Sunday, September 25, 2011

Outsourcing

Part of this is from an email I sent this weekend to first-time founders, who are so far down the path toward outsourcing I doubt they'll turn around.


It's hard to watch. Something like $200,000 over 4 months.

There's a huge risk in outsourcing, and huge value you absolutely will not get in doing that: 


  1. institutional knowledge in the form of your key developer, and 
  2. the discovery process along the way. 

It's hard to explain the second if you've never developed software before, but a lot of innovation occurs at the source of problems encountered on the project and there's deep value in being present for that. 

What happens after the first release? Do you think it will be perfect? "No plan survives contact with the customer". 

Well, you'll invest even more money in someone outside the company developing expertise around your platform.

This is not a lean approach at all. 


What happens when their employee--your key guy--gets a job elsewhere? Your institutional knowledge is really lost now. 

I can see outsourcing for supplemental work (import/export of data, porting work, internal analytics/reports/data handling), but not for the core app. 

I'm almost certain I could find a great lead dev to join you as CTO/co-founder. The more I think about it, the more I think you're making a massive mistake, where you'll find yourself in January with something that 

1) doesn't quite hit the mark and 
2) needs more work, while 
3) you're out of money, have no or little traction. 

Then you'll decide to learn to code it yourself, because you'll still believe in it but can't afford to pay anyone. Maybe you get to raise a bit more money, but by that time it's not enough to hold down a dev for any length of time. 

Investors want exits, right? So what's the exit possibility when your team isn't yours, besides the guys with the idea? Acquirers look for team, and especially technical talent. 

Mint's founder made the calculation that each developer is worth $500k in a sale (2 years ago; it's more like a million now), so if he simply raised $1 million he could hire 10 guys, sell the company for $5 million and walk with $3.5 or something after preferences. Not bad. 

You'd be better off spending 3 months in intensive programming training, using your project as your homework. But the best path--if you're paranoid about time--is to find and hire that programmer, the one who can make something happen top to bottom without a lot of other hands on it. 

I know of 2 right now. 

But, you've got a contract, you've spent money on a lawyer to review the contract, and you've got money in the bank, burning a hole in your pocket--you feel you're almost there. Just need someone to code it up. Right? 

Wrong. It's a bad choice. Bad, bad, bad choice. Did I say it's bad? Really bad choice. 

So, startup founders out there--what do you think?
(
Read these from Fred Wilson's blog, especially the comments: 

No comments:

Post a Comment