5 MIN READ
Picture a small homepage from 2003. A hand written tagline, a sidebar of links, an RSS icon at the bottom. The page belonged to one person, and the person updated it when they had something to say. A reader knew the page because they had been going there for years, not because an algorithm had decided to surface it into a feed. That page is still on the web, in greater numbers than the standard story suggests. What has changed is the share of attention that finds it.
The default discovery path now runs through a handful of platforms, and the platforms optimise for time on site rather than for the reader finding the writer they were already looking for. A small slice of working internet is doing the curating that the personal sites used to do for each other, and the slice keeps getting smaller. Whether the open web is gone, or just smaller and harder to find, depends on what counts as the open web, and who is still doing the work to keep it alive.
What the platforms were
From 1998 to 2010, the major platforms behaved like tools. Blogger, WordPress.com, Movable Type, LiveJournal, Xanga. They gave the writer a place to publish, a way for readers to find the post, and a comments section for the discussion. Then they got out of the way. There was no engagement metric on the dashboard, no algorithmic feed in the front, no penalty for posting once a month or three times a day. The platform made money from a hosting fee or a subscription, not from how long the reader stayed on the page.
The publishing surface was simpler. The writer wrote the post, the writer hit publish, the post went to the RSS feed the reader had subscribed to, and the writer moved on. There was no testing the headline against four variants, no picking the optimal post time, no wondering whether the algorithm would surface the post tomorrow or bury it. The post was a small file on a server the writer paid for, and the reader chose to read it. That was the whole arrangement, and the arrangement was honest in a way the current one is not.
What the search engines were
Google in 2005 was a remarkably good product, and the reason it was good is that the business model was not yet driving the algorithm. PageRank was a real signal. The search results were the most relevant pages for the query, not the most relevant pages for the company the query was being optimised against. A user typed the thing they wanted, and they got the thing they wanted, in roughly the same order every other user with the same query got it.
The personalisation layer came later. The cross site tracking, the user profile for the advertising auction, the history based result adjustments, the AI overview box that answers the question without the click. None of that was there in 2005, and the absence is what made the search useful. Two users running the same query got the same ten links, and the ten links were the right ten links for the query. The SEO industry existed, but the SEO industry was still a footnote, not a parallel economy. The arms race between the search engine and the spammers was real, and the search engine was still winning.
What the personal sites were
The personal site is the part of the old web that survived the platform consolidation better than the eulogies suggest. The food blogger on a custom domain. The photographer with the portfolio and the contact form. The programmer with the technical writing. The hobbyist with the links page. They were everywhere in 2005, and they are still here in 2026. The share of total reader attention they capture is much smaller. The number of them, and the quality of the work on them, has held up.
Three structural advantages made the personal sites durable. They lived on a domain the writer controlled, so a platform could not change the terms under the writer or quietly remove the post for a moderation reason. They ran on RSS, so the subscriber list was portable and the writer was not at the mercy of an algorithm change. They linked to other personal sites, so the network effect worked for the small web, not just for the large platforms. None of those advantages is gone. They are just harder to find, because the default discovery path now runs through the platforms the small web was never designed to feed.
What changed
The platforms became publishers. The blog subscription, the link blog, the curated reading list, all of them were replaced by the platform feed. The writer went from publishing to a list of subscribers who had opted in, to publishing into a stream that the algorithm might or might not show to the followers the writer had spent years building. The personal site stopped being a destination. It became a place the reader visited if the platform pointed them there, which the platform increasingly had no reason to do.
The personal sites did not disappear. They are still being read by the people who care enough to look. The work of finding them shifted from the platform to the reader, and that work shows up as the price the reader pays for the convenience the platforms offer. A reader can still subscribe via RSS, still follow the personal sites, still curate a personal reading list. A reader willing to do the work can find the good stuff, and the good stuff is still there. The platforms still work for the writer as a distribution channel. The work the writer has stopped doing for free was the part the platform used to do, which was finding the reader. That trade is not obviously bad. It is just different, and the difference is what makes the open web feel smaller than it actually is.
None of this is a death sentence. Publishing a personal site in 2026 is roughly as cheap and roughly as easy as it was in 2003, and the publishing tools have improved in the directions that matter. What has changed is the default reader, who now arrives at the personal site by accident, if at all, and the platforms that hold the reader’s attention have no economic reason to point them at the small web. A reader who wants the open web to still be a thing in 2030 can keep it alive by doing the work the platforms stopped doing, which is finding the personal site, subscribing, reading on the place the writer actually publishes. The cost is small. The work is also the entire point.
The bottom line
Subscribe via RSS. Pay the writer directly when the option exists. Treat the platform feed as a slot machine, not a recommendation system. The open web is not a comeback story. It is a maintenance task, and the reader who does the maintenance is the one who keeps the personal sites alive.
Sources & Further Reading
All claims in this article are sourced from primary documentation, vendor advisories, and reputable security researchers.
Spotted an error? Email the editor. Corrections are issued with a visible correction note.
Editorial standards. Every article on humanrequired.org is reviewed by a human editor before publication. AI may assist with drafting or research; final editorial control is human. Read the full standards.
