Too often, startup companies fade away with little explanation as to what really went wrong. That's why tonight's discussion at Startpad.org was so intriguing. It featured two entrepreneurs -- ClayValet founder Mikhail Seregine and Faves.com co-founder Mohit Srivastava -- who described in detail the reasons why their companies didn't make it.

In many ways, the Seattle entrepreneurs were still grappling with the mistakes. "I did this for four years and a lot of the insights I only got in the last year," said Srivastava, who has since returned to Microsoft. Nonetheless, there were some wonderful gems of knowledge tucked in the talk, wisdom that every entrepreneur whose embarking on a new business should consider.

For example, both Srivastava and Seregine stressed the importance of the founding partner. In the case of Faves.com, Srivastava said that he and his co-founder shared too many of the same skills, which in the end caused the company to put too much weight on the importance of engineering versus other business tasks.

"We unknowingly overemphasized items that were comfortable to us," said Srivastava.

Seregine faced another challenge. As a young entrepreneur, he didn't have a co-founder to help him weather the ups and downs. "The human factors of running a startup are pretty stressful and it is good not to go through it yourself," he said.

But it wasn't just personnel. Both companies underestimated the difficulty in monetizing users.  Faves.com, which built (and continues to operate) a social network and bookmarking service, still to this day attracts about one million unique visitors per month. While that is a respectable number (and CPMs of $2 aren't too bad), Srivastava said the ad-supported business simply couldn't stay afloat given the company's general audience.

In retrospect, Srivastava said the company would have been better off focusing on specific niche audiences which were more attractive to advertisers. "It was hard for me.... In building this product, I wanted it to be for everybody. But the risk in making it for everybody is that you don't make it really good for anybody," he said.  

Along those same lines, Srivastava said that Faves.com never really had crystal clear messaging about what the service did. That's because it tried to do too much, operating a social network for friends to stay connected with one another and social bookmarking service where people could track information that was important to them. Services that were more tightly focused on one of those two areas -- such as Twitter (in the case of communication) and del.icio.us (bookmarking) -- fared much better, he said.

That lack of focus slowed down innovation, creating conflict between the various services.  "You have to narrow it down to the point where it is understandable and that it is something you can say you are the best at," he said.

Faves.com also missed opportunities to license the technology to media companies, some of which were intrigued with the service but not enough to fork over real money. "When push came to shove, they perceived us as a 'nice-to-have' and not a 'must have," he said. In order to get more engaged with those key customers, Srivastava said Faves.com would have been better off if they had an adviser or employee who had deep ties to the media companies they wanted as customers.

"Especially if you come from an engineering background like myself, do not underestimate the value of relationships," he said. "I know sometimes engineers have the perspective that the guys with relationships don't do any work, they just schmooze all the time. But if you look at a business being successful -- and you look at the value they bring -- don't underestimate it."

ClayValet, which was building an online shopping recommendation service before it shut down last October, also struggled with product focus and a coherent business model.

"We really didn't know how we would make money, but we were next to people who were making money," said Seregine, joking that it was the "trickle over theory." ClayValet bumped along trying to find revenue opportunities, experimenting with enterprise sales and interactive marketing of Web sites. It eventually tried to sell out to Microsoft, Amazon.com and a Seattle startup before pulling the plug.

The service failed to gain much traction, in part because it required consumers to wait 24 hours before getting product recommendations. Seregine said there wasn't much on the site for people to do as they waited to get products details back.

And he admitted that the company had yet another fatal flaw. "The name was horrible," he said.