Why I Don't Like the Lean Startup
collegeanswerz.com
collegeanswerz.com
The author fails to take into account a huge source of human error when it comes to what he's proposing: ego.
"Startups are about identifying what people want, and giving it to them. People tend to be pretty bad at doing this because they’re not honest with themselves about whether or not people actually want it."
People are even worse at being honest with themselves about what they're good at and what they know. The result of assuming you're wrong & testing all assumptions is discovery of the "unknown unknowns." Yes, it takes time, but there are loads of techniques/processes that have been developed in the past few years that minimize that time.
If your goal is failure, hand the reins over to ego. If your goal is success, question everything.
True. There's absolutely an advantage to testing. The question (as always) is whether the opportunity cost outweighs this advantage. I think that "question everything" is a huge overgeneralization. You don't have time for that. To be successful, you have to have the ability to put yourself inside the minds of your consumers, and know what they want (at least in the big picture sense. obviously feedback is useful for smaller things).
Imagine 3 people creating 3 different startups: one's humble, another's an ego-maniac, and the third's balanced between humility and ego. The humble one's probably going to go the route of "over-testing," to some degree (though you could argue that confirming an assumption is right is just as valuable as discovering it's wrong). The ego-driven person thinks they know what customers want, but is likely wrong about some things. For the sake of simplicity, let's say the humble one will succeed as a result & the ego-driven one will fail. Obviously, that's not realistic, but if you consider each person as representing the average success of all startups that operate within their range of the "humility spectrum," it sort of makes sense.
Now imagine that the person in the middle doesn't know whether they should trust themselves or to question everything and will operate based on what you tell them. If you give them permission to trust themselves, they'll go the way of the ego-maniac. If you tell them to question everything, they'll go the route of the humble one. If you tell them "test what you need to and trust yourself on the rest," they stay right where they are because they don't have any metrics for gauging themselves.
People will easily detect wasted time/effort spent on testing & adapt to avoid it. It's much harder to detect false understanding before wasting time on development.
And even if testing does lead you to better understanding of your consumers, that doesn't change the fact that it isn't always proper. It's a case by case basis. Sometimes you are confident that you know what they want, and testing would take time away from other things. You seem to be arguing that in the aggregate, it would be useful if people of certain "humilities" tested more. That's probably true, but my two points are: 1) they'd be better off solving the underlying problem of their abilities to know what people want. (this is a good starting point for learning - http://www.paulgraham.com/startupideas.html) 2) you shouldn't make decisions based on the aggregate population. you have more information, and you can decide whether or not testing is appropriate based on your personal situation.
Developing enough empathy for consumers is definitely necessary, but also really difficult to do. Design thinking is the best approach to solving that problem I've encountered so far, but testing is still a huge component.
This reminds me of an awesome series of posts on how to run a 5-day design sprint that came out of Google Venture's design team: http://www.designstaff.org/articles/product-design-sprint-20...
I guess we agree then.