This is an interesting idea that has done the rounds a few times in the past few years, but it is also a bad idea in most cases.<p>First, if Google find out that you are advertising and collecting email addresses for a product that does not exist - they will ban you. PayPal will do the same if you are collecting any sort of payment information. There is just too much that can go wrong and it can be easily mixed up as being a scam or spam (which it sort of is, if you think about it).<p>You don't want to be banned from Google (or PayPal) - they carry that against your name forever, even if you later signup as a company or with a different account.<p>Second, this can only help you in a market where your potential customer knows what their problem is and knows to look for a solution. In so many cases of successful products this would not have worked. Not for Google, Twitter, Facebook, YouTube, Microsoft, iPhone, iPad or most any large and successful product.<p>It can also lead to false hope, or worse, an assumption that a product may not work. Twitter started out as a messaging system for cyclist couriers. If Jack had read Tim Ferris[1], took his idea to test the market with search engine ads, he may have come away unimpressed and believing there was no room for such a product - without even realizing the broader possibilities for such a messaging system.<p>It has become a bit cliche to quote Steve Jobs, but from his bio:<p>> At the end of the presentation someone asked wether he thought they should do some market research to see what customers wanted. "No," he replied, "because customers don't know what they want until we have shown them"<p>The better way to test the market is to actively market (as opposed to passively market - ie. having the market find you) - ie. go out and find your potential customers, the influencers in the market, analysts that cover the market, etc.<p>The whole idea behind minimal viable product is to lower the risk involved in testing a market - so get the MVP built and then promote it, but not just with search engine ads (the part of the market that knows what it wants, likely to be the smallest part) but also by actively reaching out.<p>The post suggests a budget of $20-50 a day just to get some leads, when with that sort of budget ($500-1000) you could just get a first version of the damn product built.<p>You can also find out how many people are searching by just using Trends or a similar product (keyword analyser). The PPC part just means you are paying $10 per email address - contacts that you may never end up using<p>I would like to see that similar posts to this one proposing this idea stress that results from search engine marketing are not the be-all and end-all. The language in this post suggests that the results of search engine marketing for a non-existant product conclude the viability of such a product:<p>> If you weren’t lucky enough to have a high conversion rate you can scrap the idea or rinse and repeat this process and tweak or pivot on the original idea.<p>This should say that your market may not be people searching for the product, not that you need to pivot the idea. This isn't how you validate startup ideas or products, it is how you validate part of a potential marketing strategy for a idea and product.<p>(Edit: and thinking about it, as a customer, I would <i>hate</i> this. If I were searching Google and comparing products that I need for a solution for and bumped into your product and gave my details, after your website suggested that the product exists - only to later find out that it * doesn't exist* - I would be very very pissed off about it. Especially if I needed a product right now and was falsely lead to believe that a solution exists).<p>[1] Tim Ferris is cited a lot as being one of the first to do SE testing (in '4 hour work week'), but I also remember Josh Kopelman writing a blog post about it a while before that as well. IIRC, Ferris suggests that you do this when you already have a product, to test potential new features, pricing ranges etc. The OP probably should have at least cited Ferris in his post, since this isn't exactly a groundbreaking idea and nor is it the first time it has appeared on HN.