[wdewind]

A lot of the "throw it up there and see what happens" movement comes from young engineers who literally lack the life experience to know what problems to solve. Given that scenario, and the intense pressure that comes from SV for young people to succeed, the approach makes sense, but there is a longer term, more considered approach to things. Especially for B2B startups, the best way to come up with an idea is to work at larger businesses and see what problems they are facing. Of course, if you are a young person who knows they want to start a startup, this seems like a roundabout approach, and one that requires a lot of faith. It's hard, and I see why people are attracted to "release, iterate, release," but I don't think it's the best approach.

This doesn't mean you don't iterate. A lot of the lean startup stuff is still useful to us (a/b testing for more focused UI etc.). But generally speaking, we know where the product is going, and our customers' needs, often better than them.

[wlievens]

Most people here mention "larger businesses". But aren't exactly those much more likely to buy their solutions from established software houses, for regulatory reasons, or higher guarantee of continuity? I don't see how a bank, pharma producer, shipping company or publishing house would buy software from an ISV...

[wdewind]

Just because you learn from certain businesses doesn't mean you need to start off selling to them. Many businesses start by meeting the needs of smaller businesses and work up to larger ones.

[NhanH]

PG is one of those who strongly advocates for "throw it up and see what happens". He is probably more influential to the movement than any young engineers, and he is decisively neither young nor inexperienced.

[jacquesm]

SV is but a very very small slice of the tech world and is not in any way representative, despite the fact that that is what you are hearing all the time. The linked article has a nice line about that.

What SV does have is a very large percentage of the successful internet start-ups. But there are so many more lines of business where untold billions are being made by companies that go for non-sexy non b2c approaches.

What works for YC and what works on the web is a very visible very thin slice of the software world as a whole. In a way you're making your life harder if you pursue that particular niche because even though that's where all the easy PR and the most visible success stories are it is also where the competition is fierce.

B2B is a lot less likely to make you billions. But it is more likely to make you tens of millions. So if you want to reduce the luck factor and increase your chances of success then b2b would seem to make good sense. It also works better if you will have to work without access to capital or a presence in a place like SV.

[YZF]

There's a big difference between an idea that makes it past many filters but you still don't know if it'll work and between a million random ideas that don't have a chance in hell in the first place.

You want to validate your business assumptions sooner rather than later and there's nothing wrong with that. You should also tweak your business based on the real world. Nothing wrong with that. Nothing really that new about that either. Other than that maybe Pivot is the new Agile?

[joeyo]

An VC only needs k of N start-ups to succeed, and it doesn't particularly matter which ones do. The high-variance approach ("throw it up and see what happens") may be optimal for the investor without being optimal for the founder.