H2G2 Refresh: design and technical challenges
The new redesigned H2G2 site
"Take 1 digital web agency,
2 internal BBC departments,
1,000,000 unique users,
and 12 years of love and enjoyment...
Develop.... but very carefully!"
When we were awarded the H2G2 redevelopment work from BBC, some of the team asked what it "was". Like most things which stem from likening one thing to another, it was fuelled by mankind's predisposition for making things easy to understand.
I've heard it called Wittypedia, which is as close to a true label as one could come.
It's effervescent, brilliant, astounding and a wonderful testament to the beauty that can be found in language and user generated content. Unlike other 'similar' sites, it's a conversation, not a lecture. And in that distinction lays everything.
We inherited H2G2 from its previous iteration, a UGC web 2.0 project inspired by Douglas Adams back in the 1980s. Unfortunately, since that date, no one had really revisited the site's framework and branding - a terrible shame when you consider the love and esteem the fiercely loyal community and greater world hold it in.
Amusingly, the site was so closely akin to Douglas Adam's and the Hitchhikers ethos that no one person, including the community, seemed to know all of the various granular functionality offered within its frame - something that would continually bite us throughout the next few months. H2G2 is as unique, funny and quirky as the man himself and our UT work only uncovered about 80% of functionality as we would find out later...
User Centric Design
We spent many long hours reviewing both the existing site's legacy content and functionality. This presented us with a number of key challenges throughout the design stage of the project. The biggest and hardest of which was the love and personality of the brand. We knew that the community were so impassioned when it came to the website that we simply had to involve them at every stage of the design process.
The knowledge of the existing community with their fantastic passion for the brand, and our cold, logical IA work to detangle the site and increase accessibility to new users yielded surprisingly positive results. Everyone's fear that change could upset the community turned out to be unfounded - in my opinion, this was 100% down to the efforts of Sam, Chris and Nick to involve the community so integrally from beginning to end. This process was a credit to both the community, and the passion of the H2G2 staff.
I personally worked very closely with Sam and Chris, thrashing out ideas for how best to represent the brand. We thrashed out many iterations of the design, ranging from a dusty old antique room full of iconography such as teacups, cocktails, plotting mice and flying dolphins mounted on the wall; to a sea / sky based images with inspired flotsam and jetsam bobbing along the curiously angled sea. All crazy and fun, but not quite right when it came to attracting a new audience - which is key to the site's continued existence.
So, we instead opted for broad strokes of inalienably subtle iconography such as the dolphins leaving earth, and a few subtle nuances lost on all but the most observant of viewers. Whereas on previous projects, we'd found that the BBC GEL offered significant challenges to users and editorial teams who wanted a less formulaic handling of their brand, on H2G2, GEL provided us with a perfect design pattern partner. and despite a few challenges; the site benefited greatly from the uniformity this brought.
The look and feel / UI work tested amazingly well within the community, with sporadic conversations starting up on the site about how pleased they were with what they had seen. We were extremely proud when Sean, one of the original creators of H2G2 wrote on a discussion that he believed
"the whole page - from graduated fade at the top to cloud-covered earth at the bottom - simply oozes graphical class".
Technical and Information Architecture
A key challenge that returned continually was the existing sites ability to allow the community to exhibit their own personality over their presence via use of GuideML, the language that runs the UGC content from the community. GuideML allowed users to put free form (and often malformed) HTML-esque content. This caused incredible problems for us as we could never entirely tell what that content might be. To combat this, Ben Poole, who deserves significant credit, worked tirelessly very closely with the DNA team to create conditional logic to strip out and replace whatever malformations were possible to bring the code in-line with BBC standards and guidelines.
Problems, problems, problems
Unfortunately, a combination of legacy content and our necessity to wait for the DNA internal team's delivery of systems and APIs, meant that the project went on much longer than expected - a key disappointment for us was missing the 10/10/10 (binary code for 42 for us geeks out there) as a launch date or community celebration.
The reasons for this were many, the APIs brought with them a swathe of necessary research, testing and development. This, combined with the ever shifting sands of requirements from both the editorial team and the community at large, meant that the PAL application was a mammoth task that bore yet more tasks in turn. The DNA team continually maintained close links and a great work practice with Ben allowing them to keep up with the ever ballooning brief.
Beta late than never
The decision was taken to present the development of the site to the wider community as a Beta launch. This served many purposes. Firstly, it allowed us to get some amazing in-depth user testing, from testers who often knew features of the site that no one else was aware of; and secondly, it gave us the opportunity to engage in a conversation with them - allowing them to comment, suggest and complain.
The beta launch went better than we could have hoped, with some amazing feedback, both positive and constructive criticism - which allowed us to make the project greater than the sum of its BBC / Aerian Studios parts. However, as you can imagine, it did put us back yet further!
It's hard to talk at length about how much difficulty we had integrating with the API's provided because so much of the problem was hidden in the details. It's also hard to explain to an outsider how hard Sam and Chris (and Ben Poole from our end) worked in making sure the project was delivered with the quality it deserved - I hope that you see for yourself and it's apparent to all.
For me at least, H2G2 was like a night out: it went on a little too long, it was sticky and uncomfortable in places and your head is still spinning - but you wouldn't change it for the world.
So you'll have to excuse me, I'm off to concoct another Pan-Galactic Gargle Blaster to celebrate the new look of an old friend.
Cheers - to H2G2!
Paul Goodenough is Managing Director of Aerian Studios - a mostly harmless web agency