1. Solve the smallest possible problem (that is still big enough to matter) for the user and know exactly what problem you're trying to solve. Google's first and primary job was very simple: Help people find stuff. They didn't start layering on everything else until much later.
2. Get a responsive and chatty audience using the product. The del.icio.us community eats new features. They pour over the service, discuss it, promote it, and complain when they don't like stuff. You couldn't have hired a better, more thorough, or more passionate group of alpha testers.
3. Don't wait until its perfect to put it out in the open. No more closed invite-only betas. Your idea of perfect may not jive with your users' ideas of perfect. Put whatever you can out there and get people using it as soon as possible. Feed them daily with new features to keep them interested and coming back. No one likes waiting six years for new releases.
4. Distribute. Don't force your users to play on your site in a walled garden. Let them take the service and use it wherever they want. (See Flickr badges, Google Ads, Amazon affiliates, Indeed jobrolls, del.icio.us linkrolls, moblogging, RSS, e-mail alerts, etc., etc....) Instead of building it so they will come, go out and get them by placing little bits of your service everywhere on the web.
5. Don't hold users against their will. If they want to leave, let them pick up with all of the content they created while they were on your site and leave... for free.
6. Be mindnumbingly simple. Extra clicks are deadly. People just won't do it. Indeed: One search, all jobs. Two boxes: What job and where. You can't get any easier than that and all it takes is for someone to put one search in for people to go, "Wait...what's this... links to Monster AND Careerbuilder??"
7. Get people hooked on free. Craigslist wouldn't have become Craigslist if it wasn't free for so much for so long. Even now, they're very profitable and they're only charging for just a few small pieces of their service in just a handful of their 120 markets. The world is changing. Service is cheaper to provide now than ever and users are expecting to get more for free than ever before.
8. Don't waste any money on marketing. Word of mouth has never ever been easier or less expensive in the history of human communication. they'll blog about it and e-mail everyone they know. And they'll tag it furiously on del.icio.us, too.
9. Don't overfund. Do you know how many times a day I see companies get funded on Private Equity Week and I'm like.
10. I hate it when someone says that a whole service sucks. Now, I say it myself, I'll admit, but what that does is it teaches you to discount and generalize, and probably miss a lot of small opportunities that add up. Now, I think Ofoto sucks versus Flickr, but people still use it. Why? There's got to be something there. AOL sucks... or does it? They still have 20 million users, so it can't entirely suck.
