Conference Notes

On writing my talk for dconstruct…

A few people have asked me to put up my slides for the dconstruct talk. However, I’m going to be doing a variant on the talk at Web 2.0 Expo Berlin in a few weeks so I thought I’d let it relax a bit in my head and see if there’s anything I’d like to adjust before I put it out into the world as it stands. If you need a version to persuade your boss of something in the meantime, let me know and I’ll send you a copy. Otherwise I’ll post one up in early November.

A week after dconstruct and I find myself jet-lagged to hell, awake and mostly alert at 5.30am PST in a Marriott hotel in Redmond. I’m here in Microsoft heartland for a Social Computing Symposium where I’m going to be talking a bit more about Fire Eagle and then I’m heading down to San Francisco for a bunch of meetings and stuff. I quite like travelling. I like the change of context. I like meeting people and the almost total solitude that you get in alternating chunks.

One of the things I like most about travelling is when you travel east-to-west and the jet-lag kicks in. The evenings sort of suck–I pass out around nine or ten for the first couple of evenings–but the mornings are pretty great. You find yourself glancing a the clock when you wake up and it’s 4am. You lie in bed getting a lie-in and when you’re bored it’s still not light. With nothing else to do, these first few hours are wonderfully relaxing and a great time to catch up on things that have rather got on top of you. I’m spending my pre-dawn hours today catching up on e-mail, posting to and generally catching up with stuff. Makes me feel so productive.

I don’t always feel so terribly productive. A little over a week ago I was also waking up pretty early and working, but that was mostly because of my talk at dconstruct. And while finally delivering the talk was tremendously enjoyable and invigorating, writing it was anything but. It was instead a bloody horrible experience. And sitting here in bed–it still dark outside–it occurs to me that maybe it’s an experience I should bring out into the open, both to get it out of my head and to give other people some perspective on what it’s like to try and drag a talk together when you’re not actually a terribly extraverted person. Maybe it’ll help someone. Who knows.

Let me take you back to the middle of August – a full three weeks before dconstruct. At that point, I’d already been having real trouble getting my talk together and it had started to stress me out. I’d been thinking around the subject for a month or so, trying to scrabble my thoughts together, but it hadn’t got very far. So I came up with a solution – I’d take off three days to bring it all together. The three days bled into the weekend, and the weekend into a bank holiday. Six days to work on a talk! What bliss! What ecstacy! That’s going to be fine! What could go wrong?

It was a disaster. I woke up in the morning, stared at a screen for a while, wandered around my flat, wrote things on post-it notes and stuck them to every surface in my house. I stared at myself in the mirror. I bought lots of dental floss and had a really good go at my back teeth. I paced a lot. I did occasional spats of press-ups to try and get the frustration out of my system and ended up just feeling a bit sore. My brain refused to focus on doing the talk itself, just on worrying about writing it.

I found myself doing absolutely anything I could possibly do that would be in any way useful in my life instead as a way to legitimate my lack of progress. I bought light-bulbs. I sorted out all my old VHS videos and got rid of them via freecycle. I did inordinate amounts of washing. I went through all my books and took the ones I didn’t want to a charity shop. I emptied a cupboard. I even considered hoovering, which should give you a sense of my state of mind. Day Three came and went and nothing was flowing.

I’d managed to phrase my talk as a follow-on to Native to a Web of Data, which is probably the talk of which I’m most proud. I felt that the pressure was on to match it with a sequel. Also about forty of the people I respect most in the industry had decided that they were going to come down to Brighton for the day and I was going to be performing in front of all of them. Worse still, somehow I had landed the responsibility of being the last on stage for the day in front of six hundred people or so. To put it mildly, I was terrified.
Another three days passed and I’d got nowhere substantial. My break now over and back in work, I started trying to get it together in the evenings after finishing my normal day’s workload. On a couple of occasions, I worked in the office from 10am until 1 or 2am the following day to try and get something out. Mostly this was unsuccessful.

If any of you guys out there think that this just happened to me for the first time with this talk–by the way–I need to disabuse you of it. Most of my talks are this much pain, as most of my friends can attest. I get enormous stage-fright. Standing up in front of people to give talks of this scale (several hundred people) scares the living crap out of me and always has done. In the end though, normally it’s worth it. I battle with talks to such an extent that they’re normally always pretty decent, they force me to think through and systematise things that have been in my head for a while, and the reception is normally pretty good. But god, they’re a nightmare to write.

And speaking of my friends, there are a few of them I need to thank explicitly because they normally end up being the catalyst that helps me pull the bloody things off. In particular I should thank Matt Webb, Cal Henderson, Simon Willison, Natalie Downe and Denise Wilton who have all calmed me down (and sometimes just told me off) when I’ve been freaking out about a talk.

Anyway, so the talk for dconstruct didn’t actually start really firming itself up in my head–didn’t really start coming together properly and making sense to me–until probably two days before the event itself. The day before the event I finally had a structure and all the points but still absolutely no slides. The day before the event I started knocking out some slides but found myself rapidly running out of time. At one point Cal sat down with me and just went, “Okay, what’s your next point? Write it down. Leave it. Now what? What do you want to say next? Write it down. Next.” He was–frankly–amazing. If he hadn’t have been there, I’m not sure I’d have won the battle this time.

That evening, we were supposed to go out for a dinner and a drink with the other speakers, but that was completely impossible for me. My heart had been pounding in my ears for about three days solidly, and so instead of going and drinking and socialising I found myself barrelling along, fighting and wrestling with this aggravating thing that I’d decided to create, sitting in Simon Willison and Natalie Downe’s flat scratching out slides. At around eleven I gave up for the evening. My brain felt like it had been hit dozens of times, as if it had been tenderized. Nothing I did to it could get it working again.
Five hours after going to sleep at my hotel room though, I was up again – writing slides, bashing points around. It’s worth saying that at this point I’m still fiddling with the character spacing on some of the slides, that they look right being still as terrifyingly important as what they actually said. Such is the depths of my lunacy. By 10 am I was not with the rest of the speakers upstairs watching the talks, I was downstairs hidden in the Green Room blocking through the thing section by section. And with occasional breaks to go and buy fizzy drinks and things with sugar in them, that’s where I stayed for the rest of the day. No doubt I made my fellow speakers nervous with my frantic energy and buzzing adrenal frenzy. Everything tasted of metal.

And then finally, suddenly, around midday – after weeks of fighting it just started flowing naturally.

I finally finished the talk at 3.30pm, having seen not a single other speaker on stage all day, hiding in one of the chorus rooms in the basement. I had to get my laptop onto the stage by 4.30pm when Matt Webb’s talk started (my talk was directly after his), so I had nearly an hour to do my one and only rehearsal. Reassured and comfortable that it was going to be okay, I read it out loud to myself.
And it stank. It really really stank. I’d been trying to just get a sense of the time, but I started to realise the gaps and absences in the argument, where it was running long, where I sounded like a hippy and where I sounded like a corporate drone. Quickly squeezing in a few extra bridging comments in my notes, and adding in a couple of extra slides, I turned around and had to call it finished.

I always find at conferences that the forty-five minutes or so before I’m actually talking are among the easiest of the whole time, and this was no exception. It’s like my brain accepts that there’s nothing more that it can do, that there’s no longer any point in worrying, and it just gets out of the way. It’s great. My brain-state was also helped substantially by Matt Webb’s talk on The Experience Stack. I found it beautiful and fascinating, and as elliptical as ever. You have to work with Matt’s talks–he makes these extraordinary leaps–but if you do you’re normally rewarded. There’s always something that makes me raise an eyebrow in suspicion or disbelief, and always something else that makes me see something old in a new way. I’m a systematiser, on the whole. He makes inspirational leaps. In many ways we’re completely different from one another, but those differences often got us to good places when we worked together. In this case, he got my brain moving again and engaged with what he was saying, and it got me nicely focused for my turn.

Which by all accounts, after all this stress and pain, seemed to go fairly well. I’ve heard lovely comments from a bunch of people and despite all the misery it took to get there, I genuinely enjoyed presenting it. I don’t think it’s as good a talk in the end as Native to a Web of Data or Greater than the Sum of its Parts, but I think there’s substance in it and some challenges to our current ways of building and thinking about the web. That makes me pretty happy.
I’d love to know what you guys thought of it – and whether you had any points or suggestions about it. It’s probably going to get another airing at Web 2.0 Expo Berlin in about six weeks, so any thoughts that you had would be very very much appreciated. As I’m doing the talk again though, I’m still undecided about whether or not I should post the slides. If you’re keen to see them, can I ask you to wait a month? Or if you’re really desperate and need them to make a case to your manager or boss, then let me know via e-mail and I’ll send you a version of them.

With dconstruct over, the rest of the weekend became about letting my hair down, but it wasn’t going to start immediately. I stuck around for some of the main after-party, but by eleven in the evening, I was basically unconscious with exhaustion. I ducked back to my hotel room to dump my bag and passed out almost immediately and woke up still fried the day afterwards. Only after a lovely breakfast with fellow dconstruct nerds and a good chunk of BarCamp did I start to feel human again. And only a couple of hours at the pier on Sunday afternoon could really shake all the cobwebs out of my head. You can see all my post-dconstruct pictures on my Flickr stream: After dconstruct, but here’s one to capture the mood:

I suppose in retrospect I’d call the whole thing a difficult but worthwhile experience for me. I need to find a better way to deal with my stage-fright and to get these talks out without so much pain, but at the moment it’s survivable. The most important thing is that they’re enjoyable and interesting to the people who watch them. So if you were there, I hope you enjoyed the talk as much as I (finally) enjoyed delivering it. And if you have problems speaking in public, then maybe this post will have given you some perspective. You’re not alone!

8 replies on “On writing my talk for dconstruct…”

It’s just great to see that I’m not alone. I have a batch of presentations coming up, starting on Tuesday. There’s about 8 or 9 different presentations and workshops running over the next two months. Some of them I’ve done before and know that if I just repeat them I’ll be stale – they need rejigged. With others I’m left wondering how on earth I ended up giving that title to the organisers!
The greatest stress for me comes not from numbers or standing on stage, but misrepresenting someone in the audience whom I respect. Nightmare!
Maybe we should start a club or something…

I though your talk was the best at d.construct, it was one of those talks that made me want to drop everything and come and work on FireEagled. You had great slides, examples and spoke at a good steady pace. You even managed to crack a few jokes to keep the energy up at the end of the day.

I even had to just check the website to make sure I wasn’t talking about another speaker, because It’s that hard to think you were nervous. Heck, I get nervous about asking questions at events like that.

I think experience is the only way to totally cure fear of public speaking, university helped me a lot but I have a log way to go before I could do anything like that.

Keep it up Tom 😉

Thank you for sharing this. It makes me wish my talks were ambitious enough to support this kind of epic preparation! I mean, I do agonize as well, but I usually find up obsessing about fonts (how ridiculous is that) instead of anything productive, like the clarity of my ideas. ^_^ Actually, your post also makes me realize that I wish I were talking to the same audience in this kind of serial way that you do. You have the chance to build on your previous talks and create a kind of mini-culture around your presentations. That is so cool.
Now remember: when you are in San Francisco you are supposed to let us know so we can hustle up a Werewolf game for you!!

Cal sounds like a useful guy to have around.
I started writing for money nearly twenty years ago, and the blank screen still terrifies me. I’ve got various ways in –
Write a flowery introduction, you can delete it later
That idea that’s nowhere near good enough – write it down anyway, you can always delete it later
What do you want to say? What do you actually want to say? Just write it down, you can always delete it later
– but they all boil down to just start writing. It’s hard, though.
(I do quite a lot of deleting.)

Thought your talk was good, although a little familiar, having seen you talk at Future of Web Apps a couple of years ago (the ‘testicle’ slide seems a little overworked).
For me, the summarisation you did at the end was a little laboured – perhaps because of nerves – and didn’t quite seem to consolidate on your talk (repeating points rather than bringing it all together).
I liked the stuff about the ‘web of data’ though. A couple of things you could elaborate on:
* is it important for website to identify a ‘primary’ data type/’social object’ (eg photos for Flickr)?
* if your data isn’t private (photos), but public (eg books, movies), how can you find, or invent, unique identifiers? (why doesn’t IMDB have an API?)
* You argued ‘the more data, the better’ (eg for Flickr) – but how do you navigate/visualise this, and how important is ‘newness’?
Hope that helps…

Hi Tom, I’m surprised by how much you went through to get the presentation out – you looked like a pro. Guess all that pre-gig stress got the nerves out of your system… I really liked the talk, it was my favourite of the conference. I’d like to hear more of your thoughts on things in the web of data world like decentralization (I don’t want a “mashup server”), identity, screen scrapers (to create API’s with), who actually creates mashups, consolidation (is it necessary? how will it happen?).

this is inspiring to read. thank you for posting it.
i talked to you at FoWA in SF after your “greater than the sum of its parts” presentation and you shared the idea of cursing the audience secretly from backstage before speaking, as a way to manage stage fright. i love that idea.

Comments are closed.