Reading Time: 11 minutes

This year I helped organize the inaugural PromptConf, a one-day tech conference showcasing the joy of computers.

We divvied up responsibilities among the six co-organizers. I’m writing about some parts in which I invested time and energy; views are my own, and do not necessarily reflect the whole organizing team’s positions.

I have a soapbox when it comes to tech conferences: I think that most conferences do not put enough time, effort, and thought into their programs, and that’s why many conferences either kind of suck, or are only appreciated for their afterparties and/or “networking”.

Planning an engaging conference program isn’t about getting one big decision right. Instead, it’s about getting lots of little decisions right. Let’s talk about some of those.

Choosing a Speaker Lineup

A good speaker lineup has a heavy percentage of speakers with an engaging speaking style.

How heavy? Depends on the conference. Ten minute talks are relatively low risk. Put two speakers with poor stage presence back to back for an hour each, though, and you’re going to lose your audience for the rest of the day. Speaking experience plays some role in this, but plenty of people skate by for years booking gigs without learning to speak.

Before you blame “cherry picking diverse speakers” for this, know that most of these skaters-by are people who never have to learn because everyone just assumes they know what they’re doing. That doesn’t describe “diverse” speakers. I’ll let you put two and two together here.

That topic brings me to the second thing. To me, a good speaker lineup represents the diversity of the tech field: gender, race, sexuality, technical specialty, experience level.

I do not recommend cherry-picking a diverse speaker lineup. I recommend sourcing a diverse applicant pool and then picking the talks that best suit the conference, knowing that the selected group will likely reflect the diversity of the applicant pool. We talked about this quite a bit in the previous post on conference organizing.

Sourcing a diverse applicant pool is an advanced maneuver. A homogeneous organizing team does not have the network to pull that off. Keep that in mind when building an organizing team.

Selecting Talks that Suit the Conference

To select the talks that best suit the conference, it’s important to know what the conference is trying to be.

This is the part I think a lot of conferences are missing: a clear vision of what guests and speakers should get out of the event.

At PromptConf, we wanted to focus on the joy of computing. The talks leaned toward how to do cool stuff with computers. Two reasons for this.

First, lots of conferences already cover practical computing: how to use Spring Boot, how to get started with Neo4j, Kubernetes workshops. We wanted to create a space for exciting and inspiring implementations that aren’t backed by a quota. You think Google would let a year go by without some quota of Kubernetes talks? No. But no big company has a quota for annual talks about sweating the small stuff in software design.

Second, we wanted folks from underrepresented groups to get a chance to talk about computing for once. We talked about this in the last post, but often women, black people, latino people, queer people, et cetera get invited to conferences to give talks that highlight what it’s like to be in their underrepresented group. We don’t get to just talk about computers the way any straight white guy gets to do. PromptConf wanted to make a space for that.

One side effect of having a clear vision: there will be talk submissions that don’t fit the vision. It won’t make sense to put those talks in the lineup, even though they would be objectively “good” talks on their own.

Mentoring Speakers

I mentioned that a good speaker lineup has a heavy percentage of speakers with an engaging speaking style. This does not mean all those speakers have to have a ton of speaking tenure. With some practice, lots of folks can give an engaging talk.

That’s where mentoring comes in. PromptConf offered each speaker a mentoring session prior to the conference where one of a few organizers who does regular speaking would review their talk.

This has two effects.

 1. The mentoring session gives speakers another set of eyes on the structure of their talk.

Like code reviews, talk reviews from other people give speakers a chance to collect perspectives they might not have otherwise considered. “You lost me in this portion” or “I would have liked to hear more about that part” can shed light for speakers on what will keep an audience engaged.

Notice the focus here on the content of the piece, not the delivery. Delivery, in my humble view, comes naturally as a speaker gains confidence with their material and with speaking.

Effective delivery also has almost nothing to do with some of the popular metrics for “good speaking.” A common bogus metric: use of “um” and “uh.” You know who notices when speakers do that? Speaking coaches. Not audiences, unless it’s egregious enough to signal a total lack of content preparation.

You think I’m making this up. I’m not. In universities, it’s common for lecturers to undergo review of their class by a plurality of more senior lecturers. The lecturer being observed sometimes self-evaluates a video recording of their lecture before the reviewers chime in. In a three hour lecture, just about anybody is going to say “um” a few times. Inevitably the observed lecturer self-flagellates over their use of the dreaded “um” and “uh.” Then the more senior lecturers comment that they didn’t notice that at all. These are senior lecturers—more attuned to speaking foibles than some random conference guest.

If the content is tight and the speaker is prepared, the delivery will most likely be good enough.

 2. The mentoring session forces speakers to prepare their talks ahead of time.

This is critical for two big reasons. The first one we already talked about—speaking style, particularly for new speakers, improves alongside the speaker’s confidence with the material.

The second one is that effective editing requires distance. The original writer—the person with the clearest personal vision of what this talk is supposed to be—returns to the talk after some time, detached from the choices they made in the first iteration. Suddenly, they can clearly see where things might be removed or rearranged. These changes make the talk much, much stronger.

This is why it grinds my gears when speakers slap together a brand new talk the night before because they figure they know what they’re doing and can wing it. Newsflash: y’all can’t. You might think you fooled us, but we noticed, and we’re all on Twitter now instead of listening to you. That one person who came up and said they liked your talk does not speak for most of the audience. Speakers: don’t accept a spot that could have gone to another speaker and then do this. It’s disrespectful to the organizers, the other applicants, and the guests. The exact people you want to fool will catch you, and you might lose a future booking over it.

Speaking Order

Okay, so you picked your talks and your speakers are dutifully getting ready. These talks have to happen in an order. I keep some things in mind for setting the order.

1. The first talk has to give energy to the crowd. This means it needs to be both about the audience in some way, and by someone with stage presence. If you ignore every point on this list except one, this is the one to remember. I talked about this one more over here, so feel free to take a look at that.

2. Put talks together into batches. Maybe they’re on similar topics, or maybe they share a similar theme or perspective. Batching makes each talk more memorable for the same reason it’s easier to remember phone numbers in (3) 3 – 4 batches than as a string of ten numbers. Batches also give speakers opportunities to play off each other, if they wish. Finally, guests can plan their quiet time or other obligations in accordance with the batches they need the most. I like to try to give each batch a name. I don’t always succeed, but I try.

3. Match the order of batches to the guests’ (average) energy levels. People’s energy levels, on average, go:

• highest in midmorning (after eating breakfast but before a full day of talks),
• second highest right after lunch (after eating, but before the sluggishness that follows a heavy meal),
• then the beginning of the day (before eating, as the coffee is going in),
• the end of the day (final sprint mindset)
• midafternoon (peak lunch sluggishness)

Talks that require intense attention (like highly involved technical explanations) go in high energy slots. You can tell you did this right on n day of a conference if the crowd is about the same size at the beginning of n + 1 day as it was at the beginning of n day.

4. Match the order of talks in each batch to the guests’ (average) attention levels. Within a batch, audience is most receptive at the beginning. This goes down for each successive talk. So “controversial” stuff (or stuff that is likely to challenge people’s existing assumptions) goes at the beginning of a batch.

The exception: if a talk is going to be emotionally arousing, it goes at the end of a batch. People will need a break before being able to think through another talk. So if you choose to include personal “my journey”-type talks in your program, this is where those go.

You’ll notice that this puts a limit on how many “my journey” talks you can have. That’s also a critical affordance for managing the audience’s energy. We talk about that in more detail over here.

A set of talks will deliver a much more impactful experience in a thoughtful order than the same set of talks will deliver in a random order.

Conclusion

I have a soapbox when it comes to tech conferences: I think that most conferences do not put enough time, effort, and thought into their programs, and that’s why many conferences either kind of suck, or are only appreciated for their afterparties and/or “networking”.

Planning an engaging conference program isn’t about getting one big decision right. Instead, it’s about getting lots of little decisions right. They include:

1. Choosing a heavy proportion of engaging speakers
2. Selecting talks that suit the conference
3. Mentoring Speakers and/or otherwise requiring speakers to prepare
4. Arranging the talks in a thoughtful order

These decisions, in my view, should not be made at random or by default. They set excellent conferences apart from the ones that feel like six average tech meetups strung together.

If you liked this piece, you might also like some of these other things:

The other day Gabby Hempel (thoughtful and funny infosec person/pen tester, totally follow her!) asked on Twitter for resources to help her prepare to guest lecture in a Windows course. I went through my blog posts and stuck a ‘guest-lecturing’ tag on the five posts that I thought would be most worth her time. You can check that tag out, too.

If you’re looking to write a book about programming, I tried to tease out exactly what (to me) makes a programming text good.

You can also watch me yell at computers (just kidding, I have a pretty long fuse…most of the time 😉)