Duke University unveiled a new, attractive, easy-to-use web site with several additional features last week. And David Jarmul lived to talk about it.
How? One particular reason, the director of Duke News Service and associate vice president for news and communications told Local Tech Wire, was having a defined mission.
“It’s so easy with web projects to get carried away with arguments about Java script, or design or some other technical issue,” he recalled about the process. “What you need more than anything is a clear vision of what you’re trying to accomplish, and why, and how it fits with everything else your organization is doing. Everything else is details.”
LTW sought out Jarmul for its weekly Executive Q&A because of the impressive job Duke did in totally revamping its site and also the business processes employed in the project.
We also asked Jarmul for his own “Top Ten” list of do’s and don’ts. Be sure to see the sidebar story: www.localtechwire.com/article.cfm?u=3526
What were the factors that drove the decision to redesign the web site?
Duke’s web site was more than six years old. The design was out of date and the information was poorly organized. There were long lists of “stuff,” in no particular order. Both editorially and technically, the site wasn’t working. For a top-rank university like Duke, that was a real problem. Increasingly, the web is our front door to the world — to prospective students, parents, faculty, journalists and others — and it needs to portray the intellectual energy and incredible diversity of Duke.
Were these concerns driven by your users/customers rather than simply adding new technology and features because they were available?
Our goal was to serve our audiences effectively, not to use new technology. In fact, we consciously avoided some options, such as Flash on the home page, because some of our users are still using dial-up modems from home or elsewhere. Also, some alumni and others are new to the online world, and we wanted the site to work for them, too.
Why did you start from scratch as opposed to just remodeling?
We had two main goals. The first was to provide better navigation for the mass of information at Duke about everything from athletics to biomedical research. We sorted all of this information through the eyes of our consumers — applicants, faculty, parents, donors and so forth — rather than according to our organization charts. Our other goal was editorial. We wanted the site to convey a sense of Duke’s mission and ideals, serving as an online commons for the community. There was no way to accomplish these goals within the previous site.
You reached across numerous departments to involve them in the process. What were the benefits of doing this?
I can’t overstate the importance of this. Before we wrote or programmed anything, we met with people and groups across Duke. We met with the admissions office, the student affairs people, the development people, the international people and so forth. We asked the tech-transfer office what companies need to know about Duke, and graduate students about their special needs. We brought together the communications people from the law school, the medical school and so forth, and sought their advice. As we designed the new architecture, and then the pages, we checked regularly with the appropriate people. We recruited beta testers to bang on the new site and, after we incorporated their changes, we made the site available in a pre-launch phase for anyone to offer suggestions. Finally, after receiving hundreds of messages, all of which we read and considered, we went live in February.
What were the costs? (How did you avoid took many cooks stirring the pot?)
Our budget for the whole project was about $50,000, not counting staff costs. We spent the largest share on a wonderful freelance designer, Elizabeth Kairys, and the next biggest piece for a freelance writer, James Todd. We also hired help with the rollout campaign, and bought some photography rights from our talented university photo office. The rest we did ourselves, over a period of a few months. Our web manager Ben Riseling, who was previously at salon.com, took the lead on the technical end. Two of our writers, Dennis Meredith and Blake Dickinson, oversaw the information architecture and writing.
I spent a lot of time working through political and administrative issues. We were the core team although, as I mentioned, we also got terrific support from the rest of Duke’s communications community and elsewhere, from the president on down. We also have a great working relationship with our IT department, with whom we’ve partnered on several big projects.
You have added some new features – the e-mail newsletters, a new calendar, and a news site. What were the reasons for adding these?
We have three new e-mail newsletters, which we’ve branded under the name eDuke. One provides a daily news feed from the university. The second has online clippings about Duke, so people can see what the world is saying about us. The third is a monthly newsletter that people can customize according to their interests. Thousands of people now subscribe to one or more of these newsletters, which pull together information that was previously hard to find. The same is true of our new online calendar. Before, Duke had an arts calendar, a sports calendar, a Chapel calendar and a this-or-that calendar. Now people can find lots of different events all in one place. Our news site, finally, replaced one that was derived from the printed news-release model. We overhauled it to be more helpful for reporters and others.
Duke is a private enterprise. What lessons can other private enterprises and/or educational institutions learn from the process of the Duke project?
It’s so easy with web projects to get carried away with arguments about Java script, or design or some other technical issue. What you need more than anything is a clear vision of what you’re trying to accomplish, and why, and how it fits with everything else your organization is doing. Everything else is details.
What mistakes were made in the process that you can share with others, and how can those be avoided in the future?
To tell the truth, we’ve been pleasantly surprised by how smoothly this process has gone, which I attribute mainly to the support we received from the president and other senior officials of the university. There was a lot of pent-up frustration with the web site at Duke and, although people didn’t hesitate to tell us when they didn’t like some idea, we encountered little resistance to the effort overall.
The content management system seems designed to capitalize on the skill set of editors who edit/write and developers who write code. What factors led to this decision?
We haven’t solved the content-management issue yet. Most of the new site was written in plain HTML. Over the past several months, though, in concert with our IT colleagues, we’ve carried out a campus-wide review of content-management systems and plan to buy one to serve everyone from the library to the engineering school, with our operation included as well. We know CMS is going to be a big part of Duke’s web future, helping us to process and share information more easily.
Duke News: www.dukenews.duke.edu