First off, allocate expenditures intelligently.  Spend less money on ineffective advertising, and more on IT, especially if you are counting on an IT project the way the campaign said it was.  Secondly, don't talk about how much more awesome you are than the competition at something when you are not.  Thirdly, recognize your limitations, and address them.  Fourth, do some actual load testing.  Fifth, spend some money as mentioned in step one, and make sure that there is plenty of capacity and eliminate potential bottlenecks. Sixth, allocate some time and resources to training, so that you don't have tens of thousands of new users trying to figure software out at the moment it is needed.  That was like a pitcher trying to figure out a fastball while he is standing on the mound in the world series.  Dumb, just dumb.   

Lastly, and perhaps most importantly, be intellectually honest and discard subjectivity in favor of objectivity.  Everyone wants their projects to be successful, but you can't just hope that everything will work out.  But if you follow best practices, do the work thoroughly and professionally, and it probably will.