Not long after ticket sales closed for SRCCON 2015, we added an item to the OpenNews planning document: rethink ticketing for 2016. The first SRCCON sold out in minutes, which caught us off-guard. When our second conference sold out in seconds, it told us a couple things: year one wasn’t a fluke, and we needed to change the way we handled sales. It didn’t feel right for a poorly timed meeting—much less the speed of an internet connection—to be the factor that kept someone from getting a ticket to the next SRCCON.

We talked with other event organizers about other ticketing options, and settled on lottery-style distribution as the method that met the most of our goals. Here’s how we ran the SRCCON 2016 ticket lottery and what we learned.

Ticket numbers

We feel strongly that SRCCON works best as a medium-sized conference: large enough to invite a diverse mix of attendees with complementary backgrounds, but small enough that people feel like they can engage with each other in every session–and fit into one big room for dinner. This prevents us from simply selling as many tickets as possible—the atmosphere and session styles we want to encourage at SRCCON don’t scale indefinitely. Still, we’ve increased in size each year of the conference, from 171 in 2014 to 226 in 2015, and 271 this year in Portland.

The number of tickets we’re able to put on sale is lower than our total attendance cap, though. We have to set aside tickets for:

• facilitators with accepted sessions (this year, we had 57 sessions and a total of 96 facilitators)
• staff and volunteers
• conference sponsors
• participants in our [scholarship program] to help people from smaller news organizations and underrepresented communities attend

And this year, as in years past, we held back a small group of tickets to help ensure that the conversations that happen at our conference include voices from across our entire community, reflecting diversity of gender, race and ethnicity, organizational size, and professional background.

After accounting for those groups of tickets, this spring we were able to put 90 tickets into our first lottery drawing.

Before we ran the lottery

Our goal is to run an inclusive conference, and both open sales and lottery drawings put at least some measure of an event’s diversity beyond its organizers’ control. We try to mitigate those effects by doing the widest possible outreach before tickets ever go on sale, so that we’re drawing from a group of people that really does reflect the entire community we serve.

We also believe that SRCCON works best with a good mix of veteran and first-time attendees, and considered a number of ways to make that happen. After several conversations, we determined that we’d feel good about an attendee list that was at least 38% new attendees (yep, we rounded off the Golden Ratio). This particular goal informed the script we wrote to pull random names from the pool. With first-time status as a column in our CSV of lottery entrants, the script could make sure that new attendees, as a group, had no less than 38% of the chances to be drawn. But as it turned out, our initial pool was 81% first-timers, so we didn’t have to use this feature at all.

We also chose to make tickets non-transferable—if your name was drawn in the lottery, the associated ticket would be for you and you alone. Our hope here was to avoid a couple possible (and related) scenarios: purchasing a ticket for an organization, and only then figuring out if someone actually could use it; and larger organizations asking everyone to enter, maximizing the chances of winning tickets and intending to transfer them to the people who really wanted to attend.

We talked through all these issues before we ever opened up a Screendoor form and started accepting applications for the SRCCON lottery. The form stayed open for a week, from May 18 to May 24, and 234 people submitted their names.

Running the lottery

We ran the lottery script later in the day on May 24, and sent out notifications to everyone–selected and not selected—immediately afterward. The 90 lottery winners had until May 27 to purchase their tickets (with some extra grace for folks who were on vacation or otherwise out of contact). From that group, 71 people purchased their tickets, with the rest declining, discovering they were unable to attend, or otherwise changing their mind. That left us with something we certainly hadn’t experienced in our first two years of running SRCCON: leftover tickets.

We chose to release those tickets in a second-chance lottery, and gave everyone who missed out during the first round the opportunity to put their name in again. From that pool of 139 people, we drew 23 tickets for this second round (the 19 left over from round one, plus a few tickets freed up from sponsorships), and distributed notifications in the same way as the first round. We had a small number of second-chance lottery winners turn down their tickets as well.

What we learned

The quick sellouts before the first two SRCCONs told us there was plenty of demand to attend the conference, but really couldn’t tell us much about scale. Collecting names for a lottery helped clarify a minimum number of people who wanted to attend, although we recognize that it doesn’t capture potential attendees who didn’t enter their names because they knew they’d be unavailable, avoided the lottery to give new people a chance, didn’t hear about the conference, and so on. Still, we have a much more realistic sense of the interest in an event like SRCCON now.

We also learned that people are much more likely to enter their names in a ticket lottery before they’re absolutely sure they can attend an event. During the first two years of SRCCON, open sales meant that attendees had to buy their tickets immediately, and we had relatively few people cancel their purchases. The lottery, however, made it possible to enter first and decide (or ask for permission) later, and we saw about 20 percent of both sets of lottery winners throw their tickets back into the pool. This gives us something to think about—we may decide, for instance, that attrition means we don’t need to set aside the same number of holdback tickets in advance.

But honestly, this was one of the only quirks we ran into with our ticketing changes. The lottery ran incredibly smoothly. We published an explanation of the system in advance, and although we were worried that not everyone would love the changes, we saw hardly any negative feedback. We’ll continue to talk through the process as we get closer to SRCCON 2017, though, and we’d love to hear from you as we try to make this year even better.