2019 Talk Selection Process

We’re so excited for SeaGL 2019 in LESS THAN TWO WEEKS! Our speakers have accepted, our keynotes are gearing up to deliver some phenomenal insights, and we’re getting the rest of the details worked out in the coming weeks. Today we’d like to share the process we used to manage the Call For Proposals at SeaGL. This document is as much for your edification as it is a record for ourselves for future years!

First, we decided on the CFP announcement. Last year’s was very good, so a few needed alterations were made but it was more or less published as is. That was published 25 June 2019. The CFP was open from 25 June to 14 August. The initial plan was to open til 11 August, but in leaving it a couple more days and scaring up a few more submissions, we had a significantly better return.

The next six weeks were a big rush of tweeting and emailing personal contacts to request a proposal to our CFP. Getting the word out to many area meetups in Seattle and Portland as well as at open source conferences like Libre Planet and ScaLe was also critical in getting new voices. We also held office hours every Wednesday at 12pm Pacific, where we advised folks on their talk proposals. Further, we reached out to local meetup groups in Seattle and Portland.

During this time, we began the process for keynote searching. Within our community, several individuals were nominated, and we reached out one by one until we had a WONDERFUL lineup! Next year, the plan is to begin this keynoting process well before the CFP, rather than during!

Also during that time, the program chair gathered a group of individuals to review, some with experience at SeaGL and some new. This year we had seven members of the review team and coordinated to read and rate all talks. In previous years, this has been on a 3 point scale, summarized as “No, No Opinion, Yes”. This year, we stretched that to 4 points, with not-voting being one the lowest one could “vote”. We will return to the 3 point scale next year, as it was difficult to tell what one had voted on, and how many reviewers had yet to vote on all talks.

After we all did the first pass, we wanted to aggregate first-time speaker and underrepresented speaker information with some SQL in the Open Source Event Management software. Next year, this will be done first. This information is critical for assembling the schedule and speakers. However, after collecting this datadump, we found that these talks had already been highly rated in the first pass!

Total, we had 121 talk submissions. This was lower than previous years by about a third. We’ve heard from sister conferences that their submission numbers have also been low, so we chalk up our numbers to similar effects, as well less outreach than in previous years. 18 of these submissions were by underrepresented speakers (14%), and 13 were by first time speakers (10%). So nearly one quarter of submissions were by underrepresented & first time speakers, which is lower than previous years. We have more work to do in the future on outreach! We’re incredibly proud of our community, and it is extremely important to us to represent this community accurately, including first-time speakers and underrepresented community members.

At this point, we created the list of accepts, waitlists, and declines. Through OSEM (aforementioned Open Source Event Management software), we were able to send the accepts/declines out programmatically, and then handful of speakers with waitlisted talks were emailed manually. As happens, most everyone accepted, a few had to decline, and we bumped a couple folks up on the wait list.

After solidifying and confirming the lineup, the schedule started to come together. Some of the thoughts on this were to keep tracks/similar talks in the same room, and to make sure there was no time slot with only one “type” of talk. It was also important to make sure that there were non-code talks in every time slot.

We then needed to solidify the room numbers we would be using, so that those could be affixed to the online and printed schedule. The printers have a tighter deadline than the conference itself, so getting that worked out in advance was critical. As a result, we’ll have printed personal conference schedules for everyone!

Thanks for reading this bit of conference and CFP geekery!