A redesign recipe for tough times
By David Walker (Google profile)
Web site redesign work has never attracted more interest. The recent boom has ensured that most of the Web sites that are needed have already gone up. The end of that boom has deadened any sense of urgency among organisations still planning their first site. And it's a rare Web site built in the past four years that couldn't use substantial improvement. So redesign is in vogue. When Kelly Goto and Emily Cotler published a new guide to Web site project management in August, they called it "Web Redesign: Workflow that Works" - misleading, but probably a sales winner.
Finding the resources for a redesign takes more work than it used to, though. Increasing economic gloom and well-justified scepticism about IT spending are squeezing site budgets. Even at the wealthiest sites, managers are having to sub-let the in-house massage facilities to insolvency specialists just to fund a few more fancy graphics. How do you redesign in times like these?
- Clarify your mission. Find out what the site's organisational sponsors expect it to do in the next year or two, and align your redesign to that mission. Chances are that the sponsors' ambitions are both clearer and more sober than they were way back in October 2000.
- Gather your data. A redesign has one great advantage over a greenfields site: users are already visiting it. They're carrying out transactions, leaving records in server logs, sending emails, talking to call-centre staff, commenting to management. You can find what's working and what's not. You can also put reliable figures on what it's costing. That frenzy of site-building between 1997 and 2000 was often justified as necessary experimentation. Now it's time to look at the results of the experiment, and keep looking at them.
- Redesign by inches. The big, showy launch of a new design offers maximum risk and makes a big, sudden hole in the budget. A succession of smaller changes can be managed and measured more easily, and each success will earn you the right to make the next incremental improvement.
- Aim for provable short-term gains. For instance, work out how much you'd gain from site navigation improvements that led ten per cent more people to buy the site's most popular products or reach its most popular content. "The way to sell a year and a half ago was to talk about getting a return on your investment after 12 months or 18 months," reflects Netscape co-founder and current Loudcloud chairman Marc Andreessen in a recent New York Times article on corporate Web site spending. "Nowadays, the arguments have to do with cost reduction in the first month."
- Test key features of your redesign against what's there now. If you're promising ten per cent improvements in transaction rates, ensure the redesign achieves that. Even a half-dozen test runs with friends and volunteers off the street can suggest whether the new design improves noticeably on the old.
- Trial your redesign. If you can, put a version of your planned changes in a corner of your site and track any changes in activity there.
- Find cheap wins. When a site-wide redesign is beyond your existing budget, adopt smaller projects which can produce demonstrable gains for minimal cost. They'll keep the momentum going within the site team and among the site's organisational backers.
- Design for maintainability. New graphics on the site every week costing you a packet? Try replacing them with text-only displays in a new design, and monitor the results. (Swapping text for graphics has been an accepted strategy ever since IBM slashed graphics and download time and simplified navigation in its 1999 redesign - a project which allegedly helped boost sales 400 per cent.) Where you cannot give up GIFs and Flash files, consider a design which lets you automate their creation, using a database tool like Macromedia's Generator. In most cases, your redesign should allow just about every piece of site content to be generated from a well-designed database structure, even if it's generated merely as static HTML.