As part of another exercise, I have been looking through the speaker lists for the seven DDD Reading events that we’ve run in the UK so far.
Thinking about the number of unique speakers who submitted sessions vs. number of sessions submitted, I pulled the the information from each of the first seven events to produce this graph below. Note that this graph isn’t limited to those speakers who made it on to the agenda – it contains data relating to those speakers who submitted one or more sessions.
DDD1 was obviously the first DDD event – it had a hand-picked agenda. As I recall there were a couple of cancellations, so pretty much every speaker who submitted a session made it on to the final agenda. I can’t remember the precise details behind the use of “developer developer developer” as the conference name, however it is obviously a hark back to Steve Ballmer’s famous pitch (of which there are plenty on YouTube). Of course, this is also the reason why the speaker to sessions submitted ratio is virtually 1:1. DDD1 came about after I was invited to look at an Excel spreadsheet containing a list of speakers and sessions…the next thing I knew is that I had a spreadsheet with an agenda on it, then a date, then a registration process, then an event (coinciding with my wedding anniversary)…it happened so fast!
DDD2 was the first, and last, DDD to have 30 minute sessions. We had hoped that the 30 minute sessions would be appealing to new speakers who may have felt that 60 minutes was too long. Whilst that was perhaps good for the new speakers, it was a scheduling and practical nightmare…although it seemed to work pretty well on the day. This is what happens at events – the attendees can have a whale of a time, whereas the organisers can see all the imperfections that make an event hard work for them. Luckily the DDD team are hard-skinned individuals who are keen to experiment and learn, so we take on-the-day problems with a pinch of salt and think “learning experience”.
DDD3 was almost a 1:1 speaker to session ratio. It was also the first time that Dave McMahon used his Sunday name “David” as part of a session submission. The key change for DDD3 was the introduction of community voting. Instead of relying on instinct to build an agenda, we asked the attendees to vote for the sessions that they wanted to see on the agenda. This has two main benefits. Firstly, it help us with room scheduling – if we know a session received 120 votes then we know to put it in a room that can handle at 120 people. Secondly, it means that the agenda is community-driven, we get to build an agenda that the majority of attendees are going to be interested in. Of course, like all systems, it’s not as precise as this and we do have to make some scheduling decisions, e.g. there might be two speakers discussing the same topic, we have to work out how best to handle that. I think that tells you that we don’t often take the first 20 sessions – it should come as little surprise to you that we look at the first 30-40 sessions, often settling on the 20 out of 35 sessions. With more and more speakers submitting two or more sessions, the chances of them getting more than one session in the top 25 is moderately high.
DDD4 was an eye-opener, fewer unique speakers, but those that did submit sessions were submitting one or more. This is a trend that has continued to this day.
DDD5 was originally planned to run on the first day of England’s World Cup match. You would be surprised how hard it is to plan developer events around the variety of sports that are competing for the same audience!
DDD6 “sold out” within 24 hours. DDD is clearly a popular event. After all, it is free, it is held on a Saturday, there are bacon (and veggie) rolls in the morning, there’s a free lunch, 20 full sessions, lunch-time grok talks, swag, discounted books, access to many of the Microsoft DPEs, an evening geek dinner (at your cost!)…it is no wonder that 350 attendees are driven to the registration site within seconds of it going live. We also had SQLBits (thanks!) sponsor coaches from Reading railway station to Microsoft’s TVP campus – feedback for this was good, so we’re hoping to run the same coaches at DDD7.
DDD7 saw us skip the summer DDD slot. We ran DDD 1-6 every 6 months, after DDD6 we looked at the year ahead and felt that so much was going on in the “2008 space”, that postponing DDD until November 22nd was a good idea. It is, after all, post-TechEd and post-PDC. I understand that there is a football match planned for that day…clearly football schedule planners don’t consult us when they are making their plans! We also stretched the time-frame for DDD7 – we have a time-line that we follow for all DDDs, this one we changed to suit the fact we had time on our hands. We opened session submissions earlier than usual and kept it open for longer – this has had the advantage of attacting new speakers on to the submission list. DDD7 is also the first DDD since DDD2 to have girlgeeks (aka ladies, with the exception of Barry, “he ain’t no lady”) on the agenda – Sarah Blow, Kalen Delaney, Annie Matthewman being the first to hold that honour.
Now that I’ve got some statistics rolling, I’ll publish some more over the coming weeks. For example, there is one speaker who has made it on to the agenda for all seven DDDs…more later!
In the meantime, DDD7, November 22nd 2008, Microsoft Thames Valley Park, Reading – on a Saturday, no cost to attend, bacon rolls, free lunch, swag, registration opens “real soon now”.
Follow DDD on Twitter here.