Improving the check-in process for next Ignite

[Igal Koshevoy]

Congratulations to all on another excellent event. Thanks go out to all
those that organized and volunteered at the event.

Before doing Ignite Portland 3, we need to figure out what to do about
registration, because that turned out to be a roadblock. The decision to
stop mandatory registrations was a very good one. The registration
system's logs showed that we were registering an average of 10 people a
minute, but with 700 people, it'd take nearly two hours to register
everyone.

We can calculate the number of computers we'll need for future events
with a calculation like this: (700 expected event attendees * 40 seconds
per registration) / (30 minutes as the desired amount of time to
register everyone * 60 seconds in a minute) = about 16 registration
computers given that situation. We had far fewer computers than that.

Another problem was that the registration computers were sharing the
Internet connection with everyone else at the event, and were struggling
to get enough bandwidth to talk with the web server providing the
registration program. A solution to this problem would be to setup a
local WiFi mesh just for the registration systems and let them talk to a
local server on the premises, thus avoiding slow Internet access.

However, something else we may want to serious consider next time are
tickets. Folks can register ahead of time and print out tickets
featuring a cryptographic code that uniquely identifies that ticket, to
prevent them from forging or photocopying them. We open the doors early
and let only those with tickets in, scan these to validate the ticket's
legitimacy and mark the people as present. At a later time in the
evening, we open the doors to everyone, regardless of whether they have
a ticket, and keep letting them in until we run out of room. This could
dramatically speed up the registration process and make it a bit more
fair to those that registered ahead of time, but would also add a layer
of technical and social complexity.

Thoughts?

-igal

[Igal Koshevoy]

I'm sorry, I just realized that I sent this to the entire list, not just
the event organizers. I'll take that discussion there, but you're
welcome to make suggestions.

-igal

[Todd Kenefsky]

Thanks Igal.

We're certainly aware of the long lines and we're looking into some
ways to either accommodate more people or at least avoid having any
people from coming out and being turned away. At the first Ignite we
had about 15% fewer people show up than had RSVP'd. This time we had
about 25% MORE people attend than had RSVP'd so the turnout took us
all by surprise.

Thanks to everyone for showing up, being patient and making for such a
fun evening. The crowd was awesome! Feed free to leave feedack and
suggestions for improvement here or on IgnitePorland.com.

[Satyajit Grover]

Hi,

I'm new to this group. I enjoyed IP2 and decided to come here. I look
forward to helping out with future events.

It seemed from Raven's applause survey that the majority of people
heard about IP2 on The Oregonian. So it seems likely that not everyone
from Upcoming showed up and that the surge was due to the article,
unless of course the people reading the article also RSVP'd. That may
affect capacity planning for the next one.

Thanks,
Sam

[Josh Bancroft]

The issue of registration/lines/RSVP is going to be covered, in detail, I'm sure, by the next one. We'll figure something out.

Some of my preliminary thoughts on the matter involved really cheap (like $1 or $2) tickets, or free tickets, or some other way of "pre-registering" people.

One problem we ran into is the way Upcoming.org identifies people who've said "I'm coming". It shows you their Yahoo/Upcoming username, unless they've specifically gone into their profile and added a real name. So even though Upcoming provides a "print guest list" feature, it doesn't have real names on it, for the most part.

And as Dawn points out, we can't NOT use Upcoming, at least in part, because if we don't put it there, someone else will, and if it's not promoted, it won't get many "I'm coming!"s, and people will think it's lame.

A complex problem, to be sure, but a nice one to have. We'll figure out something by the next time! :-)

Josh

[Steve Lewis]

I am not the only person who noticed, but the radio buttons created a
challenge for some folks. The trackpads and trackpad buttons were a
problem for others. Finally, we had too little space for folks to
work at the registration table. (For each machine we would need about
28 inches of table space for elbow room, and having some on low tables
and others up at a greater height would improve accessibility, while
we are wishing.) If folks fail to select a radio button, a better
solution might have been to just pick one for them and process the
registration (one woman snapped "I don't care" when I tried to help
her through that, and I guess that's fair enough). I understand many
of these details may not directly translate to future events, but they
might give a richer understanding of why it took an average of ~42
seconds per person for registration.

I also was poked about pre-registration a few times through the night,
as well as having some sort of name tag printer automatically spit out
a name tag when folks register at the computers. Geez. Don't ask for
much, do they? :)

--
SteveL

[Goddess of Good Ideas]

So two things:

 

1. Let's clarify again why we need to have the secondary on-site registration (not necessarily in this public forum.

 

2. If we find that we do, then use Southwest Airlines approach...so:

 

We feature a link from the Upcoming site...so for those who indicate they're coming...we let them know that to actually seal the deal they have to click and go to a registration site...

 

On that site (a page off of our wiki even) is a form that asks for their (legal) name (if we determine that a legal name is necessary for protection of both IP, the hosts and the attendee). It also asks for other contact info (persumably email) and has an opt-in check box (or boxes) to indicate that they want to receive email on upcoming IP events (or even to opt-in to be alerted of other events like BarCamp, etc...and yet another opt-in to receive info on sponsor products...that is clearly another discussion).

 

Completing that registration form triggers an email to them with a sequential Letter/Number combination like Southwest Airlines (A-1 to 200, B-1 to 200, C-1 to 200, D-1 to 200, and so on...or similar) People print out their email, show up and line up in the order of their print-out (if there is more than on of a Letter/Number...someone cheated). (I have NO idea how hard this would be to develop...)

 

Once we open the doors, we check the Letter/Number off a list as people come thru (no computers needed...just pens) to gauge who and how many actually show). As the line startes moving, late arrivals may find that they've missed their spot in the line. They can either negotiate another spot with someone in the line showing their higher number or they go to the end of the line if they aren't sincere or convincing enough. (It will inspire people to work up compelling stories for why they are late to entertain others in the line.) This allows people to also get somewhat of an idea how early they need to show up in order to be assured their designated place in line.

Those who show up without a printout of their registration email with the sequential number....go "Stand By" (first come-first served) until we reach capacity.

 

Just an idea....

 

Ann

---

From: ignitep...@googlegroups.com [mailto:ignitep...@googlegroups.com] On Behalf Of Josh Bancroft
Sent: Thursday, February 07, 2008 10:23 AM
To: ignitep...@googlegroups.com
Subject: [IgnitePortland] Re: Improving the check-in process for next Ignite

[Josh Bancroft]

I think we can find something a lot simpler that works, but you've got the right idea. :-)

--
Josh Bancroft
http://www.tinyscreenfuls.com
503-334-1889

[Josh Bancroft]

Oh, and for the record, this is going to be covered in the "after action report" meeting the organizers are having next week. So other than suggestion an approach that hasn't already been mentioned here, there's not any need for us to continue this thread to the decision-making point.