Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The $20 Starbucks Test

news.ycombinator.com

Post reply on HN
Print view View original

1–10 of 19 posts

The $20 Starbucks Test

#1

Post by ced83fra »

The "$20 Starbucks Test" is the best test of a startup idea. You don't even need a prototype. You just need to get to your local coffee, walk up to strangers and talk to them about your startup idea pretending it's your brother's. Then collect feedback.

Great... great ?! However good this idea might be, some might have some problem to strike up a conversion with a random stranger.

How would you (have you) overcome this ? What pushed you to do it ? Do you have any idea of a startup-idea-test that does not involve coding ?

Any advice welcome.

Re: The $20 Starbucks Test

#2

Post by al2o3cr »

One gotcha with the Starbucks approach is that it's very dependent on how much the audience for your startup overlaps the audience for Starbucks. This is going to heavily depend on both your startup and your area - asking random people at my local Midwestern coffee shop about a new AP database-as-a-service is not going to go well, but it might at some spots in SF.

One approach I've seen used to counteract that issue is the old "landing page without a product", where you use targeted ads to put your landing page in front of people who might be interested.

There's a second gotcha, common to both approaches: there's a significant leap between "gets excited about the product / signs up for the mailing list" and "actually pays for the thing".

Re: The $20 Starbucks Test

#4

Post by brudgers »

al2o3cr wrote: post #2

One gotcha with the Starbucks approach is that it's very dependent on how much the audience for your startup overlaps the audience for Starbucks. This is going to heavily depend on both your startup and your area - asking random people at my local Midwestern coffee shop about a new AP database-as-a-service is not going to go well, but it might at some spots in SF. One approach I've seen used to counteract that issue…

The story patio11/Mckenzie tells about Appointment Reminder is about half way down the page here (search page for 'chicago'):

https://www.conversionaid.com/podcast/patrick-mckenzie-kalzu...

Re: The $20 Starbucks Test

#5

Post by ced83fra »

onion2k wrote: post #3

The "$20 Starbucks Test" is the best test of a startup idea. Is that best in a quantifiable sense, or best as in "the one I like most"?

It is the best in that you will have the maximum amount of feedback/suggestions/critics. It helps you discuss with your potential users and helps you refine your ideas. It's one aspect people critically need at the beginning. It's one aspect people do not have with the anything that uses ads on internet (landscape+ads to see if some facets of your idea are successful, or prototype+ads).

Re: The $20 Starbucks Test

#6

Post by onion2k »

ced83fra wrote: post #5
onion2k wrote: post #3

The "$20 Starbucks Test" is the best test of a startup idea. Is that best in a quantifiable sense, or best as in "the one I like most"?

It is the best in that you will have the maximum amount of feedback/suggestions/critics. It helps you discuss with your potential users and helps you refine your ideas. It's one aspect people critically need at the beginning. It's one aspect people do not have with the anything that uses ads on internet (landscape+ads to see if some facets of your idea are successful, or prototype+ads).

It is the best in that you will have the maximum amount of feedback/suggestions/critics.

I can't imagine that's true. In a couple of hours I reckon you could talk to maybe 20 people in a busy Starbucks. Spending a couple of hours writing a great Show HN or Producthunt post could get you more feedback from a wider audience without costing you $20.

It helps you discuss with your potential users and helps you refine your ideas. It's one aspect people critically need at the beginning.

Strangers in a coffee shop aren't "potential users", even if they think they are (nor are HN readers for that matter). Any randomly selected group is effectively useless for gathering data - there's too much noise and not enough signal.

If you want good feedback you need to approach people you know are actual potential customers. You can't rely on luck. Go to places where they gather - user groups, meet ups, websites, etc.

Re: The $20 Starbucks Test

#8

Post by tmaly »

I just started reading The Mom Test. I highly recommend it. Your questions, even pretending to be about your brothers idea, are still introducing a bias. You will get bad data. The book shows what type of questions you should be asking to avoid these issues.

Re: The $20 Starbucks Test

#9

Post by nefitty »

$20 could buy you a lot of honest opinions on Mechanical Turk. I've been considering this route lately. I believe it even lets you solicit work from people based on specific demographic data.
Post reply on HN