I'm a firm believer of the venture capitalist-style approach to solving problems. Rather than doing many small things that you hope add up, it's much more effective to work on projects that are high risk and high reward.