Jakob Nielsen is a smart guy. The godfather of web usability knows his stuff. You only need to read his column (www.useit.com) to know that he knows he’s smart. He tends to where it on his sleeve, but he’s earned it – and I enjoy reading it.
As far as usability goes, he’s the king. He’s trying to be a real intranet guru, but he really has not proved he can move beyond usability. He over-emphasizes usability – which is an important aspect, but not the most important.
In my years of working on intranets, I have calculated the value of usability at about 15% of an intranet’s value – with content and planning & resources (governance, process, publishing, funding, staffing, etc.) representing 30% each.
Usability however has its place and Dr. Nielsen has written a great piece on standard deviation and the number of users to test when doing usability testing (see Quantitative Studies: How Many Users to Test? at ww.UseIt.com). In short, Nielsen says that testing “20 users typically offers a reasonably tight confidence interval.”
I’m no expert in usability testing (I leave that to Nielsen and to Terry Costantino and gang at Usability Matters), but I’ve done usability testing enough to know that 10 user tests is plenty for the vast majority of sites. Not all, but the vast majority. After 10 user tests you’ll likely learn very little – there’s a great deal of diminishing returns after 8 tests. Try it you’ll see.
So, if I’m right (just pretend), and an intranet’s value is mostly content, as well as planning & resources, then performing 30 or 40 minutes of usability testing with 10 employees has extremely little value. In fact, try preparing a business case to redo your intranet based on 10 or 20 usability tests.
My point is not to call out Jakob Nielsen for his shortcomings – his value to the web profession is profound and rarely challenged by few others – but to highlight the overstated value of usability. I’ve seen far too many RFPs in the past year that focus on usability testing as a means of redesigning a website. Usability testing is for enhancing and improving site navigation and information scents – often best used during the prototype phase. It is not a tool to lead the redesign of a website or intranet. This requires careful assessment and planning and attention to content and information architecture.
Save the usability testing to refine your navigation and key information paths – 10 user tests is usually more than enough. Don’t emphasize usability too much – particularly if you’re building a business case for a redesign – but don’t overlook it either; usability is a critical component, but it shouldn’t get more than its fair share of the budget.
