I'm almost at the point of launching my project -- in point 2, but I cut a lot -- and the things I'm most worried about are:
1. Once it's up and people actually use it, I have to be much more careful about uptime, monitoring, and changing how data is stored. This is by far the scariest thing.
2. Handling a large temporary spike in load when I publish the URL here or elsewhere
3. How -- if I actually knew what I wanted to charge for -- to integrate payment into the application, and what I'm supposed to do tax-wise with the money I'd get.
1. Yeah, that stuff is a pain, but it's a "good problem to have." Nothing to be scared about, just annoying.
2. Cache your landing page and upgrade your server beforehand. If it's any help, I survived getting to #4 on the reddit home page with a Django app that used middleware caching for anonymous users on a 2GB Slicehost slice.
3. Incorporate as a business, avoid processing credit cards at first (use PayPal, Amazon, Google, etc.). If you make money now, you have until next tax year to worry about the accounting stuff. Another good problem to have.
I've fallen for both 2 and 3; they are absolute killers.
There is a fourth too; stagnation. I actually had a startup a little while back (actually before hanging out here). I had the idea nearly 5 years before and spent about 3 years building it in fits and spurts (major warning signs there).
We went through about 4 rebuilds before I got bored for the last time (through total luck I was able to sell parts of the code/company).
[this is somewhat of an aside] I think it has made me a bit of a cynic about starting up. I see all these big ideas but there is little or no money in most of them.
I know the current fad is to fund lots of small, lean ideas. But ultimately I think a better approach is to contemplate your ideas, get hooked onto something really significant to you and then build it. I'm still looking for mine - but it feels the "right way".
1. Once it's up and people actually use it, I have to be much more careful about uptime, monitoring, and changing how data is stored. This is by far the scariest thing.
2. Handling a large temporary spike in load when I publish the URL here or elsewhere
3. How -- if I actually knew what I wanted to charge for -- to integrate payment into the application, and what I'm supposed to do tax-wise with the money I'd get.