HN works because PG is here. Reddit worked because PG was there to help it start.

Digg works because Kevin Rose is there.

StackOverflow works because Joel and Jeff are there.

Unless you've got a large audience to toss at the project on day one, expect a long, hard slog.

/. stumble upon livejournal etc etc

Fame makes it a lot easier, but you aren't doomed to failure if you aren't famous. If you have the dollars to do good marketing then I think it would be easy to make it work.

http://talk.collegeconfidential.com/

It isn't perfect, but it is chock full of information about the college admission process, and its online atmosphere of politeness and regard for other users' questions readily differentiates it from most other competitors running sites on the same subject. The site operator is pretty good at getting press notices and other free publicity, and quite good at search engine optimization, but mostly he's good at building a helpful and friendly online community.

Mountain biking has a solid fan base, worldwide, and the guys that are into it aren't afraid of spending several thousand on a bike, plus all the bits and pieces that go with it. They're also generally young, single and male, which dovetails nicely with an internet demographic. After all, you can't go out at night. That means there's ad dollar to be made. There's also a vast amount of technical know-how that goes into the sport, both from a how-to setup your bike, how-to do particular techniques, and where to go. A lot of this is currently buried in forums with crappy interfaces and broken searches.

A niche area like this isn't going to get you bought for 10 million, but a couple of talented guys could make a nice living doing something they love, which isn't to be sneezed at.

There are so many other niches that spring to mind crying out for a better problem-solving model than forum postings, which is really what SO is. It's true that Yahoo answers could enter any of these markets, but with branding and endorsements, a dedicated site would outstrip Yahoo answers without much trouble.

I expect that if you're creative, dedicated, and put in the time to tackle the "long, hard slog" you'll have something of considerable value by the end of it. In fact, if you can get the momentum necessary, you'll have precluded others from even entering the field.

But you're definitely corect that the right persona and pre-existing followings basically super-jumpstart these types of sites.

I would say no.

What you need are credentials - doesn't necessarily need to be fame. Admittedly -- It's probably much harder, but you could demonstrate credentials through some other means.

It would probably have higher traction among, say, graphical designers than medical staff.

For law and medicine, for instance, you'd have to fight the mindset of "online is not reliable". In programming it is fairly easy to check if someone's answer is trustworthy because you can go and test it yourself relatively quickly and unambiguously. In other fields establishing the credibility of someone answering a question is a lot more tricky. That said, I think bar associations have (bad) sites to ask each other questions.

I agree with Jeff that you can't make a general site of this sort without an intense amount of moderation. (he said so in a recent SO podcast)

In short, it all comes down to the community. And it is probably hard to figure out what incentives and moderation to have for a given community if you are not a part of it yourself.

Programming gets you a much higher volume of traffic. The same programmer, might come to your site 100 times during the week to find solutions to the different problems. Its also a field where you have a lot of amateurs and students looking for solutions.

The other fields you mention under specialty, are all the "oh shit my car broke down what could it be?" type. Its a one time problem, after which the person won't come back for 2 years. So you'll never see the same growth as SO, because you don't have any repeat users.

Why? Have a look at reference sites online. There were plenty of sites like experts-exchange before StackOverflow, and if you trawl Google Groups, you'll see a similar thing. Plenty of people asking technical questions, not so much anything else.

Certainly, you don't see medical students out there asking the best way to diagnose a rare disease.

There is one exception though -- Travel sites. Travel questions and forums are pretty big, and I think a Q&A site for that could be enormous.

Meanwhile Yahoo Answers, AskMeFi, etc continue to fill the gaps for other niches that might actually work.

In fact, I believe it so much, that's my startup: http://www.chuwe.com - "Stackoverflow for startups and small businesses."

As brandnewlow pointed out, starting it is a massive uphill battle - chicken and egg problem as it were. The software's only small part. We're attacking it in what we hope is a novel way, but you would have to find an approach unique to the area you were targeting, I imagine.

If you're interested, I've been planning on open-sourcing chuwe in a bit after we make it a bit more robust. Contact me and we can work out something if you'd like to just take the software and modify it to your own niche!

Something else to consider are the 'game' factors. StackOverflow implemented badges and they work well initially (and teach the user how to use the site). At the end of the day the things that tie you into a site are the things that make you want to keep going at it. On most discussion boards it's reputation/karma but little things like badges and medals help too.

Obviously it depends on whether the field needs/can have a question and answer platform, but I can see this discussion model working for most technical areas of knowledge.

as a business? hard to tell.

Yahoo Answers holds the potential to encompass every use case for StackOverflow and all your offshoots. However, I sincerely believe that the added context that a site like StackOverflow provides (being that it's curated specifically toward programmers) does provide real value. That sort of context injection could certainly be carried over to other verticals (provided the sites can attract seed contributors and curators).

The trick is finding a niche that you're interested in, can serve well with the site's format, and has money.

If you find that niche and you approach it in a novel way, then you can (likely) serve it infinitely better than a general questions and answers site like mahalo/yahoo answers.

The reason StackOverflow works is that computers work based on their code, and so a certain compiler will interpret C# code in a certain predictable way, so the people answering the questions don't have to be anywhere near you. And they are able to answer in a relevant way. The same is not true for other fields.

I just imagined a SO for the medicine field, I'll read it everyday!, imagine a doctor asking for help to another doctor?, and not only that, a doctor asking for feedback on a research... wow