2 ms·
This could have been a really smart way to get Medium started a few years ago. The way the site has been built out, though, this will be quite challenging. Rig
by GCA10 10y ago
This could have been a really smart way to get Medium started a few years ago. The way the site has been built out, though, this will be quite challenging.
Right now, Medium is a very noisy site. Noisy in a good way. Lots of energy. Lots of advocacy posts, full of swear words for emphasis. It's become a place where bold new voices periodically show up to make a statement. (Talia Jane, etc.) I try to skim it a couple times a week. Skim is the key word; the site's stories are very hit or miss. It's possible to breeze past five duds in a row so quickly that I hardly notice, and that's okay.
Nothing about current Medium feels like it has the makings of a judicious, curated concert hall of careful writing. Signing up a few famous writers won't fix this mismatch. Erratic is a big part of Medium's brand right now, and to forfeit that in a belated dash to "quality" feels dismissive of what the brand stands for.
It's really hard for me to see Medium as a place for the kind of WSJ/FT paid content that's aimed at rich high achievers who want business or lifestyle tips that the masses don't know about. It's even harder to see it succeeding as a paid online home for long, magazine-style features. Too many rivals in every genre are giving it away.
Building up a paid online audience is hard work, and it goes slowly at first. Jessica Lessin at The Information is doing it about as well as anyone, and she's defined her mission very tightly.
This latest strategy switch by Medium seems very prone to trying a lot of different formats in a hurry, without the patience to stick with one quietly for 2-3 years, waiting for the first little bits of traction to take hold.
- tima101 10y agofair observation overall, but not sure about "patience" - in 3 years, who knows how much of current VC funding remains.
- GCA10 10y agoGood point. In the do-over category, starting out with a small business that could recalibrate at $500k/iteration instead of $30m/iteration might have helped.