If 37 Signals Had Designed YouTube or Flickr They Would Both Be Dead By Now

If you ever checked out Flickr in the early days you would have encountered something very different to today's Flickr. The web-based photo sharing application actually started out as a kind of chat room on steroids. The big idea was real time photo sharing -- you could upload photos to something called a "shoebox" and then drag them directly into live conversation windows. This was all executed in Flash.

YouTube started out as a video version of the web social phenomenon HOTorNOT. Founder Jawed Karim explains:

"I was incredibly impressed with HOTorNOT, because it was the first time that someone had designed a website where anyone could upload content that everyone else could view. That was a new concept because up until that point, it was always the people who owned the website who would provide the content."

Of course both these applications are now a world apart from where they started. Flickr have eschewed their Flash roots and moved entirely into HTML/AJAX for interaction. The real time elements are long gone. Similarly, YouTube shows no sign of it's HOTorNOT past and has become the defacto destination for sharing any kind of video online.

The key to their success was a combination of two things: 1. having enough flexibility to allow emergent behaviour from users (in YouTube's case "as the site went live in the spring of 2005, the founders realized that people were posting whatever videos they wanted.") and 2. accepting and adapting to this new use of their tool. Flickr realised people really just wanted a better way to share photos on the web with their family and friends. YouTube saw people wanted a way to share any kind of video easily without the encoding/hosting headaches of Real Player or Windows Media Player.

But imagine if 'original' YouTube had been designed by 37 Signals. It would have been beautifully designed for sure, emcompassing only the functionality required to allow the posting/browsing of HOTorNOT style videos. It wouldn't have had the wiggle room to allow for a different use. Likewise with Flickr. So sometimes it's perhaps better (maybe even 1.65 billion times better) to have a bit of looseness in your applications. When things are so tightly defined and resolved to making a single goal you do lose the flexibility for users to define for themselves what your product is really good for.