Business value migrates in the direction of the user interface. Experience shows that no matter how much upstream data collection, information transformation and intelligent integration occurs, the payoff in the experience of the client occurs at the user interface. Whether business analyst, electronic file clerk, top executive, system administrator or power user, the "Ah ha!" moment occurs at the user interface.1 Information delivery provides a last opportunity to synchronize the experience of the user with the business process that is unfolding in the IT system. This is a basic principle in thinking about system architecture and the design of a usable software interface.

This column is not intended to be a tutorial on the placement of widgets on a graphical user interface (GUI) but rather an inquiry - one might say meta-analysis - of how meaning and business value are created at the user interface. The business insight and intelligence delivered at the interface is often a function of substantial upstream data integration; however, the end result is still what gains the kudos. This leaves most of the IT staff - data administrators, DBAs, developers - as unsung heroes. And that is where this column proposes to leave them in order to look at issues of the usability of the front-end system interface to promote business value.

The best user interface is invisible. This means that it makes the context of the business process usable without causing friction. Functionally, visually or conceptually, it does not get in the way of the analysis of the business variables. Of course, this implies just how far, science fiction aside, we have to go to attain ease of use.

For example, updating or accessing data by clicking and dragging is superior to cutting and pasting or being forced to leave one application for another in terms of translating the user's intuition and intention into action. When the interface is smart enough - that is, skillfully deploys metadata under the hood to track actions - the business context is preserved and modified as the user drills down. Transparency is enhanced. Still, flexibility as a concept has its price, and an approach that is good for ad hoc queries may not be suitable for incorporating data into a rigorously formatted production report or performing analytics in a sequential ("in-line") process.

The visual display of quantitative information remains a priority in most business contexts. Best practices indicate that if a group of points in a chart is circled with the cursor, an action called "lassoing," then the context should be extended by the system to allow further analysis. When dealing with charts, saying the chart is "live," means that clicking on a data point will show the underlying detailed data for that point. If a map is displayed, it should be live, allowing zooming and drilling into related dimensions, including non-geographic ones.

While many scenarios exist in which a transparent and easily usable interface looks to a business analyst exactly like an Excel spreadsheet, even more scenarios do not. Because the business world is complex, the display of information must also represent complexity. Yet, it must do so in a way that facilitates comparison, contrast, calculation, selection and analysis of the data. Because this is easier shown than said, an example will be useful. Figure 1 shows the "before" display of a "before and after" example from Stephen Few's Web site. Figure 2 improves considerably on the design by labeling the bars with region names, eliminating the need for a legend. Most importantly, users often want to use the visual display to make a comparison between variables (key performance indicators) that are essential to an aspect of their business operation. The redesign makes it possible to do that at a glance by region, from best to worst, rather than having to disentangle the contrasting variables, in effect, off the chart in one's head.2 For yet another compelling comparison from a different design master of how the exact same table of railroad transit time data can be rendered with confusion or with clarity, see Envisioning Information by Edward Tufte (pp. 54-55). The separating lines in the table are formed by tiny dots, making a distinction but not a barrier and calming the dominating grid.3

Figure 1: Sample Graph Design

Figure 2: Sample Graph Redesign

One of the innovations that has transformed the user experience of (and market for) front-end tools and technology is end-user self service. Rather than give the user a fish in the form of a predefined report, give her or him a tool that enables the individual to learn how to fish for themselves. Viola! The IT department saves in administrative and support effort because the user generates his or her inquiries and reports without a formal IT department service or change request. Along side of portals, custom designed GUIs, in-line analytics and search engine style interfaces, end-user self service has expanded the realm of users (clients) wo are able to access corporate information making it even more apparent how valuable this increased access is. However, in spite of significant and on going success in the market, several limitations have emerged in the area of usability:

  • More service is needed by way of administration and support than was originally intended by self service - an estimated 20 percent of the users are sophisticated enough both to formulate questions and dig for answers - especially in a complex business scenario requiring more than looking up the balance due of a customer; while the remainder still require extensive guidance and direction. While IT saves in administration, it is sometimes "cost shifting" to another support function such as a special information center help desk rather than pure cost savings to the enterprise.
  • The idea was to empower the masses of corporate users to think originally. However, that is happening much slower than expected, if at all. In the meantime, the standard practice is to provide employees the information they need to do their job right now in the corporate cubicle culture where speed and quantity of output, not original thinking, is the order of the day.
  • Once an ad hoc, exploratory query has been executed three times, is it really exploratory? It should be optimized, tuned and rationalized in order to make sensible use of system resources (assuming the query is not a trivial).

Those organizations that decide they want to use the self-service query and reporting tool as the standard for all user interfaces - do not laugh, I was just meeting with such an enterprise - risk missing the business value of all that great information because only a small fraction of users are able to use the tool as intended in its full power. The vast majority still require support in getting to the "Ah ha!" moment in the form of a custom designed interface. What are the recommendations?

  • Usability testing becomes a priority. In order for business value to be given its due at the user interface, it is necessary to compare the steps in the workflow and business process with those visible at the user interface. This does not require a laboratory with one-way mirrors, cameras and recording equipment to capture the user's facial and verbal reactions as they extend from uncertainty through confusion to understanding. The same result can be attained with a sketch pad and yellow Post-it Notes, provided the users know the job and the testers are good listeners. However, the results require the ability to customize the interface to match the lessons learned in the order and sequence of presentation and operation.
  • No one wants to hear that they need to buy another tool. Yet in order to meet all of the requirements of their user community that is what is necessary. Considering the diversity of users, this will remain the case for the foreseeable future. End-user self service and enterprise reporting tools will usefully be complemented by those enabling custom design and rapid prototyping of in-line analytics. These approaches - self service, enterprise reporting, custom prototyping of analytics - are in many ways mutually exclusive. In addition, delivery mechanisms such as portals, fat clients, or thin clients cut across all three. Finally, while the market for self-service is arguably consolidating, the market for rapid prototyping is still getting up to speed. The prediction is that more vendor offerings in analytic prototyping would mean a bigger market for all the participants because it would increase the velocity of competition, marketing buzz, use cases and success stories. Stand by for update on this trend.
  • Power users will benefit the most from end-user self-service. Electronic file clerks and low-level administrators will probably find the most use in enterprise reports, whether delivered through a portal or other mechanism. System administrators will find themselves confined to the administrative workbench that accompanies the infrastructure tool they happen to be administering. A thin client over the Web is essential when trouble shooting a system at an inconvenient time or while either the system or administrator is at a remote location. Top executives will prefer dashboards, which can be delivered in a variety of ways. However, the breakthrough area for business insight will remain the delivery of main-line, hard-core, in-line business analytics.

In the final analysis, the vast majority of business analysts will find they benefit from an optimized interface that is designed and developed to match their particular business context and analytic requirements by a rapid prototyping tool. For the rest, there is Excel.
References:

  1. These roles are distinguished (in so many words) by Keith Gile, "Profiling the Analytic End User for Business Intelligence," June 26, 2003, www.forrester.com.
  2. For a wealth of additional examples see www.perceptualedge.com or one of Stephen's books such as Show Me The Numbers: Designing Tables and Graphs to Enlighten (Analytics Press, 2004).
  3. Tufts, Edward R. Envisioning Information. Cheshire, Connecticut: Graphics Press, 1990. This is a rich source of entertaining and insightful examples of what to do and not to do. Yet another relevant quote is "Simpleness is another aesthetic preference, not an information display strategy, not a guide to clarity. What we seek instead is a rich texture of data, a comparative context, an understanding of complexity revealed with an economy of means" (p. 51).

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access