Agree on taking responsibility. There are a lot of excuses he could have made: economy, bad PR firm, etc. By taking blame he can now move on to fix things.
Also note, this is a perfect response in a PR crisis - taking responsibility (if not legal responsibility), and not behaving like it's everyone else's fault but your own.
If I buy something, and the vendor screws up, if they give me a bunch of excuses, I am more likely to think they'll screw up again. If the vendor takes personal responsibility, there is less chance that they'll make the same mistake again.
I would have much preferred a $99 conference; I saw the announcement of this year's one and was seriously thinking of attending since I was likely to be geographically nearby this time, but $499 was really a bit much for me. I can drop $99 for a one-off day out, but $499 really seems like you need a company paying your way to it for it to be interesting. I guess this is a big difference between the nature of StackOverflow versus an Oracle Users Conference or similar.
Sounds like one major problem is that they really need Joel at the conferences for a bit of "star power", but he obviously doesn't want to spend half his year flying around to SO conferences. Not easily solved, I guess, but I hope they come up with something.
If it were me, needing to be at every location would have been a major factor contributing to this change.
Part of Joel's shtick for DevDays was that he showed how proper project management and source control (via FogBugz and Kiln) could really help a small team. Kiln was brand-new at the time, and was a big part of the demo, so I thankfully had an excuse to do the European DevDays stops to answer questions and do demos.
And it...well, it starts awesome, if you do it right: you get to see a bunch of stuff you've never seen before, you get to fly first-class if you're lucky, and you meet a bunch of brilliant people. And I think Joel really likes speaking; if you feel comfortable talking in front of a crowd, it's awesome to see all those people wanting to hear what you have to say. So in that sense, doing a pile of small DevDays is absolutely wonderful, in my opinion.
But then there's dealing with the fact that almost no venue was properly set up properly. That supplies we badly needed didn't ever quite manage to get where they needed to be in a timely manner (or, in the case of Europe, clear customs at all). That a major prop we had was going to have to be driven around London in a truck by yours truly. That we had to have SSDs FedEx'd at tremendous cost because the laptops couldn't comfortably run VMware properly, and that sometimes, it didn't really matter anyway, because something would go wrong with the projector, and you'd end up being forced to use large televisions instead. Or a speaker would cancel and be replaced at the last second by someone else, or the WiFi would just not be there even though it was absolutely required for someone's demo, or half a dozen other things. And that's just the stuff that goes wrong at the time; you're leaving out all the stuff that had to be planned beforehand.
This wore on me, and I only had just a few stops. And Joel (and I) did it two years in a row: DevDays first, and the Fog Creek World Tour second. I have to imagine it wore on him, and I think that might go a long way toward explaining why he tried to consolidate the conferences.
But Joel does love travel, and I think DevDays were absolutely wonderful conferences, so that may have absolutely nothing to do with it. I think it may just be as simple as what he wrote in the post: that they tried to make it bigger and more conference-like...which in turn, made it Not Really DevDays, at which point it was competing with things like OSCON and Open Source Bridge. That's a tougher market.
I love short, vigorous conferences. You get great ideas, and go home feeling energized and ready to try a bunch of new stuff you just learned.
Going to longer conferences (like WWDC, etc.) ends up causing brain fatigue after 3rd day of 5 - it's like the Gary Larson "Can I be excused? My brain is full" Far Side comic.
Also, the "hallway track" is what I get the most out of - conferences that optimize for that by offering meals and communal events on-site maximize this.
It's sad, but the price really did kill it for me.
I've only ever been to a conference on my employer's dime once. $500 is a big gamble for me.
(I understand that companies sending devs to conferences seems to be the Done Thing in the States, but I haven't seen it happen too often in Sydney here.)
Every weekend of every month there is a city throwing an exchange. You can go and stay on the cheap as people volunteer to host, and there is always incredible dancing at exchanges.
It's closer to vacation/meetup. Of course ... Swing dancing is inherently a party.
DevDays 2.0 really seemed to be out of touch with the original DevDays concept.
As Joel says in the post, $99 was a good amount for someone to be able to just take a day off work and pay for the conference themselves. Had I not lost my job in mid-2009, I was seriously considering attending the one in Toronto. I could have planned a trip home at the same time, and just stopped over in Toronto rather than immediately grabbing a second - considerably smaller - plane the rest of the way.
Even had there been a DevDays Toronto again this time, it would have been out of my price range to attend, and it would have meant finding accommodations for a night and all that goes with that. I couldn't have justified it to myself at all.
Good on Joel, though, for straight-up admitting to the problems with the new conferences, rather than manufacturing excuses.
Tangential, but a good point was made that some people had booked flights and accommodation already. I wonder if they are still deductible if the event is cancelled?
I have never heard of this event until now. I would love to attend something like this. How/where was this event marketed?
Also, $99 sounds like a great price entry point. Defcon started out on the cheap, and a few years later spawned BlackHat. Sounds like a good model to follow.
Exactly. It's actually great that DevDays was able to charge only $99 AND compensate speakers and workers+! It actually means they one-upped Defcon in that regard.
BlackHat-caliber conferences, as you imply, will come over time as Defcon-caliber conferences explode in popularity.
Again, I know next to nothing about DevDays, and I could be totally off the mark. But it seems to me, like others pointed out, Joel tried to do too much too quickly. Jeff would have had a tough time getting BH off the ground after just two years of DC.
+My assumption here is that by stating Defcon was NOT able to do these things, DevDays was. Sorry if I misunderstood that.
To be fair, venues for conferences don't really have obvious economies of scale: not only do larger conferences often cost more per person, but the quality of the venue and the "rowdiness" (for lack of a better word; arguably the "liability" ;P) of the crowd drastically changes things.
For example, if I wanted to a one-day event with a bunch of reasonably happy low-key developers, it is quite easy. Starting on the small end, with something like 5 people, I can do it in my apartment (and people often do this, for small "meetups"), and if we went up to 10 people, everyone could come to my office.
Even if I wanted something just a touch larger, the venue costs are still going to stay near zero: if I wanted 20 people, I could borrow someone's conference room, and if I needed to go up to 50 people I know someone nearby who would probably loan me their auditorium/theater setup.
With 1-300 people (most conferences), there are lots of choices, from hotels to "community centers": often these are really cheap ways of getting a bunch of people to be able to congregate. If you have a tech conference, you might even be able to get really good rates on something like the eBay "Town Hall".
However, DEF CON is simply /not like this/. DEF CON is a four-day conference set at a fancy hotel in Vegas. This year, the estimated attendance was 10,000 people; and these are not just your "quiet developer type": these people (while largely not gambling, which is important), are drinking constantly and hacking stuff.
Yes: /hacking stuff/; like the elevators, which were half offline for a day because someone was trying to mess with them (I'm certain their other guests were not happy about this); or the lighting/sound system, which someone managed to find out had a default password, so someone central control and would cut the speaker's mic at just the right moments to drive them most crazy.
I kid you not: the point-of-sale terminals at the hotel were offline for a while one day because someone took them offline; there were rumors at the conference that someone had setup a "rogue 3G cell" (to sniff unencrypted connections; apparently people used to do similar things often at previous events), and everyone who attends the conference simply "knows" that you bring enough cash to never have to use the ATMs at the hotel: they are probably skimming cards.
And yet, this event is still managed to be run for only $150/person. Insane, huh? ;P Frankly, I have a ton of respect for how DEF CON is able to scale with this kind of insanity, and not run anything but a horrible loss (and for all I know, they do).
If you're curious around the financials of running a conference (albeit a security conference), I highly recommend the following video from Shmoocon 2011‡
They go over all the associated costs of putting on the conference as well as revenue.
The key difference does seem to be that the jump from $99 to $499 turns DevDays from a conference that you can pay for yourself into one that most people are going to have their employers to pay for.
$499 for a two day conference is pretty cheap but what makes DevDays a hard sell to most managers is that the point of it is that it introduces you to technologies that you're not going be using day to day in your job.
This is a particularly big disappointment for me: Not only was I scheduled to speak at the London event, but there were to be talks on CoffeeScript in all four locations. While CoffeeScript talks have become de rigeur at Ruby/Rails events, Dev Days presented an opportunity to bring the language to a wider developer community. Plus, the talks were going to deal with more advanced topics than the usual "Here's why you should learn CoffeeScript." My talk was cheekily entitled "CoffeeScript: The Only Language You'll Ever Need"; I was planning to talk about tricks and tools for using CoffeeScript in all walks of life.
Of course, my sympathies go out to the Dev Days planners; this must have been a very hard decision for them, and an understandable one. Consider me RSVP'd for Dev Days 2012.
I don't think the multi-city approach is necessary. It seems to be more of a headache than anything else. Stack Overflow has a big enough following that they could hold a 2 day conference in SF, LA, Boston, or Seattle and it would likely sell out. It works for RailsConf, PyCon, CodeConf, LessConf, etc., etc.
Someone should probably do a startup that solves the problem of "overpriced" conferences. This has always been a problem for me when wanting to attend a conferences, first comes the plane ticket, then hotel, and then the conference ticket itself (which sometimes is ridiculously high).
The problem of "overpriced" conferences is that conferences are really expensive to put on. Everything from from the location to the food costs a lot, and every place has it's particular rules about what you are allowed to bring in versus what they force you to use. And then you have to pay speakers...
We are just putting the finishing touches on ProductCamp Utah, and with no profit and no paid speakers, it is still extremely expensive.
There's a large number of experienced conference runners and planners who will disagree, and collectively have been running inexpensive but good conferences for DECADES.
They're called "SMOFs." They run science fiction conventions. (Not "Star Trek" conventions but "science fiction" conventions - books.) And the conventions are not for profit membership organizations typically with hundreds or thousands of members attending over a weekend at a cost of a few tens of dollars each.
The SMOFs are generally willing to help others run conferences by giving them advice, doing (paid) consulting work, or even fully (paid) managing conferences. A great example is Expotech, who ran MacHack.
I was co-chair of Python 2006 & 2007, and in preparation I went to SMOFcon. SF con issues have some relevance to tech cons, but only partially so:
* SF cons don't usually provide food for all attendees, which is the big cost driver for tech cons. For PyCon 2011, catering was 54% of the budget.
* Tech cons mostly don't have to worry about people in costumes, keeping under-18s out of certain events, or as much rowdy late-night behaviour (e.g. room parties).
* Tech cons really, really have to worry about the WiFi network. Recording every presentation may also impose more A/V requirements.
Going to SMOFcon was somewhat useful to me, but the differences are significant.
"SF cons don't usually provide food for all attendees, which is the big cost driver for tech cons. For PyCon 2011, catering was 54% of the budget."
Isn't the catering cost a catch-22? The only reason to expect food at a tech conference is because you (or more likely your employer) is paying out the wazoo for a conference ticket to begin with.
I'd be much more likely to go to a tech conference that was $99 and gave me no food than go to a tech conference that cost $500+ and was fully catered. I can take care of my own food needs for far less than $400 over the space of a couple of days.
That could certainly be done, though I didn't convince myself it would work for PyCon. I suspect catering is definitely optional for small events or unstructured things like BarCamps, but for large events it becomes a necessity.
If the conference has 500 or 1000 attendees, the hotel restaurant probably can't handle that many people, so there have to be several restaurants or a mall food court nearby, enough of them so that they don't get overwhelmed by the attendees. That rules out venues that aren't downtown, because suburbs may not have that density of restaurants within walking distance, but such central venues have more expensive room costs.
Another tradeoff: The lunch break would probably have to be 2 hours, because 1 hour is a bit tight for walking time + waiting for service + eating, which means less time for conference programming. Attendees like lots of programming, the more the better; in fact, PyCon attendees asked for breakfast as well as lunch, so it's easier to attend an early-morning event.
And the monetary risk of running an event is large, so like Hollywood film producers, you quickly stop wanting to tinker and try something new; it's safer to stick with a format that's worked in the past.
Oh yeah, and if you're paying speakers, You're Doing It Wrong. The speakers should be drawn from the attendees/members.
If a speaker is a "guest of honor" (say a major keynote) you should comp them their conference membership, and possibly cover their hotel room and travel (which you'd arrange). That's it.
No paid speakers. No "sponsors," either. Just members. Hundreds of conventions with a great many thousands upon thousands (potentially millions) of members make this work every year in sci-fi and the areas that have branched off from it like anime and comics.
An unconference (like us) can get away with that. A "real" conference often cannot. While certainly not every speaker should be paid, having known names for keynotes is important. Even just comping travel gets very pricey for small conferences.
Also remember in the scifi, anime, and comic con worlds, many of those people who offer to speak for free are doing it as a marketing expense. Nothing wrong with that, but it should be recognized because, depending on your conference, it can alter the flavor. We've been very explicit: no direct marketing.
Have you been to a BarCamp? They're usually free to attend, they happen all around the world (so you hopefully don't have to travel too far) and the quality of the sessions can be incredibly high.
They can be incredibly high. I've found our local barcamps to have been a mixed bag, and it really depended on the quality of the people who showed up that day. They used to run it where everyone pitched during the morning, then self organized around the pitched sessions. That made the sessions totally up in the air - you had no idea when you woke up if it was even worth showing up. Later ones set up early pitching/voting on the web, which seemed to help some, but because there was no ticket cost, the organizers were still on the hook trying to determine food cost and such in advance, when just a little bad weather might keep away loads of people.
The area ncdevcon.com (next week in Raleigh) started charging for tickets this year - a nominal $60 for 2 days - precisely because when it was free, people wouldn't show up on the day because they had no real investment.
For indieconf.com, we've charged - both for last year's even and this year's event - and are posting speakers/sessions well in advance so people know what they're getting.
There's room for all styles of events, certainly, and perhaps I have bit more invested emotionally than others, being an organizer myself, but the barcamp self-organizing doesn't always produce great results. This may also have something to do with geography as well, but I don't think anyone's done an actual study of barcamp attendee satisfaction around the globe. Maybe that's something you could add to lanyrd :)
Interesting. I wonder why they didn't just choose to promote the events more aggressively, etc. -- surely Spolsky could have gotten more registrations if he had tried.
I think there is a market for a $499 conference, but perhaps it needs to have a better reputation among management first.
I felt like it was a no-brainer to go to the 1-day $99 DevDays in SF. When I saw the price and duration of this year's event I didn't event want to ask.
I hope he was, but I'm not certain he was. In any case, I can assure everyone out there who doesn't vacation in the Hamptons that a jump from $99 to $499 is an absolutely huge difference, both practically and psychologically.
I'm not sure how he couldn't have predicted this outcome.
I was referring to him calling the price hike a "modest" increase. I read it as him being intentionally inappropriately flippant about the thing that killed the conference (in a mildly humourous and self-effacing manor).
I would be interested in seeing an online version of any of the presentations. Pretty please ?