Categories
dConstruct08 User experience

dConstruct ‘Designing the social web’ 5 Sept 2008, Brighton UK

dConstruct in Brighton this year was held in the full glory of an English beachside summer – sleeting rain and gusty winds. Inside the Brighton Dome, 700 web geeks gathered to hear what ended up being quite a mixed bag of presentations from speakers about various aspects of ‘designing the social web’. Light on the technical detail, all but Tantek Çelik, focussed predominantly on the psycho-social aspects of the social web.

The conference opened with Steven Johnson (Everything Bad Is Good For You) comparing the similarities in the way in which data visualisation combined with hyperlocal amateur knowledge helped prove the source of the 1854 cholera outbreak in Soho, London with the new opportunities that are emerging with social mapping and visualisation technologies in the present. Johnson ‘s 2006 book The Ghost Map is a detailed look at the cholera outbreak and the second half of his talk focussed on his own social mapping project Outside.In. Currently US-only, it is very similar in style to Everyblock, bringing a personal hyperlocal focus to news, people and events in local communities. As Johnson says, usually you care most about things that happen within a small radius of where you are right now – 1000 feet/350 metres – and so Outside.In and its services like Radar use geolocation to deliver this information to you as it happens. He contrasted his service, which parses social media like Twitter and place-centric blogs, with the more obvious (and I would add, easily ‘monetized’) geolocation services which already exist around restaurant reviews and local businesses. Johnson was an engaging speaker but his US-centrism/universalism did emerge at times – and his postulating that Brooklyn contains four of the top ten US local blogging neighbourhoods also seemed to colour his perspective (Johnson lives in Brooklyn).

Next up was Guardian columnist and gaming academic Aleks Krotoski. Aleks was very engaging and energetic, pacing up and down the stage, waving her arms animatedly whenever possible as she explored some of the things that web designers could learn from game designers and vice versa. Her social psychology research around gaming has looked at the different models that game designers use to keep players engaged over long periods of time – the ‘stickiness’ that web designers long for. Of course, as she pointed out, game designers can play the niches and their proven publishing-style business model gives them a distinct advantage over the web where most web applications needs to appeal to a far broader audience and have very few proven sustainable business models (advertising and . . . ). In the controlled systems of games players can be given ongoing ‘carrots’ to keep them engaged and willing to move on to the next level/challenge, and even in immersive sandbox environments like Grand Theft Auto the player is made to ‘invest’ significantly in their game experience – enough to keep them playing for long and repeated periods. Games also can operate as ‘enabling systems’ whereby social value is emergent through community building, storybuilding, and even extends to the obvious player communities around World of Warcarft or in simpler terms the creation of game FAQs, walkthroughs etc. Closely related are what Krotoski termed ‘psychological systems’. These operate around relationship building in-game and also frequently leverage the ‘collecting urge’, and increasingly the generation of in-game assets (see Second Life). Krotoski closed by postulating that a lot of what we see in the next generation of games around at the moment – the social gaming of the Wii and NDS, the online/offline mix that Little Big Planet and Spore are going to offer – might have had their genesis in the long-lost Sega Dreamcast and for reasons of platform competitiveness and industry secrecy, many of the opportunities that the web has explored have been late to rise in the gaming world. On the flipside, the web could learn a lot from the engagement models of the gaming world.

Aleks was followd by Joshua Porter. Josh’s recent book Designing for the Social Web is a quick reference set of design patterns that explain some proven methods for building engagement and community when designing social sites and applications. Josh’s presentation was a little disappointing in that although he presented a series of design patterns and techniques to exploit users’ cognitive biases, he was light on evidence, and like Steven Johnson made some terrible US-centric/universalist statements about behaviour. The biases he focussed on were;

– representation bias – making highly visible the behaviour you wish other users to have on your site even if this is not typical of other users. eg. Yelp’s ‘featured reviews’ and Freshbooks’ ‘what our users are saying)
– loss aversion – the couching of desired behaviour in terms that avert loss or risk.
– ownership bias – reminding users that they should care because it is ‘their’ stuff. eg. Flickr’s use of ‘your’ in all their UI

His presentation drew extensively on the June 2006 article Eager Sellers, Stony Buyers by John Gourville, which explores techniques used to convince customers to change their behaviours etc.

As the first question from the audience asked, only half jokingly, “Isn’t this evil?”.

In a similar vein, Daniel Burka from Digg and Pownce, presented a series of slides that explored the methods that Digg and Pownce use to encourage users to firstly sign up to their services, and secondly, participate in positive ways. Whilst visitors to Digg can always use Digg without creating an account (much like the 95%+ of visitors to Amazon who use the site for product research and as an image library for their iPod), Digg’s aim is to sign up as many people as possible. In order to do this it needs to ‘go beyond altruism’ and offer real benefits to those who do sign up, as well as significantly reduce barriers to entry, and in the case of Pownce, allow logins with accounts from other services (cf Opensocial). Burka cited Geni.com as a best practice example of encouraging sign ups – it not only shows users what they can get from the site, it also starts them off in the process of creating their family tree, and makes it very easy for them do complete their signup with a minimum of information.

Encouraging positive behaviour and deterring trolling and gaming of the system is the next challenge. Burka outlined the benefits of using personal profiles with photos to build trust amongst users, as well as tweaking text copy to break through ‘tension points’. He pointed to Get Satisfaction‘s use of emoticons as a good example of conveying mood accompanying messages as a way of reducing the chance of user comments being taken in the ‘wrong spirit’.

Tantek Çelik followed with a detailed presentation on using microformats, specifically hCard to explore social network portability. The presentation and its information about implementing social network portability with hCard is available through the microformats wiki.

The final two sessions were more conceptual and were fun. Matt Biddulph and Matt Jones of Dopplr gave an initially grating but finally witty and funny presentation on, well, Dopplr. It was much more than Dopplr, but they used Dopplr as a case study and set of examples for how it is not only possible but also highly desirable to build web applications that are about slotting into and contributing to the ‘coral reef’ of the web, rather than trying to work as a walled garden or honey pot. They paid special attention to the notion of ‘delighters’ or in their world, ‘data toys’ – surprises that make their service pleasurable and fun to use. The last session was from Jeremy Keith. In a lovely and somewhat laconic presentation, Keith exploded the notion of predictability in scale free networks , drawing on sci-fi and pop-sci respectively. It was a fitting way to end the day.

Then it was back out into the rain to the afterparty – which our party decided was a veritable ‘bbq’.