Why this exists: a look across the other platforms

An independent developer publishes a post. Within a week, four other blogs link to it: one disagreeing, one quoting a paragraph, two filing it in a Friday roundup. Somebody submits it to Hacker News, where it does modestly and falls off the front page by lunch.

Six months later, which of those facts can still be looked up? The Hacker News submission, with its score. The four links, which are the more interesting fact, are recorded nowhere at all.

That is the whole reason this site exists. But "nobody does this" is a claim that deserves the list it is drawn from, so here is the list.

The aggregators

Hacker News, Reddit, Lobsters. They rank submissions, which is a different object from a post: a post exists on them only if someone submitted it, within the hours when submitting works, and the record they keep is of votes rather than of citations. A post nobody submitted did not happen. A post that four other writers thought about for a week and then linked to did not happen either, unless one of them also filled in a form.

The curated lists

Kagi Small Web, ooh.directory, Marginalia's export. These are genuinely good, and they are lists of names: a set of blogs worth reading, maintained by people with taste. What a list does not hold is edges. It can tell you a blog exists; it cannot tell you that eleven other blogs pointed at it last month.

We can put a number on how much overlap there is between a good list and a corpus assembled by fetching: of the 40,477 distinct hosts in Kagi Small Web, we already held 40,215 — a little over 99%. The list was not where the missing blogs were. The missing blogs were in the links our existing blogs had already written down.

The readers

Feedly, NetNewsWire, your own reader. A reader is vertical: it holds your subscriptions, delivers what they publish, and knows nothing about the blogs you did not subscribe to. That is not a flaw, it is the job. But it means the aggregate view — what is spreading across a few thousand blogs at once — is structurally invisible from inside any one of them.

The editorial aggregator

Techmeme, which is very good at what it does, keeps a rolling leaderboard of who is breaking and who is being linked. It is also about the technology press: outlets, reporters, and the news cycle. A person writing carefully once a month about running Postgres in production is not in that universe and is not meant to be.

The adoption indices

RedMonk's language rankings, the ThoughtWorks Radar, the big annual developer surveys. These are the closest thing the industry has to a measurement of what developers actually use, and their inputs are thinning. RedMonk's own January 2026 write-up openly asks whether Stack Overflow should remain one of its two axes, with no replacement public dataset named. The Radar's FAQ says plainly that it captures Thoughtworkers' experience on client work and makes no attempt to survey the market.

Neither is doing anything wrong. But there is an obvious vacancy for a public, reproducible dataset about what practitioners are writing about, and nobody has filled it since Technorati — which did index blog-to-blog links, and stopped in 2014.

The AI visibility tools

Otterly, Peec, Profound and the rest of a category that did not exist three years ago. They prompt the answer engines and record what comes back: whether your product is named, and beside whom. Real product, real budgets.

What none of them can do is check the answer against anything. When an engine says a tool is the standard choice for something, the useful question is whether the people who actually write about the work agree — and to ask that you need to hold what those people wrote. The visibility tools measure the answer. Nobody holds the ground.

Community intelligence

Common Room and its neighbours watch the places with APIs: GitHub, Slack, Discord, forums, the social networks. Somebody writing on their own domain is precisely the case that has no API, no platform and no event stream. It is a feed, and feeds have been unfashionable for long enough that the tooling around them is mostly abandonware.

The backlink tools

Ahrefs, Semrush, Search Console. Search Console shows you your own links, sampled, for domains you have verified. The commercial crawlers index the whole web, which sounds like more but is less: a blog is a grain of sand in a corpus built for SEO, and a link index built for SEO has no way to tell a person describing what they migrated to from a footer, a blogroll or a paid placement.

What is actually missing

Look at the list and a pattern shows up. Every one of these is a platform keeping the record of what happens on it, or a list of names with no edges, or a measurement of somebody else's answer. The writing that happens on people's own domains — outside any platform, which is the entire point of it — has no record at all, and the links between those domains are the only social graph that independent writing has ever had.

So we built the index for it. Today it holds about 51,000 blogs, 46,000 of them with a live feed, 1.4 million posts, and 3.4 million outbound links, two and a half million of which resolve to another blog in the corpus. That graph is what the most cited pages, the record page behind every post and every blog, and what is being talked about are computed from.

Three things follow from the list above, and they are the promises worth writing down.

Every figure counts distinct blogs, never posts. A citation index that ranks by volume returns whoever publishes most, forever. Counting each blog once per figure asks how many different people found something worth pointing at, which is the question a link was always meant to answer.

The method is public. A number here is checkable rather than trusted. That is the difference between an index and a score.

We do not republish anyone's text. The pages here are titles, links, dates and counts. To read a post you go to the blog that wrote it — which is also the only outcome that is any use to the person who wrote it.

The writers are not the customers. The record of who cites you is free, and stays free; claiming your blog is proof of control, never payment. The companies being written about are the ones who will pay for the same graph pointed at them.