FOWA 07: Khoi Vinh – Managing User Interface Design

(New York Times)

Khoi Vinh and his team don’t design the news, don’t do the illustrations – that’s done by visual journalists on the editorial side. They instead do the platforms, frameworks, templates.

Contrast with print design, which has had some art direction, and the online version of the same story which is just a standard design that looks identical to other articles on the website. Using templates. Trying to innovate at the template level.

Too difficult to actively design the news because the tools cannot keep pace with real time. Internet allows instantaneous publishing, but it does not yet allow for instantaneous design.

One of the projects is the NYTimes.com/travel. Old travel site was very standard, just journalistic content. Now using travel tools, booking tools, rating tools, user generated content. Design at the user experience and template level.

Working with photo editors, talking about how they want to display slideshows, do captions etc.

Occasionally they do go to extra lengths to develop customised layouts for the home page, e.g. to mark the 5th anniversary of September 11. This is highly specific design presentation, and the way we should be designing. But it’s time consuming and not currently compatible with syndication models. Even though they are using CSS and HTML, they are foisting presentation layer into the RSS which it’s not used for, so it breaks.

Special projects with sufficient lead time, e.g. elections in 2006, developed a number of templates for displaying the information. But need a lot of warning so that they can get the templates up.

One day, our tools will evolve such that we will be able to design in real time. This will be a new kind of publishing, one that will be dramatically more mature than what we have today.

Technorati Tags: ,


Content and Functionality
One of the biggest things they are trying to do at the times is move away from being just a platform for new delivery, and more towards being a platform for news-centric interactivity. Adding lots of functionality, thinking about community, changing role for designers. Users no long just consuming the news online, but have a totally different mind-set. Different role for graphic designers. Traditional media uses narrative design, but that’s changing. Communication on the web is about conversational design, where the platform enables dialogue with the audience, and the design is part of the conversation itself.

For years it was a one-to-many relationship, and now it’s many-to-many, and the NYT is trying to find its place amongst these user and peers.

Changes mean that users demand new functionality.

Building applications:
– MyTimes, RSS feeds
– TimesFile, hoping to launch in a couple of months. Can save and tag articles, a bookmarking tool.
– TimesTopics, encyclopaedia online of Times Articles about people
– TimesReader, more controlled, more visual experience of reading the news.

Content is evolving into functionality. Dispersing this functionality throughout the site, e.g. an inline audio player that allows people to listen to MP3 conversations with journalists and other content that goes with the article. Previously would have sent you to a different window, but all done inline now. Putting video inline with articles too. Giving people more varied journalistic experience.

New sharing tools, including Digg, Facebook, Newsvine. Can tag it right there on the page without having to go to another site. The Permalink icon gives blog readers long-term access to an article long after the article itself has gone behind the paywall, so when you link to it the link doesn’t break. Not a feature they shout about.

Search engine optimisation. Trying to make sure that content is visible. People end up arriving at site through Google. [Not that this is new!]

People talk about a tension between editors and users. Parallel tension between designers and information consumers. Issue around control. As tools progress, they give more power to control the presentation of our interfaces.

HTML was poor language to realised the designers vision, added hacks to do that, tables, Flash, CSS etc. Can realise the design vision much more easily. Users are wanting to assume this control for themselves.

Site where you can create your own user experience. Not necessarily good design, but the fact that design is disseminated.

Contradictory trends
High-def vs YouTube
Skype vs SMS
Times Reader vs Memorandum
Digital SLRs vs Camera phones

People embrace both higher quality choices, such as digital SLR, as well as the lower quality or otherwise restricted options available.

Web pros can’t get enough Web 2.0, and willing to cede control to have those features. Users have no idea what Web 2.0 is.

NYT are not early adopters, but selective adopters. Figure out what the basics are as to what the customers want. Adding too many features frustrates users.

Getting it done
Management is the “art of getting things done through people”. Design management is “creating conditions under which great design can happen”. Have to look at that responsibility and understand that the main role is creating the right conditions.

Principles for Design
No such thing as a free feature – thinking of applications as software masks the cost of functionality – free isn’t really free. People want to throw stuff up just to see how it behaves, but have to look at the hidden costs:
Additional:

– interface constructs, screens, widgets
– code
– testing and QA
– long term support
– feature noise – everything on the page is something that users have to evaluate and then tune out if it doesn’t make sense to them. Have a home page already packed with information so need to economise on that.

Cost of expression
– content authors bear almost all the cost of expression in traditional media
– in digital media users bear at least half that cost and the more complex something is the more people turn off.
– Software is like a physical machine, which emphasises cost of functionality.

Over-determination is good
– Every feature should have multiple valid reasons for being.

Options are obstructions
– Every extra setting or preference is something that isn’t resolved in the main interface, and it’s an obstruction. Unresolved features.

Offend experts, not beginners
– Don’t want to offend anyone, but if they do it’s experts. Beginner/intermediate confusion is more important to resolve than expert inconvenience. Beginners are more likely to just walk away from confusion immediately. Most users are intermediates, most features are for experts, so try to design for the overlap.

Navigation within reason
– Don’t need to get everywhere from everywhere. Amazon has 36 product categories. NYT has a ton of navigation and are trying to reduce it.

Test like you mean it
Ue real users, not executive. Important to get their feedback on products, but they are not normal users, and by using them you privilege them over real users. Test for usability not acceptance.

Writing is interface design
– Language informs interaction. E.g. Twitter asks “What are you doing?” and when you take that label away it changes the meaning.

Let tabs be tabs, let buttons be buttons, let links be links
– Thinks should be what they look like they are. Shouldn’t try to redesign things just because you think they are a sort of cliche. Shouldn’t have to interpret the interface.

Design, don’t decorate
– Wants to be clean, efficient.

Context over consistency
– Within a family of like interfaces, avoid monotony, embrace variety.

Use a grid
– Makes it easier for users to know what is where.