Why a curated list still travels when almost nothing else does

A single-page site listing tweets that made money was shared by people who share almost nothing. "Twenty accounts worth following for X" is the one format that reliably travels on a platform that throttles almost every other kind of link. Lists of tools, lists of people, lists of posts — the format is older than the feed and outlives every algorithm change.

It is worth asking why, because the answer says something about what people actually want from a share.

A list is a gift that costs the sender nothing

Most shares are a statement about the sender: this is what I think, this is what I made, this is who I am. A curated list is different. It is useful to the recipient — here are the twenty people you should follow — and it costs the sender nothing to give. That combination is rare. The sender looks generous and well-read; the recipient gets something they can use in the next minute; nobody had to write anything.

The lists that travel furthest exploit this exactly. They are specific enough that the recipient knows immediately whether it is for them ("for people building in food tech", not "for founders"), and useful enough that forwarding it is a favour.

It is checkable

A claim travels on trust. A list travels on verification. Every name on it is a link the reader can open, and within thirty seconds they know whether the curator has taste. That is why bad lists die instantly and good ones compound: a reader who opens three names and finds three good accounts trusts the other seventeen, and sends the whole thing on with their own credibility attached.

This is the design principle behind the creator directory on this site. Every name on it is there because of posts you can open, ranked by what those posts did, with the pattern named. It is a list whose evidence is one click away, which is the only kind worth making.

It has an argument, or it is a phone book

The lists that go nowhere are the ones with no point of view. "100 marketing accounts" is a phone book. "The 20 accounts whose launch posts I would actually study" is a list, because it takes a position — these and not those — and the position is what the reader is really sharing.

The position can be narrow. It should be. "Creators worth following for AI" is a better list than "creators worth following", and "the people writing best about AI tooling" is better again. Narrow lists find their audience; wide ones find nobody's.

How to make one that travels

1. Pick a subject you can defend. You should be able to say why each entry is on the list and why three obvious names are not.
2. Make every entry checkable. A link, and one line about what the reader will find. The line is the curation; the link is the proof.
3. Order it. An unordered list is a shrug. Put the best first and say why, even if only in a word.
4. Keep it short enough to finish. Twenty is the natural ceiling. Past that, the reader stops trusting that you read them all.
5. Say what it is for. The first line of the list is a first line like any other: it has to work in a feed. "Twenty accounts for people who write about X and are tired of the same five names" does more than "a list of accounts".

Lists of posts, not just people

The same logic applies one level down. "The twenty-five most-shared posts about food tech" is a list of posts — checkable, ordered, narrow — and it travels for the same reasons a list of people does. The difference is that a list of posts teaches something on its own: read twenty-five together and you see what they share, which no single one of them could show you.

The topic articles on this site are exactly that: the most-shared posts on each subject, ordered, with what the analysis says they have in common. They are built to be sent — and, because every entry links to a post page with its mechanics, to be learned from once they arrive.

A good list is the smallest useful thing you can give a stranger. That is why it still moves when nothing else does.