The turning point for my first startup came one rainy Friday afternoon, in February 2007.

As TechCrunch reported:

A new open source cross-site social networking service called Explode launched today and looks like a very appealing alternative to the now Yahoo! owned MyBlogLog. Built by UK open-source social network provider Curverider (whose primary product, Elgg, is similar to PeopleAggregator), Explode offers an embeddable widget that links out to users’ respective profile pages on any social network but allows commenting and befriending in one aggregated location. I found Explode via Steve O’Hear’s The Social Web, one of my new favorite blogs.

Steve O'Hear had written:

Although comparisons with Yahoo's MyBlogLog are inevitable, Explode isn't primarily intended as a way to turn an existing blog into a social network, nor as a way to monitor traffic. Instead, think of it as a loosely joined network that works on top of existing social networking sites and blogs, to allow a user of one site to 'befriend' and communicate with a user from another.

Curverider co-founder, Ben Werdmuller, says that Explode is very much a 'work in progress', and that they are already working on an API, and also have plans to add support for OpenID.

Dave Tosh and I had been working on Elgg for three or four years at that point, and had obtained a two-person office above a bookstore in Summertown, an affluent area in North Oxford, awash with coffee shops and restaurants. Our startup was fully bootstrapped; not a single penny of investment had been put in, except for our own work, and the non-monetary support of the people around us.

As Februaries in England can sometimes be, it was a cold, overcast day, with nothing to recommend it. Because we were bootstrapping, we had been working hard, as always, and we were exhausted. Also because we were bootstrapping, our waistlines had seen the effects of long hours in front of the computer sustained only by relatively cheap food, so we'd decided to buy ourselves gym memberships and try to go regularly. (I think we managed eight or nine times.)

As we worked out, we decided that, screw it, we weren't going to work on Elgg for the rest of the day. There was nothing to be gained, we were feeling kind of burned out, and surely there had to be something more fun we could do.

Elgg was a full-blown social networking engine, and although we later completely rewrote it, it was still a pretty powerful piece of software used by companies and institutions all over the world. And I had a simple idea.

When we got back, I got to work widgetizing all of Elgg's functionality: writing templates that would take pieces of its output - the friends list, for example - and embed it in JavaScript such that it could be made to display on any website. This was back when people still had websites where people could embed HTML and JavaScript, and the effect would be that any website could be part of a larger social network. That wasn't a small idea: it captured questions people were already having about siloed sites, and would later be tackled by projects like Google Friend Connect. (Of course, the IndieWeb movement is answering those questions today.)

Dave created a template for the site so people could sign up and get their widgets. At the last minute, we created a logo. I made it in about 5 minutes in MS Paint, which seemed appropriate for a hacky site, and we both agreed that a ridiculously bad logo was probably good: it captured the scrappy ethos of the whole afternoon project.

After four hours of work in total, we made it live, and pinged some people we knew, like Marshall Kirkpatrick, who at the time was a TechCrunch writer (now the CEO of the excellent Little Bird), and Steve O'Hear, who had previously written about us in the Guardian. For us, it was a bit of a lighthearted thing we'd put together at the end of the week, and we didn't really expect anyone to cover it. To his credit, Dave proactively got in touch with everyone anyway.

Boom. Steve wrote about us, and then Marshall wrote the story on TechCrunch. This was more coverage than the tech press had ever given us.

The fact that we had executed quickly on our idea, which in turn had developed out of having worked on social networking platforms for years, mattered. The fact that we didn't try and make it perfect, or spend ages putting it through careful design cycles or product feedback loops, was unimportant. We had a working first version, and we put it in front of people, who thought it was compelling. This directly led to many more customers for Curverider, as well as our first investors, who got in touch because they had seen the coverage.

Those four hours of work - and the quick decision to muck about, be creative, and follow our whims for an afternoon - literally changed the fate of our company, and both of our careers.