I never been on such a website, but doesn't this make people angry? Or is the trick to do it also withe fake product and company name, so that you know it would work, but noone can hate you afterwards for lie to him?

Kickstarter seems like a better approach, since everyone KNOWS there is no product to buy.

I can't imagine being anything other than angry after filling out a payment form and then being told that there was no reason to have done that. On a new web application?

Although I consider the practice unethical, the other side would (does) argue: "Who cares? You are trying to get 20,000 customers and you pissed off 100 to get there."

I still can't bring myself to do it, though.

You wouldn't really have any customers if you weren't accepting payments though, it would be more like you pissed off all of your potential customers just for a bit of market data.

This is advice Tim Ferriss actually gives in his first book. I generally agree with your sentiment.

If I put my credit card number into a site and the next page said "actually this isn't real, and we promise we didn't save your card number", my next step would be to call my bank and cancel the card.

Surely the next step for a lot of us would be to devote an afternoon to bringing public misery upon the site operators. Don't take this advice, kids.

That's funny and true.

"Build a fake checkout"

Are you serious? That's a criminal act called fraud.

Please don't follow this advice. Just do something legal like having users sign up for a mailing list.

I'm genuinely curious - if no money ever exchanges hands, how can this be classified as fraud?

This will probably seem obvious afterwards...but if you sign a contract with someone to provide a good or service for $0 and then fail to do so, that could be considered fraud.

You could put some sneaky stuff in a ToS in theory

This is the worst advice I've ever seen on HN.

Much better to have PPC + landing page and a "tell me when this is available" signup.

Isn't this exactly what leads to the "a lot of people said they'd pay for it" false assumption? Not such a sound advice either...

The distinction is your landing page is only found by people looking for the solution you want to offer. They've already decided they have a problem, and need to fix it.

Contrast that with polling a random sample of your target demographic, where any number of things can spoil the results. People could say they like the idea just to be polite. People could have never thought about the problem you're solving before, but think it's neat. People can misjudge how much they care about it. People can have an enthusiastic initial reaction, but think the idea is dumb on further reflection.

It's the same sort of problem with polling for election results: you can't ask a whole bunch of random people who they're going to vote for and turn the raw results into a pie chart. To get useful data, you have to carefully predict which of your respondents will actually turn up for the election, and which ones aren't going to change their mind between the time you poll them and voting day. And even then, in the end, it's not the pre-election polls that determine the new leader, it's the election itself.

That distinction is what's being presented here; asking a bunch of people is only useful if you're very careful about filtering out the junk answers. You should instead measure people who have already decided, without your prompting, that they have a problem and have decided to seek out a solution for it.

I'm sure this works for things which you might buy having only see the website (and maybe a few screenshots) but beyond a certain level of cost/complexity you actually need to have some real software for them to try before they'll pay.