In determining what makes a bootstrapped startup fail versus one that succeeds, I often think about what really are the critical points that make or break the project.  At least, in regards to the “web2.0″ model where you just write some SaaS website that ends up getting a bajillion users and you drive off in your Maserati.  This all probably in no way applies to a startup thats say, trying to build reusable rocket ships.  You know, those funny startups that actually aim to make a physical product.

From my perspective, there are three major components that all have to be in line for success to occur:

1. The Good Idea
2. Technical ability to actually write the thing: A featureful, high quality, well tested product.
3. The business sense to actually execute on it.  To market it correctly, to secure the funds to keep it alive to fruition, the ability to land clients and sell your product.

Now in all honesty, those first two always seemed to me to be inconsequential.  Hell, #1 practically isn’t even a requirement, considering all the awful, horrible products I’ve seen advertised on tv for $20 german rags, and they certainly seem to be making a profit.  Ideas, it seems, are rather a dime a dozen.  Everyone has a great startup idea, and in reality 1000 people before them thought up the same thing anyway.  Great ideas are great, but great ideas do not a startup make.

Now the technical aspect of it, I’ve always kind of taken for granted as well.  At least in the circles I run in, everyone is technical.  The only cost to coding the website is simply the time it takes to code it — the technical difficulty of these things is rarely above trivial.  In fact its always been a generally assumed truth that the reason there are so many web startups is because the barriers to entry for making a web startup are so shockingly low.  Anyone can whip out a text editor, check out extended-bort and add some CSS styles.  Its much harder to say… boostrap a startup that makes a new kind of microchip processor, or a new kind of windshield wiper.  To me, the coding of a startup’s site is the least of my worries.

The business part, especially the sales, is the part I’ve always known I’d need help with.  I just don’t have that skill set, I don’t have the connections or the rollodex of magical leads that will buy a license.  I don’t reallly even know how B2B sales pitches work, how to get the ball rolling.  My marketing, accounting, and hiring knowledge is limited to twitter, quickbooks, and monster.com.

So from my perspective, the business end of it certainly represents the biggest hurdle.  However I realized today, that for someone from the opposite side of it, the technical coding of the site is now a high barrier to entry.  They have to raise capital in order to hire programmers and be able to pay them to see the fruition of the project.  At that point can you even consider it bootstrapping?

It turns out, of course, that both requirements are equally critical to the success of the startup.  With no technical competency, your business people have no product to sell.  With no sales people, your fantastic product will sit there with no clients.  Stalemate.

The question then, is whats easier:  for the technical person bootstrapping their own website to learn the business aspects?  Or is it easier for an MBA grad, with zero technical background, to be able to pick up a book and suddenly be able to have the ability to write a scalable, full stack web application, along with all the SA hats one has to wear to even get the thing running….  and then throw on some creative abilities as well for CSS and logo design.

You can guess which side I’m throwing my bets on.  The technical bootstrapper will at least be able to get his product into creation before needing any serious VC backing.  And probably even through grassroots marketing get some customers via word of mouth, twitter, blogs, hell just emailing out around to people.  The technical bootstrapper may not be very good at executing the business side but at least they can take a stab at it and hopefully get some revenue coming in, learn the ropes a bit, and then be in a good position to either hire a sales/business person, or take VC funding and hire a sales/business person.

Maybe I’m dead wrong here.  If you’re a budding CEO of a 1-person startup with a great idea, maybe you’ll easily be able to go out and find a team of developers who know what they’re doing to build your idea for you, and get paid in stock options to boot.  Maybe it will be an awesome website that will hockey stick its way into revenues within 90 days and get that 1% share of that $6b market you know about.

I just don’t see it.  With no technical competency to bootstrap yourself with, you’re stuck unless you give up the bootstrapping idea and go get funding of some sort to be able to pay out salaries and wrangle up some nerds.  Suddenly, making a web based startup doesn’t sound like it has low barriers to entry, at all.

At some point my installation of capistrano here in OSX got corrupted or something.  Either way, whenever I would try to run `cap deploy`, on a project that everyone else could deploy just fine, I’d get an error message like this:

Brians-Laptop-2:railsapp brian$ cap deploy

/opt/local/lib/ruby/site_ruby/1.8/rubygems.rb:636:in `report_activate_error':
RubyGem version error: capistrano(2.5.5 not = 1.4.2) (Gem::LoadError)

Which was weird for a couple reasons.  One, the capistrano gem that was frozen into the rails project was specified to be 2.5.3.  Capistrano 2.5.5 (which this was saying I had activated) should have run just fine that way.  But then second, where was this 1.4.2 mention coming from?  Nothing in the project talked about capistrano 1.4.2.

So, taking a look at my installed gems, it gets even more fun.

Brians-Laptop-2:railsapp brian$ gem list | grep capistrano

capistrano (2.5.5)

Well, lets just uninstall that sucker.

ERROR:  While executing gem ... (Gem::InstallError)

   Unknown gem capistrano >= 0

Wooo fun.  Gem list says I have it, gem uninstall says I don’t.  Anyway, finally figured out that at some point capistrano had been installed to my home directory.  Something like ~/.gem/ruby/1.8/gems/capistrano/blahblahblah.  Thats the one it was somehow using.  So, delete/move that .gem directory.

Then if you actually look at the stack trace from the original `cap deploy` command, you’ll also notice something similar:

from /Users/brian/.caprc:1:in `load'

Well I don’t know wtf this .caprc file is doing in my home directory either, but get rid of that too.  Finally, there, I was able to `sudo gem install capistrano` and correctly install capistrano ( 2.5.5 ) to its correct location.  From that point, cap deploys should work just fine.

Hope this helps anyone else googling around for the same arcane set of keywords I was.