Six things I think I think about UIs

10.04.2006
Software without a good user interface is -- well, it's useless. Unfortunately, UIs aren't nearly as well understood as more objective "speeds and feeds" kinds of product characteristics. Products with snazzy demos often prove disappointing in real life. Meanwhile, truly good UIs are hard to explain beyond an "I know it when I see it" vagueness. But difficulty notwithstanding, UIs are too important not to write about. And so, to steal a line from sportswriter Peter King, here are six things I think I think about UIs.

1. A good graphical user interface is the most important feature a product can have. In many cases, the GUI is the feature set, whether we're talking about operational applications, business intelligence or IT administration tools. For example, the general user analytics market is dominated by friendly BI tools. More powerful products exist, but they're mainly used by specialists, since everybody else is turned off by their clunky GUIs or command-line interfaces.

Equally dramatic is the importance of GUIs in IT administration. Check Point Software Technologies rose to dominate the firewall market because it offered a usable administrative GUI back when other products operated from command lines. Similar phenomena occurred in database administration tools, network administration and many other sectors -- and would occur in more sectors if vendors would wise up. (I've been nagging the BI industry for years about developing a stronger capability to manage alerts and custom key performance indicators.)

2. Web UIs are now, finally, much superior to the client/server systems they replaced. Usually, changes in computing platforms lead to improved user interfaces. More powerful mainframes and cheaper minicomputers allowed real-time apps to replace batch systems in the 1980s. Client/server GUIs blew out character-based apps in the 1990s. For years, Web technology was a partial exception, since it actually caused a retreat in GUI flexibility. But the group of technologies collectively labeled AJAX, has finally restored client-side parity -- or, if you like multimedia, even forged ahead -- and Web apps now boast top-tier GUI functionality. What's more, they have major application navigation advantages, via hyperlinking and search, over anything that came before.

3. BI look and feel is on the upswing. One UI area that stagnated for years was BI. Not only was it hit by the transition from client/server to the Web, but vendors also spent years upgrading their server-side infrastructures. Now, however, we're beginning to see UI advances again. Some of them are subtle, such as the care being taken to optimize the use of screen real estate. Others are flashy, such as the long-overdue mainstreaming of some cool data- visualization technologies. Either way, UI is once again an important decision factor in selecting a BI vendor.

4. Portal technology is headed for a boom. "Portal" might seem like a somewhat passe Internet bubble buzzword, but portal technology is actually very important in at least three ways. First, it's increasingly central to application navigability. Second, it's an increasingly important BI underpinning (both SAP's and Oracle's analytics are portal-dependent). Third, it provides the integration framework to combine operational apps, analytics and collaboration technology in new ways.

5. Natural-language interfaces are advancing too slowly. Unfortunately, big vendors remain clueless about language-based UIs. Enterprise search is a fiasco. Most single-site Web searches are even worse; in almost every case, they're inferior to just Googling on "search string + site name." Serious natural-language/voice command/ control and navigation should be well established by now, but in fact they're barely a blip on the radar screen, InQuira and Sybase Answers Anywhere notwithstanding.

6. Microsoft Office is a huge question mark. Microsoft Office is at a crossroads, with a mounting open-source challenge and a justified reputation for unpopular feature bloat. Microsoft's main strategy to reverse this trend is to make Office the gateway to database information, rather than just to static documents. Much of the action falls in two areas -- BI and live XML. The first fizzled query-in-Excel effort dates back to about 1993, but there's reason to think it will be different this time. The BI industry is making Excel ever more viable as a core analytics client by delivering reports straight into formula-rich Excel spreadsheets, for example. Or maybe not; after all, those same vendors are trying to undermine Excel by replacing it with more heavyweight budgeting/planning systems. And, as I've noted before, live XML is a great idea in certain niches, such as smart forms or complex contracts, but whether it will ever go mainstream remains to be seen. Interestingly, as we went to press, Microsoft announced that it plans to acquire BI vendor ProClarity.

Curt A. Monash is a consultant in Acton, Mass. You can reach him at curtmonash@monash.com.