"make sure the architecture and code is examplary"

If you have time to write beautiful code at a startup, I might be suspicious. If something is ugly but not broken, why spend the time to fix it (versus releasing a new feature)? If you do everything right the first time, more power to you, but I think the prior suggests otherwise.

"Look for fundamental problems and propose fundamental solutions. Shoot as deep into the software stack as plausible/possible."

Not that I explicitly disagree, but the deeper you go, I suspect the higher your implementation risk.

The payoff for reducing code complexity, and thereby improving your intuition for the system, can be very high in the medium- and long-term.

I'd also add that if your project happens to be a "me-too" (which most projects probably are) then technical excellence might even be the deciding factor to establish and justify a presence in the marketplace.

I wonder why these advisory articles always assume that the common startup is built around some truly novel idea. In my expirience the opposite is true; 1% inspiration, 99% perspiration...

> technical excellence might even be the deciding factor to establish and justify a presence [for a "me-too" project] in the marketplace.

How often does technical excellence have a significant role in establishing presence in the marketplace? (I don't understand "justify". Justify to whom?)

I think you're thinking in terms of end-user products, like a website. In this case, technical excellence may be less relevant as long as it works. However, if you're selling technology, like in the example given at the end of article, technical excellence may carry a higher payoff. Another example: id software selling their engines to other game developers.

Justify to the users, give them a reason to use or switch to your product instead of staying with one of the competitors. (edit: a better word would maybe have been maintain)

I don't know how often it really plays a role but I think many google products fit this pattern. Search, Gmail, Docs and many others are not novel ideas - they were just better executed than what we had before.

One might even go on a limb and say that facebook is succeeding because it is a better myspace...

It's not an advisory article. It's just my approach. Somebody with a different background probably has a completely different approach.

Think id software.

Haha, top down approach that will fail as surely as I wipe my ass after going to the toilet. You think you have discovered something new and hidden and sweet - well, hundreds of people have found it before you, and they have silently failed. That's why you cannot study them.

There is only one way to get rich if you are not one of the elite - get off your high horse and get dirty. Start quick, test quickly, poll users and don't charge at first. Evolve, evolve, and evolve. Then get slimy. Studiously ignore techcrunch and focus on the smaller people. All these people who speak about 'buzz', they are part of a social network you are not a part of. They don't tell you this, so you do the same things and you wonder why yours does not work. Observe this eco-system, but do not ever believe you are a part of it, otherwise you will be blocked by their constraints but have none of their advantages.

There is no design for making money that works. There is no grand plan, because if there were, someone who have tried it and executed it successfully. There is just having a product, then modeling it in ways that sell it. Then reaching a huge amount of people. Use tricks like network effect, auto spreading, etc. Don't just drop it and expect anything to happen.

Most important rule: Don't let the constraints that tie the minds here tie you down.