How not to test your startup idea; how to test your startup idea. | Hacker News Reader

How not to test your startup idea; how to test your startup idea.

wademeredith.com

35 pointswmeredith19 comments

I dont think starbucks advice is that terrible. Even though mechanicle turk does seem like a good alternative, I would much rather go to a coffee shop or a bar where I know startup savvy people hang out and discuss it in person. But one of the best advice I ever received is - Start a blog and gain some traction for it before you launch your MVP. Discuss customer stories and problems you are looking to solve. Ask people opinion about your solution, discuss/debate in public. Use these discussion to keep on refining your direction and feature set. As a positive side effect you might end up creating a passionate community around your solution before launching and have list of people interested in paying you money beforehand.

Right. I feel that this article doesn't give strong arguments as to why the advice is terrible in the first place.

Lying to people and manipulating their sympathy to validate a business idea is wrong. I imagine the OP didn't imagine this was something that had to be pointed out to people--even if you don't agree with that viewpoint it's not exactly a difficult one to imagine, and it's better to state as a premise that it's a bad idea rather than bog down a blog post--which seemed more focused on discussing alternative methods--into a tedious moral argument that won't convince anyone.

I dont think saying my idea as my friends idea is much lying. But I do think that you should be straight about it being your idea and use conversation skills to make sure people still give you honest advice.

You lie to people every day just to function normally in society. You must be a terrible sinner. Boo hoo.

While were at it let's throw in a good dose of violence, because that can help us get what we want as well.

I imagine that mechanical turkers don't future customers. You want to find people that would pay to use your product, not the other way around.

Breaking it down -

"If you spend your time building something instead of conning in coffee shops, you can put a minimum viable product in front of someone"

FALSE. I used to say that, until I think it was Harj who corrected me. His reasoning (paraphrased) was along the lines of "you'll spend 3-5 years if it's a great idea, wasting a few days/weeks pausing & checking your assumptions is insignificant time in the long run. It makes sense to spend few minutes/days making sure I don't waste weeks, months or even years.

Option A: Mechanical Turk

MAYBE. It depends on your audience. If you are InDinero building a service for small businesses, it does not make sense to ask MTurk to vet your feature set. They're not your target audience.

Option C: Get people to pay/commit to it

MAYBE. I bet I could ask 30 of my closest friends to commit simply because psychologically, people want to help. People will feel warm and fuzzy by giving you a few bucks to help you build your idea. Ask a business to commit $1,000+ - that's real validation.

Breaking down why you're wrong about why I'm wrong :-)

A) 3 to 5 years to MVP? You're doing it wrong.

B) No solution is perfect for every situation, I'll concede that. It doesn't make me wrong. In fact it assures that what I said will be right, at least some of the time. I'm being snarky, but the point is, of course MTurk isn't going to work for every application. Use some judgment. Maybe the coffee shop thing would work better for you, I just think other alternatives are more worthy of investigation.

C) Once again, you're doing it wrong. You got it there at the end though. Ask businesses or real potential customers to commit, not you're closest friends. In what scenario would that route work?

A. MVP usually takes at least a month if not more. Do you want to waste a full month of your life? That could be avoided by 30 minutes of talking to future customers.

B. & C. - "Maybe" - not wrong, just maybe. There needs to be emphasis on "real potential customers." I've made the mistake of thinking I'm doing something great because friends (who might buy it) liked the idea. Just trying to point out the need for honest feedback.

(Sorry if I came off as snarky ;) apologies)

No worries, publishing on the Internet for years has given me quite a thick skin. I'm also always willing tondo use counterpoints. It's the best way to refine an argument, or startup idea for that matter.

So what do you think is the proper way to validate your ideas before you commit time and effort to building the product?

Build a prototype and explain the vision. Prototypes are useful in so many ways - validating & going to users, pitching & getting investors, etc.

Prototypes should take a few days or a week to build, unlike MVP which probably takes months (or more).

Talking to people without a product doesn't always work. Most people can't imagine what it would be like to use a product, so they can't reliably tell you if they would use/pay for it.

See, now that he brings it up, I have the same issue about the Google Adwords campaign that most folks have with lying about your brother: you don't really have a product page or anything to sell, but your Google ad implies you do. You're lying to people to see if you can trick them into clicking on your ad, in the name of statistics. If I clicked on such an ad, I would be mildly pissed.

I feel like these ideas could not have been used to test a Facebook. Asking people if they want to use a new social network without actually seeing and touching it would not have been insightful. You could describe to the wall but because the concept was so new I feel like most people including myself would not necessarily see the value of it right away. Asking people to commit to pay for it is even more of a stretch.

It doesn't apply to all situations.

Showing people a minimum viable product is not feasable in most situations. You should try to validate your innovation idea earlier in your development process. Talking to potential customers is still best - I think.

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

This is probably the best thing you can do, if you are ballsy enough to do it. Personally I always feel a bit bad/like a con artist/ about going to people and saying "Hey, I totally have this great thing and I will likely be pretty much able to hopefully give it to you in maybe a few months, perhaps a year or so, if you would just buy it right now!" ... that's not a good way to conduct business. (imho)

Now, what does work for this sort of thing (and my conscience) is Kickstarter (and others like it). If you can get a lot of money donated from people who know what they're getting into, then you're onto something.

You don't have to get people to buy hot air. What the author has in mind is probably either a) "Assume that this product is fully functional. Will you pay for it $X?" or b) "I am planning to build this product. Would you promise to pay me $X when I deliver this product?"

Reply on news.ycombinator.com