Friday, February 5, 2010

#onw2010 Library in Your Pocket: Online Northwest Conference

Kim Griggs, Hannah Gascho Rempel, (and Laurie Bridges) of Oregon State University

Show of hands indicates few libraries represented have mobile sites now.

Background data
Over 4 billion mobile subscriptions worldwide, but only 1 billion land lines. 85% penetration rate in the US, but it's higher in many countries. 51% of undergrads own a web-enabled phone, and another 12% plan to buy in the next year. 250 libraries use TextALibrarian. EBSCO, IEEE, and PubMed are the databases with a mobile-optimized option.

OSU Experience
They began by looking at best practices from W3C. They recommend that mobile access gets its own design - desktop/laptop design doesn't work on feature phones or smart phones. The site gets shrunken and difficult to navigate. This is especially problematic when big real estate is taken up by non-essentials (like new exhibits). Mobile users have specific interests - more immediate goals for specific pieces of information.
OSU also checked other websites, from both libraries and commercial firms. Mobile sites are not just miniature desktop sites. Mobile sites should have many fewer graphics, and more linear structures - avoid multiple columns and put the most important stuff at the top. Mobile sites have less information total. Consider how far the user is willing to scroll. Consider the differences between a feature phone and a smart phone - the optimization is different. North Carolina State University Library has an excellent example of both.
OSU saw the problem as the fact that their patrons were using their mobile phones to access the usual website, but that was cumbersome. The solution was to design a mobile-specific version.
They started with scenarios for three user groups: students, library staff, and the general public.
Based on the scenarios, they decided to include:
  • Time saving applications, like floor maps linked to call numbers
  • Location sensitive information
  • Native capabilities, like autodialing or videos
  • But not everything.
They broke the project up into stages, and defined their users. Then they identified which content to make mobile by looking at analytics, and through a poll on their desktop. They highlighted fast, easy, and fun content that would be easy to transform.

They also decided which devices to target. 85% of use is on feature phones, particularly with international students.

Key content
  • Hours
  • Ask A Librarian (chat, email, ref desk hours, phone, SMS)
  • Address & phone (with Google Maps)
  • How do I?
  • Where is it? (floor maps)
They then returned to scenarios. A main goal, indicated by the polling, was to connect the catalog to physical location. They also linked the available lab computers to a map, so users can figure out where to head to use a desktop right now.

There's also a mobile specific catalog interface, including reserves. The results privilege title, location (including call number), and availability. It gives the option to text or email a record.

User studies indicated that 100 users per day stayed for 4 minutes. iPhones were 75% of users.

Mobile specific URLs - multiple, guessable addresses - help users find the site, as did mobile directories, library news, and the home page.

To develop a mobile presence:
Consider the available resources.
1) Do nothing.
2) Miniaturize using CSS.
3) Create a mobile-specific site.
A mobile-specific site has a 64% success rate. Users on mobile devices have the same limitations as users with disabilities on the standard site.

Adaptation, or multi-serving, delivers data for each device, although standards are converging. Media types in CSS offer "screen" and "handheld", both of which are used by different mobile browsers - ie, Blackberry is handheld and iPhone is screen. Screen supports more images and more complexity.

php detection algorithms can direct mobile users from the desktop site to the appropriate media type. Don't, however, force the user - allow a way back to the main site, and a way to remember that preference.

An XML file called wurfl at SourceForge can let you know the exact capability of the user's phone, which helps direct them appropriately. Allow feedback from users to improve the directions.

Content can auto-adapt with open source tools, like Wireless Abstraction Library - a mark-up language to use with wurfl. It avoids the requirement for separate copies.

There are mobile site conversion tools for libraries with less programming capability:
  • Google Conversion Utility
  • MobiSiteGalore (free)
  • InstantMobilizer
  • Drupal Mobile Tools
Another option is to use RSS feeds.

For iPhones, create a specific stylesheet. Use the tag that creates a iPhone link to add to the user's home screen.

Design Recommendations
  • No more than 3 clicks
  • Access key number for each navigation link (so no more than 10 links per page)
  • URLs short; only alphanumeric
  • Don't rely on Javascript
  • Minimize typing with check boxes and radio buttons
  • No image >80% of screen size
  • Use tried and true patterns
  • Continuously valid code with W3C mobileOK test

Since testing on every phone is impossible, start with Opera and Safari. (Be wary of online simulators.) Downloadable simulators from the iPhone, Blackberry, etc are better. You could even ask for help from a retail phone outlet - they might let you use the phones in the store.

Future services OSU plans to add include databases, room reservations, and Text A Librarian.

Q: How much content revision was required? A: Hours is based on a database, so no change, but much else was drastically trimmed.

Q: What about mobile design suggested changes for the main site? A: Nothing yet, but it probably should.

Q: What was the time frame? A: First stage: from December 2008 to March 2009; second stage by December 2009.

No comments:

Creative Commons License
Libri & Libertas: Books & Freedom in a Web 2.0 World by Laura H. Wimberley is licensed under a Creative Commons Attribution - Noncommercial - Share Alike 3.0 United States License.