What Went Wrong
1. Project Scope
Natural Selection 2 was an ambitious follow-up to our free independent Half-Life mod, Natural Selection. Our original plan was to make a quick commercial sequel on the Source engine to start making money. Instead, we spent six tough years developing our own engine and almost going out of business multiple times.
We also decided we wanted to make the most moddable game ever. We did this by implementing 100 percent of our game code in Lua. There wasn't a Lua IDE on the market that we thought was good enough, so of course we spent more time to write our own (called Decoda).
NS2 is all written in Lua script. (Click for larger version)
The final product was more commercially successful than we had originally hoped, but that came at a big cost. The game was too large in scope, took too long, and because of that, ended up being very high-risk.
If we could do this over, we would have reduced the scope of the game significantly, or made a smaller game, so we could get cash flow positive sooner. We certainly wouldn't have made our own technology as part of our first product. We bit off more than we should have, and the jury is still out if the engine flexibility, modding, or lost time will make up for it.
2. Ignoring Team Problems
We have a completely unstructured environment, without task lists, priorities, managers, producers, or even job descriptions. We rely on people to figure out the best thing for them to work on, and how to do it efficiently.
Early on, we hired some talented folks that didn't fit in. They had the skills and we liked them as people, but we now realize that they were demoralized without structure. We were so consumed with making the game that we didn't notice these team problems -- only that morale was low. We would've been better off had we realized that personality fit and talent aren't enough: people need to mesh with your working style. When people are unhappy, it spreads through the whole team.
The original happy team.
Now, when we interview, we try to determine from the candidate's personality and experience if what kind of work environment and structure they expect. We tell them that we are completely unstructured and it's up to them to figure out what they should be working on: not a manager or producer. If we get a blank stare, grimace, or nervous laughter, we know the candidate is probably not a fit. We are also being more proactive about asking people what they want to work on and how they want to work, and trying to accommodate both.
3. Inexperienced CEO
We were perpetually broke. There's no other way to say it. And as CEO, it was my responsibility to make sure the company had proper funding. But as a first-time CEO, my primary skill and interest lied in game design and development -- not finance, nor fundraising. But we were learning about fundraising as we went along.
Unfortunately, not having any money is not a good position to be in raise more money. Legal expenses were exorbitant, so we handled the contracts and paperwork on our own as much as possible. And every time we started to run out of money, it was a huge burden to drop everything and try to raise more. It was like a wound that would never heal.
Common startup advice says you should raise twice as much as you think you need. In our case, three or four times would've been better. The same work is required to raise $1M as it is to raise $100k -- you just have to talk to different people. No matter who you talk to, you have to convince them you're going to make them a profit, and you're still in for months of negotiating and legal work.
I often say that starting a company is similar to getting your MBA: it takes about the same amount of time and money (more, in our case) and will teach you everything you need to know to start a business successfully. Maybe your first business won't be a success, but at least you weren't in school, where there's not even a chance of success.
Looking back, I spent most of my time working on the game, not pitching to investors. If you decide to raise money, jump into it completely. Pitch constantly and refine your technique every time. Constantly meet prospects and work your contacts. Get multiple parties interested so you can leverage their offers against each other. Make sure each party knows you're going to get funding with or without them. In short, do it all at once and make it count.
Next time I would probably look for an additional partner who is an experienced fundraiser and CEO so I could stay focused on what I most cared about: the game.