Standard information architecture (IA) focuses on how to best structure material within a single site, while Micro-IA focuses on dividing a single topic's content across several pages. With Mega-IA, the question broadens: How can you best distribute content across your website(s) and outside services?
We're currently running user studies of social media. One of our findings is that the mega-IA decision to host some services on external sites can create usability problems for your customers.
Bad Example: Massachusetts Governor's Site
The Massachusetts Governor's Office hosts its videos on YouTube, resulting in reduced usability. As the following screenshot shows, the same video is represented thrice in the UI, eating up precious pixels and forcing users to scroll if they want to see more video options.
Countless usability studies show that it's bad to duplicate features on the same screen because it often confuses users, creating uncertainty about what they've already seen and whether the two seemingly identical features are actually different. Of course, triplicating features is even worse.
The MA Governor's site also features poorly chosen thumbnails :
Judging by these thumbnails, 3 of the 4 videos are almost identical. The Web is all about choice; video links must differentiate the videos and help users choose which ones to watch. The 3rd thumbnail — representing the gripping topic of the gubernatorial hip — doesn't visually indicate the content at all.
It's hard enough to make good thumbnails for still images (mistake #4 of the top-10 Web design mistakes of 2003). Videos are even harder to depict in thumbnails. Still, it's much better to choose an image based on editorial criteria than to let YouTube randomly pick a frame.
The above thumbnails make the videos look supremely boring. Most likely, they are boring, but most people won't ever know that because of the viewer-repellant thumbnails. Downplaying talking-head videos for online viewing is an obvious and long-established guideline.
As this example shows, just because YouTube is hot doesn't mean that you'll get a good multimedia site by pouring your video content into one of its templates.
Better: Use Categories to Guide Users
Let's compare the YouTube presence for Martha Stewart (cooking and homemaking guru) and Harvard Business Publishing:
Martha Stewart has much better thumbnails than the MA Governor, though admittedly her topics are far more photogenic. However, users get no help here (other than search) to find videos on a particular topic. Relying solely on social features such as "most viewed" and "most discussed" provides insufficient IA support.
Want to make blueberry pie? Prepare to scroll through 126 videos.
Harvard Business Publishing shows a starting point for better structure, dividing its video into 3 categories according to topic:
Of course, the thumbnails offer no help in differentiating the categories, unless you think that close-up headshots are more strategic than torso shots.
The underlying problem is that YouTube's features are designed for casual social networking, and topic-focused websites need more options for organizing, categorizing, and presenting content.
You can improve YouTube usability somewhat by writing better titles and descriptions that comply with our writing for the Web guidelines. Comparing the above examples, the MA Governor site's writing is clearly the worst, further muddling the user experience. "Governor Patrick Announces the Special" — special what? It's miraculous that this zero- information-scent video link collected a whopping 49 views from among the state's 6.5 million residents.
The Risks of Outsourcing
Clearly, usability suffers when an organization puts its website content on social sites without adapting it to the particular site's features. Usability suffers even more when those features are optimized for a use that differs from what the company's customers need.
These arguments count in favor of keeping social features on your own site where you can design them to provide a better user experience for your customers. By doing so, however, you give up on the potentially much bigger audience that 3rd-party social networking services (SNSs) offer.
Beyond usability, outsourcing social features entails an even bigger, strategic risk: by relying too much on external SNS sites for your social features, you might be handing them the keys to the kingdom.
Websites have already lost much of their value to search engines by making them the entry-point to the Web's riches. When people have questions or want to accomplish tasks, they turn to their preferred search engine — which is why search engines collect billions and content sites collect peanuts, despite the fact that they're the ones actually satisfying the users' needs.
If websites train users to turn to a handful of SNS sites for the next level of Internet activities, history will likely repeat itself, further diluting the value of those sites that actually produce content and services.
It's too early to provide definite advice for how to resolve this conundrum. One possible approach is to feed the outside sites only broadly targeted material that might go viral and/or attract casual browsers, while keeping higher-value specialized material on your own site, including any action-oriented and need-to-know content and discussions. The broad material can then drive traffic to the specialized content, as can e-mail newsletters and other standard tactics that foster loyalty to your own site's services.
While the big picture is still unresolved, there's no doubt about the everyday design implications of hosting any material on other sites. Don't just use the default template, and definitely don't assume that simply dumping stuff onto a popular SNS will automatically make you popular. As always, design for usability — both in terms of everlasting guidelines for things like IA and writing, and in terms of newer guidelines for using newer features.