from the article:

What startup lessons am I taking away from Greplin’s demise?

- Be a painkiller, not a vitamin (eg dropbox)

- Make something that people are going to use often (eg evernote)

- Hype is different from traction (eg greplin vs whatsapp)

- Give people a triggers to return to the service (eg icons, notifications)

- Make something viral or at least monetizing well enough to buy traffic (eg mint)

Gremlin is dead?!

Oh, I guess it is: http://allthingsd.com/20120618/greplin-recasts-itself-as-cue-a-personal-assistant-app/

Those are a great summary of the takeaways.

One more lesson: Don't think that just because a startup is from YCombinator and has Sequoia as a funder, that they actually know how to build a business.