The open web is not dead. But it is no longer where most people spend their time, and the gap is widening. The numbers tell the story more clearly than the eulogies.
According to Similarweb and Chartbeat, the share of total web traffic going to the top 100 websites (excluding Google, YouTube, Facebook, and a handful of search and social infrastructure) is roughly the same as it was a decade ago. What has changed is the share of attention. Time spent on the open web, defined as news sites, blogs, forums, and topical sites outside the walled gardens, has been declining steadily for a decade, and the decline accelerated after 2020.
Here is the data, with the sources, and the version of “what now” that is honest instead of nostalgic.
Where the attention went
Mary Meeker’s annual internet trends report (now discontinued, but the data is preserved) tracked the shift for years. As of the 2019 edition, Americans spent roughly 6 hours per day on digital media, of which about 3.5 hours was on mobile. Of that mobile time, the majority was inside apps, not browsers. Inside apps, the majority was in five: Facebook, Instagram, YouTube, TikTok (from 2020), and Messenger.
By 2024, the same data showed the open web’s share of total digital attention below 25 percent. The walled gardens had the rest. The growth came from short-form video, group chats, and private communities, all of which are not the open web in any meaningful sense.
For news specifically, the numbers are starker. The Reuters Institute Digital News Report, which has tracked this for years, found in 2024 that for under-30s, social media and video platforms are now a larger source of news than news websites. The order varies by country. The direction does not.
Why the open web is losing
Three reasons, in order of importance.
Distribution moved to private channels. The link economy of the 2010s, where sharing an article on Twitter was a major source of traffic, has been replaced by DMs, group chats, and ephemeral stories. None of those surface a URL. None of those generate a referrer. None of those reward the publisher for producing the thing being shared. The article is still being read. The publisher does not know it, cannot measure it, and cannot monetize it.
Search became an answer engine. Google’s pivot to AI Overviews (previously SGE) is the latest in a long line of changes that put the answer on the search page. The user does not need to click. The publisher does not get the click. The traffic that was once the open web’s lifeblood is now leaking into the same walled garden that hosts the search engine.
Monetization collapsed for everyone except the very top. Programmatic advertising pays a fraction of what direct sponsorship paid. CPMs for mid-sized news sites are down 50 to 70 percent from their 2014 peak. The top 1 percent of sites (the ones with branded demand and direct sales teams) are doing fine. The long tail is not. The sites that used to pay for journalism are now running on small teams, partial paywalls, and donations, or they have folded.
What “open web nostalgia” gets wrong
The standard lament goes: we had it good, and we lost it to Big Tech. The diagnosis is not wrong. The implied prescription is, though.
The open web of 2008 was a great place if you were an early adopter with an ad-blocker and the time to find good sites. It was a much worse place for everyone else. The publishing tools were uneven. The comment sections were worse. The harassment was worse. The way it handled misinformation, child safety, and the long tail of bad-faith actors was worse. The reason people left for the walled gardens is that the walled gardens were, in many dimensions, better.
Nostalgia for the open web is, in part, nostalgia for being a person who knew how to use it. That is a real loss for the people it affected, but it is not the loss we usually talk about.
What is actually working
The honest version of “what now” includes the things that are working, not just the things that are dying.
Newsletters. The newsletter renaissance is real. Substack, Ghost, and Beehiiv host tens of thousands of paid publications, some of them doing serious money. The model works because the payment is direct, the relationship is owned by the writer, and the content is delivered to a destination the writer controls. It is not the open web in the strict sense. It is a return to the direct-publisher model that the open web displaced.
RSS is not dead. The number of people using RSS readers has grown modestly for the last three years, off a low base. NetNewsWire, Reeder, Feedly, and Inoreader all report growth. The growth is not enough to save the open web. It is enough to support the people who never left.
Niche communities. Discord servers, subreddit communities, Mastodon instances, group chats, small forums. These are not the open web either, but they are a form of public conversation that is not on the major social platforms. The moderation is uneven. The culture is often better than the major platforms at its best and worse at its worst.
Independent blogs and writers. The number of people writing long-form, ad-free or lightly-monetized, on a personal site is small but stable. The bar to start is lower than it has ever been. The bar to be discovered is higher than it has ever been. The economics are not great. The work is good.
The honest recommendation
If you care about the open web, the version that has the best chance of working is not the one we lost. It is a smaller, more deliberate thing.
Read newsletters. Subscribe via RSS. Pay for the ones you read. Use a reader. Avoid the walled gardens for news and analysis. When you do use them, treat the algorithm as a slot machine, not a recommendation system. Most of what it surfaces is not for you.
If you publish, treat the open web as a long-term play, not a growth strategy. Write for the readers you have. Make the work good enough to be linked to from the places the readers already are. Charge money if you can. Use the major platforms as distribution, not as the destination.
None of that is a solution. The economics of the open web are broken in ways that no individual choice fixes. The thing the individual can do is keep the practice of writing and reading on the open web alive, so that when the next platform shift comes, the skills and the institutions are still there.
That is not a victory. It is a maintenance task. But maintenance is what most of culture is, most of the time.