Comment by DanielBMarkham | Hacker News Reader

Comment by DanielBMarkham | Hacker News Reader

First you might want to consider that you did nothing wrong.

I just did kind of a recap of my startup career over on my blog at http://www.whattofix.com/blog/archives/2007/10/who_me_worry.... Perhaps tonight is the time for people to think about these things.

I read a really great article that was linked from here a couple of weeks ago. I'll use a baseball metaphor to describe it. I hope you'll understand the reference. The point was that it's not hitting the home runs that matter, it's getting the times at bat. That is, if you're always swinging, and you're always learning, it's just a matter of getting enough pitches.

You want to fail fast and learn from it. What I've found that sucks is that to qualify an idea, generate the site, and really get humming might take 4-8 weeks. This means you can't just try out ideas by joining a group for a couple of weeks. Ideas need to be beaten, nurtured, and planned. So if I had any criticism, I'd say "go out there and build something that works, then ask what did you do wrong". Because failing is fine, as long as you keep failing forward.

I just want to say that was a great article. I think your baseball metaphor really gets to the point. It doesn't matter that you've tried so many times. The fact you are able to go out on a limb and try to create something from nothing shows you're already on the right path.

I remember pg saying somewhere that a lot of making a successful startup is skill, but another large part is luck. If you keep putting yourself out there, you might fail, but each time you get to that point, you increase your chances of getting lucky greatly. As long as you can take something away from each try you really didn't lose anything except a little time.

Thanks.

You mean it could take 8 whole weeks? That's like...two months! We could all be dead by then.

I'm obviously from an older school of thought. I plan on giving an idea two years...if it's not good enough to justify that, then it's not worth any of my time.

Within 8-12 weeks I should know if a project is going to fly, assuming everything works okay. That's not break-even, that's just knowing the thing has wings. My pattern is to take some time to try to knock it down, then prototype, then bring in some users, then evaluate my market traction. That's just for the initial call. That's milestone A.

YMMV. If I spent 2 years on every potentially good startup idea that came along I'd be 260 by now. Once you're flying, sure, give it a couple of years.

I've gone back and forth on the fail-fast vs. go-deep approaches. After working at my first startup, I leaned towards the go-deep approach, because they went through 4 business plans in the year I was there, and if they had concentrated on #2 (they were literally a week or two away from having paying customers), they might have had cash to survive #3 and 4.

Then when I was graduating from college, the fail-fast approach started making more sense, because people were making tens of millions on really simple startup ideas like del.icio.us, Flickr, and Reddit. These succeeded because they happened to hit the right niche, because they had supercool, detailed technology. Nowadays, I'm leaning back towards depth, which sucks for us, because our idea is a fairly shallow fill-a-quick-need idea and my cofounder is more into the fail-fast hit-the-lotto approach.

My two years in financial software has kinda rubbed off on me: I think that the real optimal strategy is to do whatever everybody else is not doing. So startup #1 failed because it was flitting between ideas on the eve of the dot-com bust, when everyone else was doing the same thing. Flickr, del.icio.us, etc. did well because they were founded right at the tail end of the dot-com bust, when the recession had killed off everyone else going after the easy pickings and left a lot of untapped niches. Now may be the time to go deep, because there are a zillion Web 2.0 folks picking off every single social networking idea, every AJAXified calendar, etc.

I know the feeling. You can't just churn for the sake of churning.

I think at the heart of every entrepreneur is a contrarian. I know that I'm just the type of person that if everybody else went left, I'd go right.

I don't see a conflict between failing fast and going deep. That's because when I'm in the mode to qualify the work (before I spend 8 weeks on it), I try to scope it down. Am I getting into six months of development effort? If so, is that because I want to make my technology hard to reproduce? Or is it just feature bloat -- I'm putting the kitchen sink in. I know now that I'm failing fast (giving it 4-12 weeks once approved) with small timeboxes, I'm seeing a lot more progress. Chase the users, right?

I don't think I'd go back to long timeboxes again, although I might very well chop off some really gnarly problems and chew on them a little bit at a time :)

Thanks for nice input

Reply on news.ycombinator.com