Sean Johnson over at Snooty Monkey has passed on some pretty bad advice on how to find prob­lems with your startup idea. The gist of it is you go to a cof­fee shop and lie to peo­ple about “you’re brother’s crazy idea” and have them tell you what’s wrong with it.

The most likely out­come is you’ll hear mostly the same obvi­ous rejec­tions of your idea that you your­self have and have come to believe are sur­mount­able. Your $20 didn’t gen­er­ate any great new insight, but was cheap check that you aren’t blind to an obvi­ous shortcoming.

A good out­come, is that you hear lots of inter­est­ing and sound new objec­tions that you never thought of before. This should give you real pause about your idea. Both its mer­its, if the objec­tions are good ones, and the extent that you’ve suf­fi­ciently thought through your idea and are being real­is­tic about it.

The argu­ment for the “$20 Starbucks Test” is that peo­ple will not be hon­est with you if they know they’re crit­i­ciz­ing your work, so you lie to them, buy them a cof­fee and get some hon­est feed­back. Now this was shared with Sean by Hugh Crean, the ex-CEO of Farecast (a very very good idea) so it has some appeal-to-authority cred­i­bil­ity behind it. But I have to say, this sounds goofy to me.

  1. If you spend your time build­ing some­thing instead of con­ning in cof­fee shops, you can put a min­i­mum viable prod­uct in front of some­one and they’ll tell you very quickly what’s wrong with it/whether they like it or not.
  2. If no one in your social or busi­ness cir­cles will be hon­est with you on a pro­fes­sional level, you may have big­ger prob­lems than whether or not your startup idea is good or bad.

How to test your startup idea…

I’m not one to offer empty crit­i­cism. (At least at this very moment; the real­ity of that state­ment waxes and wanes.) So I offer three more sin­cere solu­tions with accom­pa­ny­ing tes­ti­mony to their worthiness.

Option A. Mechanical Turk sur­veys from Amazon:

The infor­ma­tion I got back for my $27.50 was INVALUABLE. I found from that 1 sur­vey, how to basi­cally build my prod­uct for launch. What fea­tures I had to have based on how users would use the ser­vice. I also real­ized I could basi­cally cut my cur­rent fea­ture set in 1/2 because what I thought peo­ple would want, wasn’t even mentioned.

From How I Used Mechanical Turk to Validate my Startup Idea by Lindsey Harper

Option B. Run Google Adwords against a com­ing soon page:

Instead of doing for­mal focus groups or ini­tial lengthy pri­mary research inter­views, we’ll start with a more infor­mal process that will deliver results quickly and inex­pen­sively. This is not per­fect, but a great place to start in the de-risking process. It’s quick, easy, and cap­i­tal efficient.

From Five Surprising Ways To De-Risk Your New Idea Using Google Adwords by Steve Barsh

Option C. Ask peo­ple to com­mit to pay for it:

I didn’t start WPEngine until I had 30 peo­ple say­ing they will give me $49/mo. Not would, but will. In the end, 20 actu­ally did. Our mes­sage res­onates with hun­dreds of thou­sands of peo­ple, and new peo­ple sign up every day though we’ve still spent $0 on gen­eral marketing.

From Vetting a startup (or two): The sys­tem­atic birth of @WPEngine by Jason Cohen

This last one is the best in my opin­ion. Find a non-trivial amount of peo­ple to com­mit to pay­ing you for the ser­vice you pro­pose. No mat­ter what else they say to you, if they don’t hand over the money you’ll know how they really feel.