It's an interesting story that other entrepreneurs can learn from and Abbott has done a real service in opening up honestly about what's going on at Dandelife.

The basic idea behind Dandelife is that it's a place to record your life story. You can read and comment on other peoples' stories, navigable by time-line, topic or story teller. A lot of people seem to be using the service as a time-line based social bookmarking account for things they find around the web. The site spoke to a universal human need, the need to be heard. Abbott says all that content did well in search engines, too.

When the company launched in the summer of 2006, it had a knockout team of advisors including the grand poo-bah of web design Jeffrey Zeldman, Userplane founder Mike Jones (who sold his company to AOL for $30-$40 million the next month), Bruce Livingstone (who had just sold iStockPhoto for $50 million months earlier) and a list of other luminaries. It was a great team.

Dandelife won a Webby Award in 2007. It made early moves in favor of data portability, tying to the once-celebrated Attention Trust and later implementing Attention Profile Markup Language (APML), the protocol developed by Data Portability Working Group co-founder Chris Saad.

The company got lots of attention. It got written up on Lifehacker, shared on Waxy Links and hit the front page of Delicious Popular.

What Went Wrong?

What went wrong? Apparently all those great backers didn't give Dandelife very much money. At least that's what's implied by founder Abbott last night when he said the site was terribly engineered because he was unable to afford to hire better developers. What's most important, though, is that Abbott says the site suffered down time and took up all his energy in fixing the back end. He's a marketer by trade, but now admits that he's spent almost no time marketing Dandelife. Feature-creep took over instead, he says.

No matter how hard I wanted the site to be successful, and no matter how good the pitch for Dandelife was, I always feared success. Staying small and non-profitable became an excuse for failure to scale. Why won't this site grow? I kept asking myself. But in my heart of hearts, I knew why. I was pouring all of my effort into product development and in particular finding and fixing bugs, that I had no time and no confidence in marketing the site.

Now Abbott says he's going to scale back on features, throw far more time into marketing the company and, most importantly, open source the code. "I think the Internet could use a thousand Dandelifes," Abbott wrote last night. He hopes that some of those new versions of the site will pay him to consult on implementation. That could work well.

Abbot's initial monetization strategy was to leverage the life stories of users (with their permission) in service of brand advertisers - to act as an advertorial farm-team of sorts. I reviewed the service on TechCrunch when it launched and said I thought that was creepy. The company even tried to trademark the term "lifecasting", and any time spent on that was time wasted, obnoxiously even.

It appears now that Dandelife never scaled up enough to be able to pull that branding plan off. Abbot said last night that it's a great site for "people who want to make a difference in the world." Presumably that's a different group than those who want to share stories about drinking Pepsi at a family picnic or looking great in brand name clothes when they went on a memorable first date.

People who like Dandelife like it a lot. There is probably a lot of room for long form personal story telling on the web. The lesson Abbott seems to offer, though, is that launching such a company, even with an all-star advisory crew, a lot of press and solid search engine pull, just isn't enough. Prioritizing quality engineering from the start and remembering to do marketing after the launch are at least as important.