dConstuct 2008, Designing the Social Web is a yearly event held in the Brighton Dome, including speeches and workshops around the topic of the evolving social side of the internet.
After catching a few friends on Twitter and GeekUp talking about dConstruct I thought I’d get myself along to a conference again, as I’ve not been to any for a few years. Being scouse, I blagged a ticket off a friend at .net, one of the event sponsors, booked a hotel and put a call out on GeekUp to see if anyone wanted to split petrol money (of course I’d go by car, a chance for a long war-drive!)
The mighty Chris Mills from Opera took me up on the offer, so on the Thursday 4th Sept. I picked him up on route through Manchester. I’ve worked with Chris a few times before, when he was an editor at Friends of ED, and I was tech reviewing books, so it was good to finally meet face-to-face. The drive down to Brighton (from the North West – 6 hours!) flew by thanks to good conversation and an awesome selection of music from the iPod.
We got to Brighton, dumped the bags, and hooked up with some more ex-FofED’ers, Bruce and Matt, had dinner then headed to the pre-party at Po Na Na. At the club I met up with two more Northerners, Rob O’Rourke (web designing Scouser in Burnley, friend of my youngest brother) and Dan Donald (Here In The Hive), drank a bit and chatted, taking in the faces and clocking the rock stars of the standards world… ahem.
So an early start on the Friday, due to me not really reading the agenda properly saw me at the venue for around 8.50am, where the security guys were keeping the nerds at bay from their validating demi-gods. Brighton was belting down with rain, so everyone was fairly glad to be let in for registration.
Coffee and croissant acquired, a quick wander through the lobby area was in order. Silverback had a gorilla doing the rounds (ok, not a real one, a man in a monkey suit) handing out branded bananas. Nice touch!
Bumped into Pete Aylewood at the Friends of ED stand, another email-only acquaintance, and waited around for Dan, Rob, Matt, Chris or the pint sized walking ASBO Bruce to appear, and head to the first of the seminars.
Steven’s talk was my first taste of the dConstruct event, and I have to be honest I was a bit ignorant as to who he is. However, his discussion of how social networks have been in effect for centuries, with documented data from the last cholera outbreak in London in 1832 proved he knew what the hell he was talking about, and could be interesting with it. The gist of the talk was that for local news, people close to the source are always going to be best, as Reuters won’t be bothering to comment on a car crash on your street. It’s news to you, as you live there, but the rest of the country doesn’t really care.
The ‘long zoom’ of gathering information by a local networker (a man of the cloth in those days) was the key to isolating the cholora cause and Johnson paralleled this with bloggers in his neighbourhood writing up small stories allowed him to see a van on fire in his street, while he was holidaying on the other side of the States. This is where he introduced outside.in and how that gathers your local news from location aware blog searches using human readable terminology.
It was pretty interesting and quite easy to see how it would be relevant to me. For example, if I moved to Spain I’d want to know what the neighbourhood was like where I intended to live. You can get an overview from Google, find out where to eat etc, but it’s the local blogging news that’d be key.
Next up was the slightly nervous but energetic Aleks, a video games corresponant from The Guardian, took to the stage and got us started with her talk on, basically, why the video gaming industry should be taking a cue from the open source and social networking side of the web. Aleks clearly likes her games, and banged on about the differences about simple, social games and over-indulgent, beautiful console games.
She asked for a show of hands about how many people were from the games industry at the conference, and how it was strange there were so few (I counted about 10 hands). A fair point, but knowing the gaming industry a little, I wasn’t too surprised. The games industry is not an open bunch, and projects are developed in secret, mainly due to a games gestation period. The speed the web is developing, compared to the time to create a game means that having a game come out at the right time to capitalise on a current social networking darling site would be difficult.
Anyway, Aleks name checked the Dreamcast and Phantasy Star Online, making me nearly jumped out of my chair and yell support. I collared her later and offered to buy her a drink for the Dreamcast reference!
Josh’s talk was pretty good. In a nutshell, it was how people will make choices based on what other people are doing when limited data is available. Human’s act like sheep and follow the crowd if they don’t have enough facts to hand. He used a very funny, old black and white TV clip from Candid Camera, set in an elevator. Four or five people would enter the elevator, one of which was the “sheep”. The others, on cue, would turn to face a certain way in the elevator, and the sheep would always follow. This worked to such a degree, they managed to make one guy take his hat off and put it back on, following the crowd.
He then went on to show interface design techniques that reinforced how popular or good a site was, by showing a “random sampling” of users, of course who were always good looking, participating well on the site etc. Basically playing on the heuristics in decision making. Fairly straight forward if you do use any psychology when designing, but I’m sure a lot of developers in the audience will have clicked with it, as there were some real good points.
One real intersting point he noted was how one large US retailer had seen a sales increase of 20% after removing the need for buyers to register online. Let them buy without it, but paint the fact that they’re not registering as a loss for them. After people have invested time on a site, and completed a few tasks or made the site a little more personal, getting them to register is easier, as they’ve put time in and will suffer a loss as a result of not registering. People suffer from loss-aversion in everything they do.
However, Joshua’s best example of a design that did the hard-sell by repeating the joy of current users was a bit crap. He pointed to FreshBooks as a perfect example of hammering the message home, but to us English designers, the site was a piece of typical American marketing design. I personally think the English taste in advertising and marketing is pretty advanced, and his example probably raised eyebrows and a few wry smiles.
Lunch – The Opera lads, Bruce and Chris along with a few Yahoo! devs were heading for a sushi bar. I hate fish, but Chris persuaded me to come anyway as they’d have some meat dishes. As it turns out, the food was excellent (mostly teriyaki for me) and I made Chris laugh by asking “did I ever tell him I loved sushi?”
Designing for Interaction could possibly have been titled “Digging for Interaction” as the Digg site, interface and design decisions were featured prominiantly by Daniel, but since it was a (well known) real world example, and he was very candid when discussing features inclusions and drops, it made for a good focal point.
Some great insites from Daniel about how small design decisions would affect the majority of Digg users, and conversley, how some features the team thought would do well were underused by users (File transfers on Pownce).
Burka also pointed to the loss-aversion tactics in getting people to sign up, as mentioned by Joshua Porter. Let them digg a few stories and get involved, then tell them that if they want to save what they’ve done so far, register. This line of thinking is becoming more and more usual with social networking sites, reel the user in before they sign up, that it was nice to see it nailed by the speakers. He also lamented the crappy Digg sign up process and showed how they’re streamlining it and hoping to keep it as unobtrusive as possible.
It was quite good to hear Daniel candidly talk about how some features had little user-metic testing and were added on a rolling 30 day launch inclusion. Some of the larger features obviously did go through the user testing process, as you’d expect, but I have a feeling few places have the time to test every single new feature. Good to know the big guys do it too.
Tantek is well known for banging the MicroFormats drum, so I was hoping this wasn’t the sole-focus of his talk. Fortunately, it wasn’t, and he covered a lot of familiar problems that the social web presents the users, such as “social network fatigue”, the tedious process of entering all of your personal info again and again for each new social site.
Tantek was the only presenter to actually get full on geeky with his slides and drop some code. This was in the MicroFormats section of his talk, and I felt it was a little “preaching to the choir” for this specific audience. While we’re on MicroFormats, I’d been asking most people I knew this week what they thought of them, after hearing the Rissington Chaps commenting on them in their podcast. The general concensous is that they’re a good idea, but there’s no real groundswell movement that will prove to be a tipping point. This was all but confirmed by Tantek when he discussed the Operator plug-in for Firefox that parses HTML pages for the hCard mark up and hi-lites it to the user. I’d used this only on Wednesday this week, so knew exactly what it was. But I feel that if MicroFormats are so good, they should be built into default browser behaviour and deployed on all social networking sites. There’s a lot of movement to get them in the social sites, but I’m unsure as to what the browser vendors think of them.
Çelik also discussed the inherent problems of signing into a multitiude of social sites. He asked for a show of hands as to who had how many social network signups, for which a good portion of the audience had over 20 each (count me in). There’s a few services that provide a unified login place, like OpenID and Chi.mp, plus things like Gravatar that will provide one central point to changing your avatar on supported sites, but at the moment, I don’t think there’s an all encompassing service. First to market will probably ride the wave.
I think the best part of Tantek’s presentation was the dissing of these services embedding Google/Yahoo/Microsoft email address book sharing APIs. You’ve probably seen them, “Oh you’ve signed up, why not see which of your address book buddies use this site.” Tantek gave this a kicking for two reasons. One, that they usually then spam your whole address book because they’re not entirely clear and ever so slightly devious in getting new sign ups during your buddy-checking process, and secondly, because it’s teaching web users bad habits.
By getting a user to provide their username and password when inside another site/service, you’re lowering their guard against online phishing attacks. How easy would it be to start a new online community that on the surface looks like a safe, fun place, that then asks you to connect with your address book mates, and bam, robs your username and password in the process because they fake the Google Address Book API? Bad mojo, man. Don’t teach users bad habits.
The lads from Dopplr gave the best talk of the day. Their energy, honesty and simple joy in what they were doing was great to see. They talked about how Dopplr did little things to the site that a user only notices over time, and puts a smile on their face despite them having used the site for a few days/weeks/months. They showed how with their site they’re quite happy to launch people off around the web as they know they’ll come back. No walled gardens here, they want to be a trusted reference point of jumping off platform.
A good mix of design and coding theory that goes into Dopplr, and how attention to trivial details is what makes the site sticky and loved, and having heard them talking, I’d feel very happy about recommending Dopplr as a trusted site and a way to go about the social side of things.
Design Matt was great when talking about the feature on Dopplr that extrapolates your distance travelled in a year and matches it to an animal’s top speed. The site’s fastest user is the speed of a whippet, and the slowest users are slower than snail’s, prompting the guys to look for slower moving objects, like the NASA Shuttle crawler! Matt said that he spent waaaaay too long on this whole feature, in terms of research and what to design, but the end users delight proves that the whole thing has been very worth while. Nerds take note, let the designers play with their crayons, as sometimes an idea they have can’t be measured in cost per hour, and it’s worth going overboard with it.
Hmm, that sounds like a pretentious session… but to be fair, Jeremy pulled it off very, very well. Intiially I thought I’d somehow fell asleep at dConstruct and woken up in the wrong conference, such was the sheer nerdity of what Keith was talking about, but with a few name drops of an Asimov series he reeled me back in, and went on a compelling discussion of how estimates of past and future numbers can’t be based on a current time slice, bell distribution curves, Gates has lots of money and MySpace is fugly.
I really can’t begin to repeat the context in which Jeremy recanted this, as he’s obviously done this public speaking malarky before, but his talk was an all encompassing look at social systems and how users interact very differently when there’s a million opposed to the actions of one. Of how it’s basically impossible to predict what we’ll all be doing and which social networking site will be the next big thing.
edit: Jeremy’s now put his talk online in HTML, so you’d best give it a read if you want to know more…
After show party – Above Audio
Before the after-show (oxymoron?) I met up with an old work colleague from the Subnet days who lives in Brighton, so turned up fashionably late… meaning I missed the food. I scrounged a few beer tokens off a Yahoo! dev using the Chris Mills guaranteed winner line, and had a good chat with Brendan Dawes and the mN lads. The atmosphere was good, and when we moved on, we headed to the Old Ship hotel, where I Jedi-mind tricked the bouncer into letting 11 people in who didn’t have a room, downed a lot more drinks, and had my coat nicked by a fellow scouser (thanks @sanchothefat!).
All in all, I thought dConstruct was fantastic. I met a load of people I knew, met many new friends, and learnt something useful and applicable from each and every session. I’ve already said to the lads at JP74 that a few of us will be going next year, and hope to catch up with the people I met at future conferences.
Oh, and I found 987 new SSIDs on the war-drive down, in case you’re interested…