ia play

For most of my BBC career, the website wasn’t really about TV & Radio. Discussion were filled with talk of news, sport, weather, recipe finder, GCSE Bitesize, initiatives like iCan and products like Connector, MyBBC, and search.

There were lovingly crafted programme showcases for the top shows like Eastenders and Doctor Who. And there was Radioplayer, well ahead of its time.

But most programmes had no coverage (a temporary schedule snippet notwithstanding) as I discovered in my first few weeks at the BBC. Part of my job was to respond to users who had emailed us via the ‘contact us’ link on the search engine. Query after query asked for information about a programme recently and not so recently seen or heard. We resorted to back catalogues of RadioTimes and lots of apologetically framed replies.

Now the situation is somewhat different, with a number of projects having re-homed programme content on the internet, mostly notably:

• iPlayer 7 day catch-up, TV and Radio
• Catalogue (currently offline) Text based records of the back catalogue, based on the BBC’s internal catalogue produced by Information and Archives
• Archives Trial collections of archive audio, video and written material
• Programmes A page for every programme from October 2007 onwards, some with embedded audio & video from iPlayer

Diverse teams tackling the original problem (no programme support) from slightly different angles and a more experimental, innovation-friendly culture has resulted in an information architecture headache. Part one of solving the resulting problem is integrating the data from Catalogue into Programmes. I’m sure that’ll be a cinch 🙂

Michelle asked for resources when starting out with taxonomies and I’d coincidently been compiling some stuff for other nefarious purposes, after have to search for the same old set of stuff for the umpteenth time.

So I’ve made a page of CV resources. There’s the basics now but I’ll be adding more.

Another striking search/browse experience is the new Times Archive. I haven’t spent so much time looking at old newsprint on screen since the weeks I spent scanning microfiche for old new stories for Learn.co.uk

I really quite like the timeline on the homepage, which surprised the curmudgeon in me.

TED.com just won a Webby for best navigation.

Now TED is a great site to while away hours with and still feel like you are improving yourself. I’ve spent lots of time there and I’d be struck by the navigation but not always in an entirely approving way. As with most new and exciting attempts at navigation there’s something fun about it and also something kind of irritating when you need to get something done.

Good

• traditional but useful text-based left nav combined with the more attention grabbing main content
• lots and lots of different ways to slice the content – themes, tags, speakers, popularity, most inspiring and so on
• choice of list or ‘visualisation’ view

Bad

• in the ‘visualisation’ view, the mouseover interaction gets in the way. If I try to select a promo in the middle of screen but move a little too slowly then the promos I am passing over pop up and obscure the one I’m trying to get to.
• illustrating sub-categories with pictures of individual speakers. Takes a while to realise that if you click on that big promo with a picture of Jane Goodall you get taken to another index with 36 talks on the topic ‘inspired by nature’ and not straight to the Goodall talk.
• surprisingly low-key use of the tags. No tag browse on the homepage or the themes pages (where I can really see value for drilling down further with them)

But going back and playing with it some more I think all in all that the good outweighs the bad. Would like to see that mouseover sorted though. It is a bit too much like the problems I have with the Windows Start menu(!)

I’ve been asked to come up with a quantitative measure of success for the information architecture of the BBC website. I won’t be solving this is one post.

Now I’m wary of this. We all know that what you measure goes up. There’s an anecdote (office myth?) about the days when site funding was tied to page impressions and some  (mis-guided) web producers redesigned their interactive quizzes so that each question was on a single page, resulting in higher page impressions.
At the same time, Martin’s been expressing doubts about the BBC Trust chosen metric for measuring the success of the BBC search:

“They state that internal referrals from the search engine are down to 19% of all search referrals from 24% the previous year. Now, of course, there are lies, damned lies, statistics and then web metrics, but I’m unclear that you can argue how ‘good’ or ‘useful’ a site search is from these figures.

People tend to use site search when they are lost or disorientated, not just when they are seeking a specific piece of information. You can use exactly the same figures to argue that nearly a quarter of people used to get lost on the site and had to resort to search, and now only a fifth do – that could equally suggest an improved navigation user experience rather than a deterioration of search quality.”

The attention from the BBC Trust on site search is helpful and correct. But the success metrics need to be chosen carefully else we could genuinely improve the quality of the search and still get marked down as having failed (likewise we could fail to improve the quality but the chosen metric might improve, resulting in pats on the back all round but no improvements for the users). So this stuff is v. important to get right.

But back to measuring the quality of the IA in general…

Our key metrics are reach, impact and appreciation i.e. lots of people spend lots of time on the site and like it so much they tell lots of other people that it was great.

Getting the basic IA right should reduce time spent on site as it would get people where they wanted faster. They would then be appreciative and tell other people, but their time on site might well go down.

But we could make it easier to get around and increase time spent by cross-selling – providing a clear contribution in meeting their expressed needs but also in showing them what else we have that they didn’t know about.

That bit is important because some of the research commissioned for the 2004 Graf review showed that members of the public who took part in the usability tests (and were hence shown lots of the BBC site) were angry that all this free content that their licence fee had funded was there and they didn’t know about it. Cross-selling is a public service duty not just commercial good-sense.

So good IA would mean short (what does this mean?) journeys to each piece of content AND a high number of pieces of content found (and used? and liked? in a single session?).

And what if they get BBC content elsewhere, in some syndicated form? Surely we’ve got to include that too? It might be different with each platform too…the potential for cross selling on mobile might be more limited, given the context of use and time people want to spend looking at content on their phones.

I think around 12 IAs have had to manage our metadata system at one time or another.

One was not bothered by it, had no problems with the work. One other person found it satisfying and actually interesting. Everyone else seems to have found it limiting, frustrating, boring, degrading even. In the admittedly limited frame of IA, wireframes are sexier.

Maybe I’m odd but it was a task I found flow in. There was a rich repository of data to analyse, procedures that could be honed to perfection and theory that could be drawn upon. There were side benefits of learning new words (ungulates?) and watching the English language evolve (house-blinging?). It felt like a craft.

Now few of my colleagues were interested in what I was doing day-to-day but that had the benefit of no-one else meddling with it. So my success or failure on any given day was down to me. There’s a certain pleasure in that.

I also, to a reasonable extent, built my career on it. My first presentations and published articles were all formed by insights from being immersed in the metadata systems. Other people were working in the same space but for the most part they weren’t the same people who were standing up at conferences and talking.

So find it boring, by all means. But there’s opportunities there for the taking.

I’m fed up with the rail ticket machine and typing in my reference number.

Why would you use a Qwerty keyboard layout on a touch screen? How many people interact with them by putting both hands on the touch screen?

I was making a hotel booking recently and got stumped by the currency converter for a while. The UK currency was listed in the alphabetical list as ‘pounds sterling’.

Whilst technically correct, UK pound or British pound are more normal in these lists, giving precedence to the country name rather than the currency name. With most countries that makes finding the relevant currency easy but it is still a bit tricky when dealing with the UK as you never know whether it’ll be listed under U or B.

So it would help not to throw P into the mix as well.

In the midst of Jared Spool’s gadfly-esque “UCD never worked, maybe we should retire it” opening plenary at the IA Summit was a rope prop. The rope represented the number of visitors to a website. It had a knot to show the small percentage of visitors who are customers and another for the even smaller percentage of big spenders/ultra-fans. Jared suggests that a smart business person doesn’t worry about the huge percentage for whom the website clearly isn’t working and just focus on selling a little bit more to those ultra-fans.

Fair enough.

But oddly not everyone in audience is just trying to make lots of money. I was reminded of this in a recent presentation from the BBC’s marketing team about audience segmentation. Like any other organisation we have a segment of highly passionate fans but, as the presentation made clear, the BBC cannot just up-sell to our the high approvers. That’s not public service.

So BBC IAs have to think about the whole rope, I’m afraid.

At the BBC we divide our large UX team into a number of fine grained distinctions: visual designers, interaction designers, information architects and usability engineers. The reality is that the actual people in the team don’t fit neatly into this divisions, even if you could come up with a clear definition of the differences.

I’ve been recruiting juniors recently and it has been noticeable that the applicants struggle with the differences between the various UX job titles.

At the same time I’ve been having a look at the job adverts on a few job websites & mailing lists (Monster, TotalJobs, Chinwag, Mad, Jobserve, London-IA, London-Usability) to see if there was any consistent connection between job description and job title. I went through 50 ads in detail and skimmed a load more.

information architect – far and away the most common job title (3 x more than the nearest rival UX architect). Every single job description asked for wireframing skills and only one didn’t mention sitemaps/blueprints. At least half asked for experience in working with multi-disciplinary teams (project managers, designers and developers), client facing skills and a pragmatic approach to balancing user needs and business constraints. Half also asked for usability testing skills, interaction design experience and persona creation. Where tools were mentioned (a third of ads) it was usually Illustrator, Photoshop & Visio rather than any particular package.

interaction designer – rarer than I expected, when this did crop up the job description was pretty much identical to IA. The rarity may indicate a loss of popularity in favour of user experience designer.

user interface designer – similiar to IA and interaction designer but with a slightly more technical angle, often including HTML, CSS and Javascript

user experience architect & user experience designer – very similar to IA and Interaction designer. The only noticeable difference was the remit of these roles often included ‘visual design’ which I didn’t see once in an IA job description

usability specialist & usability analyst – usability engineer seems to have lost popularity as a job title and this appears to have coincided with a broadening of remit. These roles are very similar to the IA and UX job descriptions but with a greater emphasis on designing, conducting and analysing usability tests.

So essentially the job descriptions are very, very similiar for all these job titles. They *must* mention wireframes to be an IA, the UX prefix may widen the job description to include visual design, UI designer is probably more techie and a usability prefix will mean more emphasis on user testing. But these are subtle distinctions. A great deal more unites than divides.

Now job ads aren’t the end of the story but it is interesting that we’ve created so many different job titles and then essentially described them the same. No wonder the applicants are confused.