How to do good talks at conferences

Hint: Teach don't sell

Last Friday I went to Brooklyn Beta. I'd been looking forward to it a lot, as I know several people who'd been before, who have all eulogised about the conference's honesty, creativity and focus on supporting like minded internet-ty craftspeople to help each other get better at what they do.

This year they increased the attendee count a lot - from ~300 to ~1300, and you could see the folksy homeliness straining at the seams. Graphic design / human interaction limits aside something much worse happened - it became (in part) a platform for self promotion and PR. Which is rubbish.

Preaching to the choir

I'm not going to go through all the talks. One was awesome, the rest were ok - but one, by Joe Gebbia the CPO of Airbnb - was car crash awful. It was one of the most arrogant bits of conference PR I've ever seen, and, a annoyingly, a massively wasted opportunity to share his experience with peers who could really benefit from his experience.

This made me think

I got really annoyed by this. 1300 of the most interested developers, designers and internet people in the world in the room. What potential. The first thing I did when I got back to my computer was to write a long piece trashing Joe. But, after reflection and some advice from cooler heads, I've decided not to publish that. Instead, what I felt would be much more useful would be to:

  • Write about the amazing talk I would like Joe to have given
  • Note down some points on what makes a great talk 
  • Commit to only doing this myself from now on 
  • Begin the process of organising talks like this in London 

So that's what I did. Carry on reading if you think this is interesting and you'd like to get conference talks back on track as a medium for sharing knowledge and learning instead of PR.

What Joe said at Brooklyn Beta

Below I'm going to to give you some context and explain a little bit about what Joe said at Brooklyn Beta. It was an awful talk, but it also contained a lot of presentation patterns that I've seen recently.

The talk started with a very promising premise - what unexpected things had Joe learnt since his last talk at Brooklyn Beta two years ago?

When I heard that title I thought, wahey! - lots of amazing insights into the rocket ship ride that Airbnb has been on on the past two years. Maybe we'll get to hear about how you raise $117 million dollars on a $1bn valuation. Or maybe we'll get to hear about how to buy and integrate five companies in one year (e.g). Or maybe what their legal argument is that means they can confidently defy the NY Attorney General?  

A content-less sales pitch

Instead, Joe chose to come on stage as a kind of tech-CEO caricature (1 part Jobs, 1 part Dorsey, 1 part Zuck) and present a  trivial sales pitch that contained no new news, and left out anything useful or interesting. The centrepiece were five, empty brand words (e.g goosebumps and connections) supported by some cheesy customer stories, interspersed with some vapid exercises ('write down three things that give you goosebumps!'). Along the way, Joe forgot to tell us anything we didn't already know about Airbnb.

A deluded / revisionist story

Joe also did that classic Silicon Valley revisionist nonsense about the backstory of the company. He tried to create a kind of aura of inevitability, starting with a cheesy website they threw together and ending with a billion dollar business as if it was nature's true intent and nothing happened in between.

Joe, we all know that's not true! You must have hustled and coded the shit out of it. Tell us about that, please.

But. But! But!!!

But the reason I'm so annoyed is that the premise of the talk was really, really interesting. "Five unexpected things I have a learned in the past two years building Airbnb". How amazing would it be if Joe had shared the real stuff. The interesting stuff that might actually have helped his captive audience of 1300 web developers, designers and business people? Joe's a remarkable man running a remarkable company. 

Here's five things I would have loved Joe to have talked about instead:

#1. How to design at scale

This is fascinating. Joe is a great designer, and Airbnb's user experience is superb. How did he do it? How did he go from a tiny team of three in 2009 to a huge operation currently hiring over 60 positions in 12 countries whilst retaining quality. What mistakes did they make? How do they make product design decisions now? Which personalities were pivotal? How has Joe stayed in control of design so well during the past two years? The answers would be fascinating.

#2. Managing young founder dynamics

Airbnb is currently valued at over ONE BILLION DOLLARS. Yet four years ago Joe, Brian and Nathan were renting their spare room out with air beds thrown in. How the hell have they stuck together through all this? What have been some of the biggest fall outs? How did they divide up responsibilities for running the business? What advice would he give to companies with multiple co-founders that he wished he'd know two years ago? Again, I would LOVE to know this, as I have a company with two co-founders right now, and we're at the stage where we need to start dividing up the work.

#3. What was worth shipping?

Next - product design. Airbnb is a superb app, and it would be amazing to know the details about what has worked and what hasn't worked on their site. What impact did the neighbourhood guides make? Are there any really important features that bumped up growth, conversion rates or ARPU levels? How do they manage their roadmap? When you have people in 34,000 cities using your service how does local culture fit in? What tools do they use at Airbnb? How has the toolset evolved in the past two years? What did they build themselves? Why? Again, I'd really love to know this!

#4. How to make lots of money

Airbnb is a mega money making operation. Fact. You can pretend its 'gratitude and connections' that makes it successful, but there's a reason they've raised $120,000,000 as quickly as they have - its a winner takes all marketplace and that means BIG money for the winner. I'd love to get some insight into how they figure out how much money to make, and how that impacts growth. Why 9-15%? What experiments have they tried? How do they navigate regulation? What pricing tricks work best to get people to list their rooms? How have the handled handling so much cash in the past two years? Again, fascinating stuff, the answers to which would have been really valuable to his audience. 

#5. Why us?

Finally, I'd love to have heard Joe talk, honestly, about why he thinks Airbnb has won  this game so far? What unique combination of luck and sweat made Airbnb the winner in this marketplace business category? Hint - it's not gratitude! The three founders must be formidable. Timing is on their side. But why did they win? I'd love to know what Joe thinks is the real answer.

So, that's what I would have loved to have heard about. I'm sure if Joe had sat down and thought about it a bit, and considered the impact he could have had by teaching instead of selling he could come up with five much better things - its his company!

The question that next springs to mind is are there any general principles to extract from this? What makes a great talk instead of a crap one. 

Here's my thoughts:

Rule #1. Get your product/audience fit right. Conferences are about learning.

This is my biggest point. It seems like some people have forgotten the point of a conference. Conference's are not concerts. They're not press launches. They're not cocktail parties. Yes, people like to be entertained and spend time chatting, but the primary point of conferences is to share knowledge. Academic conferences do this well, if not dryly - presenters are generally always presenting new research findings, and they have to be peer reviewed before they come on stage. When they get there they are teaching delegates about what they have discovered. 

In my opinion, everyone wants to come back from a conference having learned something. If they had a laugh and met some new friends along the way, that's awesome. But if they didn't learn anything it was a waste of time - I could always have more fun going on holiday with my friends or checking out some art. But conferences are for learning, and by association teaching. 

Rule #2. Tell as much truth as you can.

Telling the truth is essential to teaching, and discovering new truths is the process of learning. So when you are giving a talk it's vital you tell the truth, and even better is if you can reveal new truths to your audience - especially when the information is normally private or swept under the rug. At Brooklyn Beta there was a very inspiring talk by Catherine Hoke from Defy Ventures where she told the truth about how her last venture had failed due to a sex scandal involving her and a prisoner. Wow. That was some serious truth telling.

Rule #3. Blend details with the big picture

People learn practical tips from the details, but are inspired by the big picture. You need to mix both these up. Again, at Brooklyn Beta Tim O’Reilly did a great job of this, sharing how he had done his marketing for his first book and accidentally built the first commercial website. If you talk is all details, it will be boring (like an academic paper) but if its all big picture it will also be boring (like Joe’s PR fest).

Image credit - Beautiful Sketch notes of Tim's talk by - Evalottchen on Flickr

Rule #4. Stay humble, be funny, tell stories

Humility is vital to delivering a good conference talk. If you are lady GaGa, maybe not. Or if you are launching a new smartphone, again, perhaps not. But if you are talking to peers its essential - you may have more success than the people you are addressing, but to inspire them and get your message across they need to see themselves in you. And that means humility.  Being funny helps too. This talk by my old boss Adil is a great example of doing this well.

Rule #5. Put some effort in.

Finally, you have to put some bloody effort in! Talks that are clearly recycled variations on pitch decks, internal pep talks, or something that you gave before - no. You are not a politician doing a stump speech. If you are going to do a conference talk, take the time to make it unique and special. Your audience will love you for it. My favourite example of this was Russell Davies talk at Playful in 2009. It was a fantastically crafted piece about the edge of games. I loved it.

Image Credit

Putting this in to action

I like to think I've tried to do this a bit in recent talks I've given. I'm no Tim O'Reilly or Russell Davies. But I'm trying!

Last night I had the chance to put this into practice. I gave a small presentation to the DigitalDumbo London audience about Makeshift, our process, and our products (slides up here soon). 

And I told the truth about things people don't normally like to talk about - how much money we've spent, our cost to acquire a customer, the things that have gone wrong, and the things that have gone right. It was short, so I couldn't go into much detail. But it was a start. It went down quite well.

Work In Progress

Writing this piece, and reflecting on the shambles of Joe's talk at Brooklyn Beta, has made me more convinced than ever that there is a space for a conference focused on telling the truth about building businesses and products on the internet. Probably this is what Brooklyn Beta was like the first two times. 

At Makeshift we run a fortnightly event called Work In Progress where we share, unsurprisingly, some unreleased work in progress on one of our projects with a mix of critical friends and new people. We generally ask another London startup to do the same. Its a Chatham House rules event, and its great. 

So, next spring I've decide that we're going to put on a bigger thing, focused on telling longer, non-fiction stories about the journey we are on at Makeshift, and hopefully we can get some of our friends to join us. Basically like our Work in Progress events but bigger. If you'd like to come you can start by emailing me: