Mastodawn

@mkennedy I believe it “sold out” at 2700, so we didn’t completely max out the venue. I’m not an organizer, but the limits have more to do with precommitments from the conference than physical size; i.e. we project we can get 2300 in the door for sure, so we won’t buy food or food-service-worker-time for more than 2700. Stuff like that.

Making these numbers too big (as they have been at a few points in pycon’s past) can be financially ruinous to the PSF so it’s always a little conservative.

@glyph Thanks Glyph, that’s good to know. Conference organizing scares me to just think about it…

@mkennedy at one point one of the organizers explained to me *just* the logistics of getting coffee for the sprints and I felt like my soul left my body for the astral plane

@mkennedy on a macro level though the game is a sort of “price is right” where the stakes are the financial future of the programming language. before you get sponsors, you sign in blood that you will get X people into hotels and Y people into the venue, and if you guess Y too low the sponsors get mad and don’t come back because you’re not getting them sufficient brand equity, and if you guess X too high the hotels and convention center clean out the foundation’s bank account

@mkennedy I think we could have done 3.5k easy but there’s no way I’m signing *my* name to that, and therefore cannot blame the organizers even a little bit for being reserved in the estimations of our growth

@glyph I agree, it’s tricky for sure. I was just thinking there must be another $300k-$400k on the table for the PSF easy. Just wait til it sells out, then open it for more, only promise the more conservative number to hotels, venue, sponsors, etc.

@mkennedy oh that’s the problem though: you promise that number to hotels, but by the time the conference is rolling around, they’ve already sold the other capacity, you *can’t* increase the cap much. You want to “open it for more” but there’s already a physics convention and a gynastics competition that the conference center & hotels sold during the same dates

@glyph True. Just don’t promise the second round of folks hotel discounts. They could find them *somewhere* in Pittsburgh even if they had to transit there. That could be a bonus of early bird registrants. :)

@mkennedy you might get away with “no hotel discounts” but the nightmare of “you can only buy the kind of badge that doesn’t get you the conference-provided lunch” is intractable. I went to one conference that tried this and it was … well let’s say “logistically difficult”

@glyph Yeah, I wouldn’t be for that either.

Ben Berry May 24, 2024

@mkennedy @glyph I think Glyph is right about this, especially since the "we're sold out" note came days before the conference.
It's different at 10x scale, but I know the caterer for PyCascades last year said "no changes within 1 week" so that was when we set our hard cap on tickets.

George Hickman May 24, 2024

@mkennedy @glyph this might have worked for Pittsburgh this year. But you never know when an unexpected boy band tour will roll around, and suddenly, your local capacity will be gone. Booking things far out, like PyCon US does, is a huge benefit here.

Source: PyCon UK has seen some stuff 😂

@mkennedy also, don’t quote me on this one, but I’m *pretty* sure that the economics of it are that you lose money on each attendee, so as you increase attendance you then need to go back to your sponsors for increased commitments, which is also itself an awkward conversation.

@glyph @mkennedy Fact check: This year it was a volleyball tournament ;)

@lorenipsum @glyph @mkennedy and a state treasurer's convention, apparently they travel to non-capital cities for a change of scenery?

@glyph @mkennedy I am very into their logo

Thomas Wouters May 24, 2024

@lorenipsum @glyph Be honest, you had @georgically mock up a fake logo, didn't you?

@Yhg1s @glyph @georgically I totally would have, except I refuse to ask Georgi to do anything for at least a month or two 😇

@glyph @mkennedy exactly that.

Also, you can't magic up 600 seats of extra food and the staff to prepare and serve it at a week's notice. Or staff to set out seats. Or bigger rooms.

Big conferences are massively inelastic.

Cassandra is only carbon now May 24, 2024

@chrisjrn @glyph @mkennedy Not only inelastic, but inherently risk-averse. If you plan for more attendees than actually show up, such that not enough rooms get booked on the group code, you can get stuck paying for the entire block of hotel rooms.

@xgranade @glyph @mkennedy Yup. Indeed, that was a huge part of why it took weeks to announce 2020's cancellation, despite it becoming more clear that COVID was a risk: making sure the Foundation wasn't exposed to literally millions in penalty fees.

Read these blog posts with that in mind: https://pycon.blogspot.com/2020/03/march-6-update-on-covid-19.html https://pycon.blogspot.com/2020/03/march-12-update-on-covid-19.html

March 6 Update on COVID-19

PyCon continues to closely monitor the Coronavirus (also known as COVID-19) situation. As of March 6, PyCon 2020 in Pittsburgh, Pennsylvan...

Chris (so far)May 24, 2024

@chrisjrn @glyph @mkennedy To put it in perspective Perl’s conference numbers for 2014 was < 500 people … so you’re effectively taking about adding on a “normal sized” conference with a week’s notice.

Hugo van Kemenade May 24, 2024

@mkennedy @glyph It sold out just before the event opened, with a spike of some 300 tickets
https://fosstodon.org/@pycon/112445571644276379

PyCon US (@[email protected])

⚠️ UPDATE⚠️ : #PyConUS 2024 is SOLD OUT. We had an unexpected spike in registrations just before opening the event today, and we've had to close sales of new conference passes. Some Tutorials are still available, prepaid voucher codes will still work, and if you didn't snag a ticket, online passes are still available at us.pycon.org. (Contact [email protected] if you need help.) We're thrilled to welcome over 2700 attendees to PyCon US this week! http://us.pycon.org

Fosstodon

@hugovk @glyph Thanks!

@hugovk @mkennedy @glyph

One of my theories about the late bump is that there are more population centers in easy-ish driving distance from Pittsburgh [eta: vs SLC], which means more people who didn’t have to commit in advance by buying plane tickets, so they could leave buying conference tickets until later in the game.

Tim Schilling May 24, 2024

@lorenipsum For what it's worth, I almost bought a ticket last minute because of this exact reason (a 7 hour drive away and the weekend opened up).

@CodenameTim Makes sense!

Jon Banafato May 24, 2024

@lorenipsum As a point of comparison and contrast, I'm pretty used to seeing last-minute bumps in registration for regional conferences. I wonder if there has been demand for those late registrations that's less visible for a conference that regularly sold out (some times two months prior) pre-pandemic. Data about post-sellout ticket demand is probably spotty at best, but I'm sure some of us could dig up graphs of registration over time for regional events.

@jonafato I would guess that regional conferences also have the effect of fewer airplane tickets required in general? E.g. I'm assuming a much higher share of people drive to PyOhio or PyTexas than to PyCon US.

Jon Banafato May 24, 2024

@lorenipsum Yes, that's definitely a factor (and often by design, i.e. many regional events _want_ to focus on their local communities). I'm just curious if there has also been some late demand for PyCon US that's been impossible to act on because the conference is already sold out. (Tried to check a few 2019 and earlier blog posts, and it does seem that six to eight weeks out is a common enough sellout time frame.)

@jonafato Oh yeah, I'd for sure bet there was late demand over the sell out cap. We probably also increased late demand by not selling out the last few years, since there wasn't a 'cost' to registering late in '22 or '23, so there was less pressure to make sure to buy a ticket early. (If that all makes sense -- I just got home around midnight yesterday and my brain is still a little soupy.)

@jonafato Also the more times I type "sell out", the more it feels like I'm a teen insulting an alt-rock band in the nineties.

Jeff Triplett May 24, 2024

@lorenipsum @jonafato Don't make Eddie Vedder and Pearl Jam sad...

I think of my favorite, local cover band https://selloutlive.com

Rock-HipHop-Cover-Pop Dance Band

Musicians in costume playing high energy hits from the past and present with a propensity to get the dance floor pumping. Not your daddy's dance band.

Sellout

Jon Banafato May 24, 2024

@lorenipsum Yep, used to be that the day ticket sales opened, I'd tell my teams to buy tickets before they forget and it's too late. Glad to be able to point to a post and say "no, really, it's going to sell out" again for 2025.

alexwlchan May 24, 2024

@mkennedy @glyph One of the tricky parts is that the venue you’d choose at X people may not be the one you choose at X+500, but swapping is basically impossible.

Even with lots of money, you can’t conjure up more toilets, hallway space, power sockets, Internet bandwidth, and so on – the physical limits of the space often kick in before you run out of room to put people.

@glyph 😅😅😅 …yep

jacobian May 24, 2024

@glyph @mkennedy the huge factor is of course that, uh, something huge happened in 2020 that changed conference attendance. Possibly permanent? Definitely long-term. Many conferences never recovered. Projecting attendance using pre-2020 numbers is ill-advised.

That said, I'd guess next year will likely have a higher cap, and I wouldn't be surprised to see '25-‘26, returning to California, to have even higher attendance than that.

@jacob @glyph COVID was primarily why I was asking. I’m trying to assess how the conference scene for Python has changed from 2019 to 2024 and beyond.

Should be interesting and here’s to hoping it keeps growing back and beyond.

@mkennedy @jacob COVID was obviously an absolute knockout blow to event organizing, but the ongoing influence of high interest rates shouldn’t be underestimated either. Much as I think that we *could* do better, I am thrilled to have broken through 2500 in 2024, it’s huge progress in rebuilding.

@mkennedy @jacob (Just for any readers who may not be familiar: high interest rates = money is expensive for corporations & startups = promotional budgets are slashed = much harder to raise money for “optional” things like conference sponsorships)

@glyph @jacob I imagine sponsorships are the primary profit center for pycon too.

@mkennedy @glyph @jacob
Yes. Historically, the "individual" ticket price is the break-even point.

And fwiw, PSF had Pittsburgh booked for 2020, and iirc we issued 3500 tickets just like Cleveland did. I think the "sellout" point was entirely down to setting a "reasonable" amount of growth vs 2023 and then getting a sudden spike in demand last week that met that "reasonable" amount quickly. No venue-related reasons we couldn't sell out at 3500 next year if we want that.

@mkennedy @glyph @jacob (and happy to talk outside this canoe for reasons why we would or would not want to do that, as it's probably best I don't speak for staff or the rest of the board)

@chrisjrn @mkennedy @glyph @jacob Yeah, and 2700 was already ~500 > than 2023 in-person. Plus it’s always hard to project what attendance will look like in a new location.

One other factor I haven’t seen in this thread is staff/volunteer capacity. More attendees = more emails to pycon-reg, (regrettably, statistically) more CoC reports, more needs for help at the reg desk, etc etc. That’s v hard to scale up on short notice (or even medium notice).

@lorenipsum @mkennedy @glyph @jacob

This is one reason why pre-pandemic, PyCon US was _permanently_ capped at 3500 attendees. It's a lot easier to bid and budget for a known number of attendees: the increase in demand does not impact the ability to deliver the weekend.

Of course there are other problems: turning people away is antithetical to the ethos of our community, and we need to do a better job of giving prospective attendees somewhere to go.

@chrisjrn @mkennedy @glyph @jacob

>turning people away is antithetical to the ethos of our community
Totally agree

>and we need to do a better job of giving prospective attendees somewhere to go.
Can you say more?

@lorenipsum @mkennedy @glyph @jacob

More? My $0.02 (entirely personal opinion, not board-vetted): We need to do a better job of supporting, funding, and promoting regional events. How and how much do we do that? That's for the board and staff to decide.

@chrisjrn @mkennedy @glyph @jacob that answered my question:) somewhere to go” could be a lot of things, regional cons make sense!

Mario Munoz (has moved)May 24, 2024

@lorenipsum @chrisjrn @mkennedy @glyph @jacob I also believe that as the "online" experience grows, that's also an avenue. Definitely not a substitute for the energy of live interactions...

But being on the "online" side of things this time around due to ... 🩹... reasons... I was able to participate in a slice of it by attending online, volunteering, and even having a chance to give a "lightning talk" of sorts.

There's "virtually" no ceiling there...

@pythonbynight ISWYDT ;)
I'm glad you were able to participate a bit, you were sorely missed in Pittsburgh! Looking forward to next year :)

Mario Munoz (has moved)May 24, 2024

@lorenipsum I'm a sucker for wordplay 😅

@pythonbynight One of many reasons we are friends:-)

(i apologize for not including a pun in this reply, i am so tired my pun generator is on the fritz)

@pythonbynight @lorenipsum @chrisjrn @mkennedy @jacob it would be great to find ways to better integrate the online experience into the live conference. On my “if money were no object” feedback list is to put a videoconference setup in every openspace room so virtual attendees could access the hallway track more meaningfully

Mario Munoz (has moved)May 24, 2024

@glyph @lorenipsum @chrisjrn @mkennedy @jacob For what it's worth, that is a recommendation I made to the online team during our after conference touchbase. I think this is a great idea.

@pythonbynight @glyph @chrisjrn @mkennedy @jacob FWIW money is really really an object with respect to AV equipment & services

@pythonbynight @glyph @chrisjrn @mkennedy @jacob (not to be a wet blanket! i love the idea; just sharing intel from the dollars side of things)

@lorenipsum @pythonbynight @chrisjrn @mkennedy @jacob This is the magic of being an “idea guy”, I don’t have to deal with the consequences of my (brilliant, wonderful, you should definitely do them) ideas