Hacker Newsnew | comments | show | ask | jobs | submit login


I did the same thing -- http://outletmapper.com

Some learnings:

1. It really does free up your head to think about harder issues (read: getting in front of the right market) if you're not distracted by code. For me, code is "my happy place", so it is too easy to spend too much time there.

2. I asked hard questions when picking someone to work with. Not just "have you used MapKit" but "how would you do hit detection inside an arbitrary polygon efficiently -- on iOS?"

3. Keep feedback to the right level. Sending back a stack trace and commenting on an edge condition was fine. Somehow sending back a patch seemed over the line.

4. The product will only be as good as your specifications. Get them right. Build test cases. Build a set of better test cases that you use -- that you don't share.

5. Get it right before paying someone. Paper, PPT slides, HTML demos all help to get the ideas baked before your money is pouring out the door.

6. If you're in the US, hire US. Why? You can have them sign a contract that is actually enforceable.


How detailed are your test cases? Do you have a sample document, I am interested to have a look at it.

#6 If you're in the US, hire US.

Mmm I disagree with this. I think the risk that someone decides to steal your idea or code and go run their own business are fairly low. I accept to take this risk to be able to reach developers that I will pay 4 to 5 times less than the local developers in my country.


> I accept to take this risk to be able to reach developers that I will pay 4 to 5 times less than the local developers in my country.

And you'll typically need 4-5x as many of them, and/or they bill 4-5x as many hours. This is not theory. I've seen this happen in reality, over and over again. Does this happen in all cases? No. But I've seen it happen enough that I make decisions based on it. And look at Mythical Man Month for a primer on why larger teams tend to be worse than smaller ones (worse efficiency, more communication overhead, more cheese-moving, etc.)

Having flashbacks to the worse quality startups I've seen that had 5+ person teams in Uruguay, Pakistan, etc. when they could have had the same work accomplished by 1-2 more elite, and local, same timezone, same culture/language, more "expensive" (nominally, but not overall) Americans or Europeans.


#4 I created an artificial outlet mall data file with all of my weird edge cases.

#6 I don't want my job outsourced. That starts at home. More generally, when I'm in a position to be an employer, I try to be of the sort that I would want to be employed by.


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact