No way -- plenty of flameouts were led by hackers, too. On top of that, the reason we've heard about the "most notorious" flameouts is because they actually knew something about marketing!

I'd bet that the vast majority of successful companies start with a problem, and then ask "how can technology help me solve this problem?" Sure, there are notable exceptions that worked backward, e.g. eBay, Amazon, Viaweb. They saw the inherent potential in a technology and asked "what problem can this technology help me solve?" And, no doubt, if you can pull that off, you're probably more likely to hit one out of the park rather than "merely" build a sustainable business. But most of the stuff that programmers build these days are solutions for non-existent problems.

I'd go so far as to say that, often, those who best understand the medium have the worst ideas of what to do with it. The more time they spend on the nuts and bolts of the technology, the less time they spend exploring other domains, where they might find more interesting and mainstream problems.

> On top of that, the reason we've heard about the "most notorious" flameouts is because they actually knew something about marketing!

Interesting point. I never thought of it that way. When a company is run by hackers, they usually only get publicity once they've created a product worthy of it. By that time, they're most of the way to success. Whereas when you have marketing guys at the top, the publicity comes when the company is less mature. So the perception that companies run by marketers flame out more often might be partially a result of selection bias.

"It makes sense that people who understand a medium will have the best ideas for what to do with it."

I think it makes sense to differentiate between ideas and style. For example, many of the most famous guitar players / trumpet players / musicians in general were self taught. I heard an interview with James Taylor where he said he doesn't even know how to read music. And Miles Davis only played like he did because he learned on a broken trumpet.

(This isn't meant to be a metaphor for hacking, but rather to point out that there is something called style that exists outside of the cognitive layer.)

It makes sense that the guys with the most interesting styles were the ones who were mostly self-taught.

At the cognitive level, it makes sense that the guys with the best ideas are those are

A) most familiar with the medium.

B) Familiar with other similar mediums they can draw from.

But a lot of people don't make their money through novel ideas so much as taking other people's ideas and adding the sex. Which seems to be more of a style thing than a cognitive thing.

I have an (utterly unsubstantiated) theory about this. I think that startups started by people with a business background have a BETTER chance of SOME SORT of success than startups started by hackers.

That being said, I think that businesses started by hackers have a greater chance for spectacular success (which is the empirical evidence that I think you're talking about-- and that kind of success is the only kind most VCs (or press outlets or bloggers) really care about or want to talk about.

In other words, business people are more equipped to solve the problems that assail most businesses-- and slog through to some sort of success/profit. Hackers, on the other hand, are more apt to ignore these problems or attempt do something revolutionary.

When biz people try to emulate hackers and aim for these revolutionary victories, that's where you often see the flameouts that you're talking about.

As PG has stated in his essays, it's much easier for a smart hacker to learn customer support, accounting systems, marketing/blogging, than it is for a non-technical person to pick up hacking in a few days.

Non-technical founders tend to overvalue:

- their lovely snowflake of an idea

- their business plan (that it will magically cause investors to beat a path to their door)

Things imho that can be hard for a hacker:

- sales

- finding and/or dealing with investors (getting into Y Combinator definitely seems to help here, but what if you don't make the cut?)

If I could find a non-tech guy who can successfully raise angel and/or venture capital money, I would totally partner with as a co-founder in a second. (given a good fit, etc)

These people tend to be hard to find though (the Steve Jobs' of the world) -- if they can raise money, why don't they just raise a round and hire a talented hacker at 1% equity instead of 30%? =)

There's a few counter examples.

digg.com would be the most obvious. Kevin Rose studied CS at college, but it had been a while and he didn't really have the skills to execute his vision quickly and simply. So he paid a developer he'd met through elance 200 buchs to write the first beta of digg. digg is now the most popular social news website.

Outside the world of startups, there's a bunch of established companies that weren't led by programmers.

Microsoft's success is largely based on business acumen - their OS monopoly came from Bill Gates' ability to purchase QDOS and sell it to IBM.

Steve Job's is also a non-programmer - the success of Apple is his vision and the ability to direct people to achieve that goal.

"Steve Job's is also a non-programmer - the success of Apple is his vision and the ability to direct people to achieve that goal."

Well, that and Wozniak's 1337 skills at hacking.

The real benefit of having a great hacker cofounder isn't the work they do themselves. It's that they can recognize and attract other great hackers to work for them, and that they're respected enough that those other hackers will give their best effort and not cynically slack off. You think that Jobs could have attracted Burrell Smith and Bud Tribble and Andy Hertzfeld alone if they weren't working with Woz? Or that he could've hired John Lasseter at Pixar without Ed Catmull and Alvy Ray?

Yes, I believe Jobs certainly would have attracted Andy Hertzfeld without the Woz. Both Jobs and Hertzfeld are passionate user interface designers, with a strong awareness of typography and user experience. The work Jobs have Hertzfeld was in an area Hertzfeld already cared deeply about.

I'm not familiar with the other developers you've mentioned.