I can tell you what I've been doing, at least. Hell, I can show you.
During the "off season," I spend most of my time figuring out what I want to improve in the following year and then trying to figure out how to actually do that. One of the fruits of that is a proposal I have for event organizers and game master: the Premier Events Reward Program.
Basically, PERP (bad acronym, I know) is intended to reward companies and groups that run popular events and adhere closely to our policies are given some of the recognition they deserve, some addiitional exposure and get first pick of event space. The program is still in the proposal stage, though, and I've released details to the GM forum for event organizers to discuss and debate so we can work out the kinks and make sure it's actually what people want before moving into it.
I dunno about y'all, but I like being able to talk about changes we have in mind and get some feedback before they're made official. Gen Con is quite literally built on the backs of our volunteers and GMs, so I like to try to include everyone whenever I can.
I've also been spending a lot of time pouring over our attendance data from 2008, analyzing both submitted events and tickets turned in to try to get a sense for trends. I've been doing stuff like graphing attendance (both submitted and actual) hour-by-hour over the whole weekend for different event types, companies and locations, trying to see where we can be more efficient or improve things - and to determine what events are our strongest performers and biggest draws.
A lot of the conclusions I've been able to draw from this data might seem a bit obvious, but there's a big difference between having a feeling about something or basing a decision on anecdotal evidence vs. actual wide-ranging data. There have also been lots of interesting surprises, like how RPGs are the only standard, ticketed event type that has fewer final players than people who pre-registered (meaning there are lots of people who sign up but flake out and don't show) or that the peak number of concurrent players for LARPs is just about equal to the peak number of concurrent TCG players (that's the maximum number of people playing at the same time, not the total number of the whole weekend).
Or what about the trend that more people submit events for Saturday, but actual attendance is higher on Friday?
I've been preparing lots and lots of graphs over the past month or so and hopefully I'll be able to cobble them together into some kind of reasonably intelligible report I can share with everyone. Not only would it be useful for event organizers when planning their events, but it would also be useful to prove to everyone just how important and useful it is to turn in your event tickets.