No Ads, No Analytics, No Thanks: Inside the Hand-Built Directories Keeping the Web Human

There's a page on the open web that is nothing but a list. No infinite scroll, no recommendations, no little avatar asking if you want to upgrade. It's a plain HTML index of a few hundred links — personal blogs, a site about vintage synthesizers, somebody's shrine to roadside diners in Nevada — each with a one-line description written by a person who actually visited it.

The person who maintains it doesn't want you to know who they are. They've never run an ad. They've never checked their traffic stats. When I asked how many people visit, the answer was a shrug translated into text: "Some. Enough."

Discovery Used to Be a Hobby, Not a Product

It's easy to forget that there was a time when finding things online was a form of labor people did voluntarily. The web of the late nineties and early two-thousands ran on enthusiasts. If you wanted to know where the good Star Trek fanfiction lived, you found the fan who'd catalogued it. If you were into obscure Japanese RPGs, somebody's GeoCities shrine was your map.

Webrings formalized this. A ring was exactly what it sounds like: a chain of sites, each linking to the next, so you could travel the whole loop and end up where you started. No company owned it. Nobody optimized it. It worked because someone cared enough to keep the chain intact.

Most of that architecture got steamrolled by search and then by feeds. But it didn't die. It went quiet.

The People Who Never Left

Spend enough time poking through the small web — the personal sites, the tildeverse servers, the directories that link only to other directories — and you start noticing a pattern. The people maintaining these things talk about curation the way other people talk about gardening.

One directory runner I corresponded with, who asked to be identified only by their handle, described their routine: a couple of evenings a week, they visit every site on a submission queue, check that the links still work, and decide whether the site belongs. Rejection is the norm. "If it's a landing page for something being sold, it's out. If it's a real person making a real thing, it's in."

That's the entire editorial policy. Compare that to the discovery products we use daily, where the criteria is opaque, adversarial, and constantly shifting under our feet.

Who Actually Uses These Things

Here's the part that surprised me: the audience isn't nostalgia tourists. Or rather, it isn't only nostalgia tourists.

There are teenagers who found the small web through Neocities and treat hand-curated directories the way earlier generations treated the library card catalog — as the only trustworthy map. There are researchers and archivists. There are people who've quietly opted out of algorithmic feeds for mental health reasons and still want to browse. And there are the site owners themselves, for whom a single link from a respected directory can send a slow, steady trickle of visitors for years — visitors who actually read, instead of bouncing in eight seconds.

One personal-site owner put it to me plainly: "A hundred people from a hand-built list is worth more to me than fifty thousand from a feed. Those hundred people chose to be here."

What Their Labor Reveals

The uncomfortable takeaway is what these directories expose about the mainstream web. Discovery didn't get better when it became a product. It got faster, and it got monetized, and somewhere in that trade we lost the annotation — the human note in the margin that says this one's actually good.

An algorithm can surface a billion options and still can't tell you which one was made with love. A stranger with a text file and good taste can.

There's also something quietly radical about the economics. No analytics means no optimization pressure. No follower counts means no performance. The lighthouse keepers of the small web aren't building audiences. They're building infrastructure for people they'll never meet, on the theory that the web is a place, and places need maintenance.

Keepers, Not Curators

I keep coming back to that word: keeper. Curators select. Keepers tend. The distinction matters. A curator works for an institution or a brand; a keeper works for the thing itself.

The directories will never trend. Most will outlive the platforms that buried them, though, precisely because they demand so little — a text editor, a hosting bill of a few dollars a year, and a stubborn refusal to stop. The web that optimized discovery into a product assumed nobody would do it for free.

Nobody told the keepers. They're still doing it anyway, one link at a time, for an audience of strangers who mostly never say thank you.

Which, if you think about it, might be the most honest version of the internet we've ever had.