I'm admittedly short on my knowledge of this newfangled "lean startup" notion. But, it sounds to me like the fundamental principle is "lazy product evaluation." Ie, instead of building anything up-front, go out, find out what people want, and build that, MVP, reacting to markets that you know, provably, to already exist.
At the risk of getting told I don't "get it", this sounds like a great way to build a very successful but horribly boring startup.
Vision is a critical part of a leanstartup and by no means is the fundamental principle "lazy product evaluation".
To quote an Eric Ries post:
"Lean Startups are driven by a compelling vision, and they are rigorous about testing each element of this vision against reality."
The two most important leanstartup principles or teachings for me are
a) I'm stupid and make really poor assumptions all the time. To combat that, I need to continually be validating my ideas against real customer feedback.
b) Validated learning is the most important metric for progress. Ultimately my company be successful if I understand better than anyone else the lives of my customers as that in-depth understanding of their hopes and dreams and pain will enable to build and market products which speak to those emotions.
Your last paragraph has a lot of words but says very little. Can it not all be summed up as "talk to your customers?" Isn't this the foundation of any business? Does anyone actually advocate not doing so?
Obviously you need to be aware of the problems your customers face as a business.
But the notion that anything is the "most important" metric of progress reeks of classic silver-bulletism that plagues software engineering.
I also am not a big believer in "validation." You're never going to "validate" your ideas through customers. You're going to evolve your ideas through seeing people use what you build. But they're never "validated" or "proven right" through intimate customer feedback. Good ideas are "validated" over the course of years, often after much ridicule, resistance, and pain.
"Can it not all be summed up as 'talk to your customers?' Isn't this the foundation of any business? Does anyone actually advocate not doing so?"
I don't think anyone advocates not talking to your customers, but Steve Blank makes a bigger point than that.
The point isn't just to talk to your customers. It's that the CEO, the founders, the people with the control have to talk to the customers.
A lot of companies have the CEO, the programming team, and the rest of the decision makers relatively isolated from the customers, whereas the sales team is the one out there talking to customers every day.
This isolation still makes it look like the company is talking to customers (they have sales out there, every day, listening to customer feedback). But because the real decision makers aren't themselves experiencing the way the customers use their product, they'll never understand it the way they need to to keep making their product better.
'Talk to you customers' is a bi ambiguoud, and could be wrong. My reading of the lean startup thing is that you look at the numbers and metrics of what your customers are doing, rather than ask them what they want. If you're not sure what feature to add, don't ask people, and do surveys, do a half version of all the possibilities, and see what features your customers are using.
Well, 'talk to your customers' is a little different from 'listen to your customers and make sure you understand exactly what problem your product/service solves for them'. And I think the only difference between evolution (as you've described it) and validation is that seeing how people use what you build, and getting qualitative data on those users is emphasised, rather than something that just happens over time.
At a minor level, that could be making sure that the next feature you implement is the one that will make your software much more usable from their point of view, rather than the shiniest or most technically interesting to implement. (e.g. it doesn't matter if your mousetrap automatically telling your computer when it's caught something, if their local mice just aren't interested in $bait_that_works_best_in_bobs_house )
On the larger scale, if your customers are using your product to solve a problem you didn't expect, you can't optimise your product so it sells even better. If the largest user of your mousetrap is the gremlin-catcher's guild, then perhaps you need to look at building more gremlin-specific features, and changing your marketing a little.
And, of course, the simplest metric for validation is that if they ain't buying it then it ain't doing what they want enough for them to pay for it.
Disclaimer: I have never run a startup, let alone used these principles. They just seem to make sense to me.
That's a good question, and I'm sure people go either way on this. I for one am definitely looking for success, but I try to find it by doing things I like and care about.
I'm comforted by Paul Graham's thoughts on this, namely that if you build something people like, you'll usually manage to find a profit somewhere.
At the risk of getting told I don't "get it", this sounds like a great way to build a very successful but horribly boring startup.