For what its worth- my first steps when I come up with a 'killer' idea.
1.) Wait a week- if still excited about it go to 2.
2.) Discuss with my other half. As a non-techy she'll often have a very different view on things which is invaluable.
3.) Find and purchase a domain, and put up a sign up page based on a cheap theme.
4.) Buy some impressions from Adwords - this is enough to give me an idea of if anyone is already looking for what I want to build, and if they are get me some email addresses.
5.) Talk to prospects found in 4.)
6.) If still convinced its a good idea, I might start writing some code.
It is so easy to get going with writing code straight away, rather than thinking about whether you have any customers and how you're can reach them. The above approach has saved me many months of wasted effort, and found me a number of lucrative niches!
This is great advice for building a project for which you can find measurable demand.
If you have picked an idea for which there is no searches yet, but people might want, don't give up. It's where you have to listen to guys like Hiten Shah and Eric Ries that say get out there and talk to customers first. Just know you'll be building the channels to reach your customers too and it's not the same as being able to look into search results that aren't being met with a product.
I don't think there's an exact answer I'm afraid, depends on the product, how much its going to sell for, and how easy it is to find good keywords that match your product (and which people are searching on!). But maybe £200-300.
What I will say- is that if a decent numberpeople are signing up with interest in your product, on the basis of a single page of html, with no release-date, screenshots or demos in sight, then its probably something they REALLY want!
re: adwords - impressions vary from industry to industry. instead of focussing on impressions as a measure of success, here is what you ought to test when running an adwords campaign:
Test 1: are people clicking on my ad? are they at all interested in what i'm offering
Test 2: if not clicking, then how can i change the ad to gain clicks?
Test 3: if not clicking, how can i change the keywords i'm bidding on?
Test 4: if click throughs == true, how is the landing page working? am i converting people via an email sign-up or some other call to action?
Test 5: if people are coming to the LP but are not interested, how can i improve conversions? A/B test components of the landing page is a good start.
This is a matter of regional vocabulary, and the local cultural expectations that go with it.
You ask any aspiring developer in Silicon Valley what they're working on and it's always "oh yea, I have a startup I'm trying to get off the ground on weekends."
In Toronto, there's a very vibrant startup community, but back there we call them "side projects". Really same idea but the word "startup" means something rather different than a few kids eating ramen. There are plenty of freelancers who band together into a consultancy and have a couple of startuppy products they're working on together. Effectively a bootstrapped team working towards a unified vision, but they'll still be shy about calling it a startup.
I suspect a lot of this has to do with the investor and acquisition community. As someone who moved down to Silicon Valley from Toronto, I was very uncomfortable calling what I did a "startup" up until I was raising funding or entertaining acquisition opportunities. It just sneaks up on you when it's an option, and these options hardly exist in Toronto.
> This is a matter of regional vocabulary, and the local cultural expectations that go with it.
This slow realization made me give up calling people out on calling themselves hackers. It used to be that a hacker was someone who was particularly adept at finding clever ways to misuse, change or exploit something. http://hacks.mit.edu/ and http://www.catb.org/~esr/faqs/hacker-howto.html were the canonical references. Hacks were contrasted with engineering -- both constructive creative output, but one involved detailed in-the-field knowledge while the other was a more principled approach. 'Street smarts' vs 'book smarts' if you will. Both were needed for a successful enterprise. Calling oneself a hacker was equivalent to giving yourself a nickname -- socially awkward and decreasing the likelihood the moniker would stick.
Now 'hacking' is synonymous with prototyping, or even more esoteric types of development (what the hell is 'hacking education' supposed to mean?). This isn't entirely untrue as quick prototypes often contain "awful, we shouldn't be doing this" kinds of code. However, they do tend to lack the righteous "awful, we shouldn't be doing this but damn is that awesome" code tradition calls for.
Like you said, definitions change and the only thing that matters is if your idea is communicated; ultimately you have to go with the flow...
Very true, that's another example of the same phenomenon. In Toronto, "entrepreneur" is often a dirty word among who actually get stuff done. It usually refers to your typical MBA with no industry experience looking for a code monkey to fulfill his vision.
I don't think "hacker" has yet been tarnished in Toronto the way it has been here, though.
Yep. I think the term "startup" is soooo over used. I don't know if it is just here in Silicon Valley (the SF Bay Area in general) but it seems like everybody wants to say they are working on a startup. Because, really, if you are not working on a startup then you're just working and where's the cool factor of that? A lot of the "startups" I see here on HN aren't really much more than "side projects"... some aren't even that. I don't have either... I just have a good job and I tinker with stuff in my spare time.... I'm such a failure. :)
True. In some countries, specially in the developing countries, going to the goverment for money is actually a common course of action for relatively risky but promising tech ideas.
I agree with this for the most part but I don't think incorporating is that much of an issue. I can't speak for the US but in Germany it takes about 5 hours to form a company (UG similar to a LLC or Ltd.) and it costs about 100 or 200 EUR. With a company in place you have peace of mind for all the co-founders and nobody has to worry about getting screwed over if everything is notarized and on paper.
Creating a company forces you to think about who owns what, who is and isn't vesting and how they vest. It raises other conversations like who _does_ what and what value each person contributes.
The post says: "Just if things are going well you’ll create the company and build the structure of it."
This is the worst time to start the conversation about all the above.
I've been exposed to bitter disputes between company owners (not my company) about equity ownership and you can't possibly imagine how nasty things can really get.
These problems are not an issue when there is no pie yet, just the pan you plan to bake the pie in. Who cares about dividing that up. But pretend your business starts generating $100,000 per day in cash in the bank. Now ask yourself "who owns what?".
You may think you know your co-founders and co-owners well, but most people are unable to predict how they themselves will react in many situations, so don't assume you can.
My apologies if this post comes off as negative. Creating a company and watching it grow is one of the most exciting and rewarding things you can do, and it's so much more fun when you have co-founders and investors to share your success with. So to keep it fun, have the hard conversations when they're easy to have.
> Set a deadline of 2-4 weeks to ship the first version of the product.
I blame this kind of mentality for much of the derivative rubbish the startup ecosystem is polluted with. We don't need any more photo/video sharing, link voting, or file sending apps.
I'm all for validating assumptions early, but some ideas (the ones that matter) are much, much bigger than 4 weeks to an MVP.
I believe that being a first time entrepreneur should not think of creating something that it's too big. Because it makes you miss the warnings that you should get.
I believe that if you want to create an Amazon (6+ years of under break-even) should be done when you have more experience as entrepreneur.
I think that a first time entrepreneur should resolve a real concrete problem, also not that big, but this way he will understand much more and he will have some chance of succeeding.
This is the opinion maturated for my personal experience.
It's one of the many different ways to start a company. Assuming you are not selling a physical product and don't actually require an office, and also assuming you are (or know) a developer that can work for free... Sure, that's exactly the right way to do it in my opinion.
Actually, that's exactly the method that me and my friendly developer are using for our own product, which should launch sometime this year. We both have a day job which would be hard to leave (we both have families to feed). And you're completely right that a single server is enough to host the MVP and we can expand further once we know we're on the right track.
You're very close. I agree with "Skip"[1] the rest.
However, don't bother with a team or co-founders. Just build an experiment yourself. Then use that experiment to find a co-founder if you don't have one.
I think co-founders should be people you've known for a long time, a decade or more (if you're 20 and haven't known anyone that long, then consider working for awhile and building out a network until you have someone you know well and trust and have known for 5 years.)
If you can't build the experiment yourself then you're not ready to lead a company in the space of the experiment (managers should know how to do their subordinats jobs).... so if you're an idea guy, figure out how to program. Do your experiment ins super simple code or easy to use tools, a real engineer can do it right later if the experiment succeeds.
And above all, don't build a company first. There is an unending list of chores that are involved there. All the time you see advice like "get a lawyer" and "get an accountant". You don't need an accountant when you have nothing to account.
[1] You actually say "build a team and skill the rest", but I'm pretty sure you mean "skip".
Exactly. My first few projects I bought domains and set up google apps for those domains, paid to have logos done, etc. After 3 or 4 failed experiments on different domains I realized how much time I was wasting setting up all this stuff when I should have just started building first.
Although, a component of why I did that was because I had big dreams. I wanted to "play startup" because I didn't actually have any clue how to build one. The obvious thing that you see about every company is their public website and their logo, so it makes sense to start mimicking that first.
The irony is that once you start to understand how to actually build a company you often end up picking a name and incorporating as the LAST thing. In fact, you try to avoid it for as long as you reasonably can because it makes you lose your focus on just simply building an excellent product.
Funnily enough I am completely the opposite way round.
I always wanted to start writing the code first, but then started realising that was a waste of time if there was no-one to sell it to.
So now I always start with trying to find customers, via a simple sign-up page or similar, before I even think about writing code.
Not saying that either approach is right, but interesting that we've come at it from opposite angles, and both ended up changing our opinions completely!
I come from this path too. I've usually found that writing code is too easy. Confronting the potential customers is harder to do and I believe the hard part of the business.
IMHO
It probably depends on what exactly the idea is, but yeah I think what you are saying has a lot of value. I never made a sign up page. It was like, "I'm going to build the next reddit.com" kind of ideas. I don't think a sign up page would get you anywhere with something like that.
What I don't understand from your article is what exactly you're asking one not to do. I like all your recommendations, but I think incorporating and getting some proper legal and financial guidance (i.e. tax implications) is important too. But aren't you technically a "startup" at that point? Does a StartUp need office space etc to be qualified as one?
Great post!
But actually, there are no ideas that are good enough to pass all the tests. And every ideas need small innovations that we all should come up with. And to come up with small innovations, we have to tacle the idea seriously. So, I think most important thing is just doing it. But we should not start it with others' money, if possible.
Step 42) When you go public with your product, make sure your infrastructure can handle the load so you don't suffer a failed launch. Many users will not give you a second chance.
1.) Wait a week- if still excited about it go to 2.
2.) Discuss with my other half. As a non-techy she'll often have a very different view on things which is invaluable.
3.) Find and purchase a domain, and put up a sign up page based on a cheap theme.
4.) Buy some impressions from Adwords - this is enough to give me an idea of if anyone is already looking for what I want to build, and if they are get me some email addresses.
5.) Talk to prospects found in 4.)
6.) If still convinced its a good idea, I might start writing some code.
It is so easy to get going with writing code straight away, rather than thinking about whether you have any customers and how you're can reach them. The above approach has saved me many months of wasted effort, and found me a number of lucrative niches!