The responses to Matt Mullenweg’s recent post on how Tumblr was early to a lot of core social features of the web are intense. Lots of complaints about some horrific redesign that has supposedly happened. So I opened up Tumblr, as I do about once a month, and it looks the same?

If anything, Automattic hasn’t done enough. Tumblr is still impenetrable to folks who are not deep power users, and the power users are still convinced that all change is bad. Meanwhile, the platform has one of the worst targeted advertising systems I’ve ever seen (or simply the worst quality ads filling their slots I’ve ever seen) while at it’s core being an absolutely incredibly good social blogging system1.

I cannot tell what has changed there, and that’s bad. I have spent years wanting to use Tumblr without ever getting it to work. I’m their best shot at conversion, and since the Automattic purchase, nothing has happened that actually helps me become a regular.

Tumblr is all wasted potential. It should have been Instagram. It should have been Twitter. It should have been Mastodon. It has always had everything it needs technically, but has failed to put it together to win from a product perspective. Tumblr should be a case study for anyone in B2C product management. The only problem is, I’m not sure any of us would know how to fix it.

  1. Theming Tumblr remains the sore point here, with a horrible, outdated, incredibly hard to work with set up for amateurs. I cannot believe how little attention theming has gotten, considering that selling themes could be a huge business. It’s easy to say this from the outside— theming on a blog system like that, and making theming better, has the potential to break everything and might amount to a rewrite. But so what! It seems like that’s the rewrite that is sorely needed. ↩︎