad.
I recently started a new blog & evaluated a load of options, Medium was one. I quickly decided it had become a cesspool of attention engineering and social widgets, and ultimately settled on self-hosting, with no banner adds or social rubbish.
Can the average public do this? I’d argue it’s not trivial nor inexpensive currently, but maybe that’s worth fixing? Encourage people to move off centralized platforms for sharing content, and maintain control.
I recently moved my company's blog to it. This is an open source, self-hostable markdown blog engine with an excellent featureset and some great default themes, and a paid hosted plan.
Official Ghost blog: https://blog.ghost.org
My company's blog: https://articles.hsreplay.net
A lot of it comes down to financial incentives, and I think VC-backed products are going to go head-first into step 3 without thinking, and then we all suffer. If there were more small, grow-slow products (probably bootstrapped) that based success on delivering a nice product that never goes to shit (instead of constantly conquering the world), we might have less of this cycle.
[0] https://write.as