[ilghiro]

I'm not particularly fan of this or similar practices because A) it's a bit morally grey to say the least, and B) you're not actually testing a business idea, you're simply testing how a product sounds.

A half decent example is Evernote (recent financial troubles aside). Imagine trying to cook up a super quick landing page for Evernote and then buying some adwords. I'm pretty sure you wouldn't get very far because who wants "Notetaking App #784" (which is about all the information you can give in such circumstances)?

Based on that you'd assume no one wants to pay for/download a note-taking app. Of course this is provably false as many people pay for myriad note-taking apps. Where's the discrepancy?

When you land on a products landing page there's far more than just product screenshots. There's also the impression created by the brand, one you've probably heard of before you got there. There's customer reviews, there's well thought out pricing designed to make you pay exactly what you can afford, there's a thousand techniques used to make sure they get the sale.

Effectively, I'd say it's very difficult to distill all this down into a quick experiment that proves anything.