05 Jan 2016

8 years ago, Superfeedr

8 years ago, I started building what is now Superfeedr. It was the early days of Twitter and Facebook, blogs were still in their early adopter’s phase and the big chat networks were AIM, Y! Messenger. Google Talk was gaining in popularity but was still dwarfed by the AOL and Yahoo! giants.

At the time, I wanted to build some sort of universal bot. A tool which anyone could use via email, SMS, chat messages or even social networks (Direct Messages) to subscribe to their favorites sites and blogs. I named it Notifixious.

I was looking for a service to help ingest all this data. It quickly became obvious that none existed, but that RSS was a great starting point. Actually, notifixious’s API became exactly just that… and very soon, we were seeing more usage through the API than through any of the “consumer” channels. Superfeedr was born.

But history is not the point of this note.

To me, at the time, it was obvious that the browser could not be the only entry point to the web, that chat applications, even though they required another software to be installed were filling a gap in the user’s needs and that eventually, we would also interact with machines via “chat” programs.

A lot of people warned me: HTTP had already won the protocol wars: IRC, XMPP, NTP and even SMS would, at best, be re-invented to sit on HTTP. I was also dumb enough to assume that the lack of users was not the consequence of a half-baked product but a mismatch with user needs.

Today, it’s obvious that my initial assumption that text would be a major user-interface was the good one. Slack, Snapchat, Whatsapp, Telegram, Kik, Intercom are proving that, more than ever, there’s a market for chat applications. Even further, I can see Facebook’s M become a much more powerful version of Siri.

In retrospect, it’s easy to predict the future but it’s hard to predict when it will happen.