Don't launch a company, launch an experiment

http://viniciusvacanti.com/2012/01/23/have-idea-for-a-startup-dont-launch-a-company-launch-an-experiment/

by mikeocool • 15 years ago

61 17 15 years ago

I feel like a lot of people use the term "startup" as a cover for "unbootstrappable". If your idea requires big-league success (think twitter), or is otherwise a failure, surrounding yourself with buzzwords will not escape you from reality.

My personal opinion is to make an MVP that has a revenue model from day one, and by releasing early and often grow your product while being paid to do so. If no one buys it, stop. It's not as sexy as a social media home run, but it just might make you some money without VC intervention.

billpatrianakos3• 15 years ago

The word startup really has lost all meaning. Everything is a startup now. If LOLCats launched today it would probably be called an Internet Meme Startup. More troubling however is where people have gotten this idea that you don't need to think about how to monetize something from the start. There are a ton of people out there who literally think they can just launch a website (in many cases it's just a static site, not even a useful app) and it'll magically make them money. Do they think ad revenue is generated simply by posting the ads? Do they believe that they'll start getting checks from VCs in the mail out of nowhere? Do they actually think what they saw in The Social Network was an exact reproduction of reality? What is it? Please tell me because it bugs the living hell out of me! I constantly get people asking me to build them a website that will make them money but the site they ask for is either a clone of an existing site with nothing but a name difference or they just ask for, and I swear to god this happens, "just gimme a template and I'll drag and drop the parts of my app in using (here they either mention Ms Word, Frontpage, or Dreamweaver)". Talk about a WTF moment.

rdouble• 15 years ago

LOLcats is perhaps a bad example as that guy was making $5600/m out of the gate and has grown to a 50 person company with $41M in investment.

rokhayakebe• 15 years ago

You know why you hear of companies that just blew past xxx,000 users in no time? Because that is not what usually happens.

If you do not know what you want to work on, then test & drop as quickly as possible.

If you know what you want to work on, then my advice is to shift you thinking to 'I am in this for the next 3 years (absolute minimum)'. Commitment changes your decisions from tactical to strategic.

Building a company requires 1 ounce of skills + 4 parts of persistence + 5 parts of patience. Get married to the problem, but be willing to change the proposed solution.

vacanti2• 15 years ago

That's good advice on getting married to the problem. We clearly weren't married to it.

That being said, you should still experiment on the problem you are solving as opposed to waiting for 6 months to roll something out.

hanula• 15 years ago

Good article as a reminder what should you focus on. That's one of my New Year's resolutions about code projects. To show something in quite early stage of development - instead of trying to get everything done. Having a feedback straight away, from users that would eventually use your software, will validate your idea. Recommended to all of those who starts projects and abandon them after a while of intensive work which wasn't shown to anybody. It of course applies to small projects, or ideas to test. If you plan to start a big company, then there are other ways for validation.

tjr• 15 years ago

I was surprised to see the experiment deemed a failure after only one week. Do most folks feel that one week is sufficient time to ascertain this? One week seems hardly enough to even start propagating a website through Google... ?

vacanti2• 15 years ago

Definitely not 100% conclusive evidence. But, after sending it to our almost 1,000 friends and families and seeing just 2 urls submitted after a week, we knew it was going to be widely used. After all, these are people very motivated to try the product (or I guess they could secretly not like us...)

jaredsohn• 15 years ago

knew it was _NOT_ going to be widely used

billpatrianakos3• 15 years ago

I don't think it's enough but this might be the new thing to do. Leo from Buffer posted something last week about how they put up a simple landing page to gauge interest and decided to only spend time on the product if there was enough interest. I think he mentions building version 1 in just 7 weeks. If you can create an MVP in a month or so (less is even better) then I think that's the way to go! Maybe Google is irrelevant. Maybe it's better to leave Google out of it. If you can send it to some choice people and you get some good feedback primarily from those people sharing your link then that's probably even better than having Google on your side. If I had a choice I'd want my product to be so good people take the time to spread the word for me rather than relying on Google to hopefully point the right people in my direction.

keiferski• 15 years ago

This is really just business 101. Outside of the tech startup world, most businesses aren't launched without significant market research. That's mostly a result of the upfront costs (huge in real-world vs. tiny in tech).

Of course, both approaches have their plusses and minuses.

danielharan2• 15 years ago

It's really easy to rationalize that your business will work, and takes discipline to try and prove yourself wrong. This is something I've seen in the lean startup movement, but rarely in business.

Most of the traditional businesses I've seen launch had business plans, and their "market research" section was a bunch of untested hypothesis. Have you ever heard a banker and adviser simply state: "I've no idea if people would actually buy this. Do you have tangible proof?"

Instead, they try to sound smart by blabbering about whether they believe your total market size, growth rates, etc...

mrushton14• 15 years ago

"Just emailing it to lots of friends and family"... Seems like this should be a part of a bigger "experiment". I'd ask friends/family before I built anything. I'm left unconvinced enough was done without more detail.

danielharan2• 15 years ago

OP mentions in another comment that they sent it to 1k family and friends.

ilaksh• 15 years ago

This is a great example and a good philosophy, but I think there is a common false belief that not having users means that there isn't a need.

I think there is a lot of social stuff going on that leads to a particular service or idea being successful. Its more about trends and social networks than people realize.

There are lots of services which are similar to very popular ones but get little traffic. Its not that their idea isn't good, its just that their service isn't popular.

I'm pretty sure that with the right marketing and social happenstance this could become a popular service and make quite a bit of Google Adwords money.

Which is not to say that the concept of an experiment is not a good one or that people should beat dead horses.

dkrich• 15 years ago

When I was in bschool, one of the founders of Aardvark came and gave a talk about their early days. What it boiled down to was a few super-smart guys taking EXACTLY the strategy prescribed in this article (namely getting a landing page up and manually handling the backend processes).

They eventually sold the company to Google for $50MM. Great for them. Where is Aardvark today?

I said it then and I'll say it again- experimentation is for laboratories, not for businesses. At the core of a business is a product or service that people exchange money for because it solves a problem. Too many startups set out with the mission of meeting their founders' needs (ie, quick money and fame) as opposed to any real societal needs. The irony is that if the founders worked from the consumer need backward and took a more long-term approach, their chances of success would likely go up substantially. The Lean Startup Method to me involves some mix of laziness, lack of vision, and giving up on your product and vision prematurely.

The faster your product proves viable, the easier it will be for competitors to rip it off anyway (Groupon, anyone?). You want real, sustained traction? Do the things that others are too lazy or unwilling to do because they are too hard or involve some risk.

I'm not advocating over-engineering, and it is great that development tools have progressed to the point that designers and developers can perform experiments for free while honing their skills. But to put an initial idea out there and let the public bat it around and bastardize it until it has some weird mixture of users and you don't even understand your own product is a path to nowhere. Remember- tools are merely used to complete a task. Let the task dictate the tools, and not the other way around.

Don't launch a company, launch an experiment

Related Stories