 |
Diffused Complexity
| |
The second article in a Think Research series that uncovers the causes of needless complexity and provides solutions.
Pumping your own gas is one thing. But what if a customer ordered a car and received a box of parts and a wrench in return for his or her money? Not likely, of course, but the truth is that it's not far from what some organizations do to their employees and customers under the banner of "saving money" or "user empowerment." IBM Research refers to this as diffused complexity, and it is an effective way for a company to pick its own pockets.
Sometimes complexity brings value
Self-service can be a powerful tool - one that yields lower costs and happier customers or employees. If a task is simple enough for an untrained person to do it in less time than it would take him or her to turn it over to an expert or an assistant, then it is an excellent candidate for self-service. Supermarkets provide the prime example of this principle, letting customers pick their own goods and, in some cases, even allowing customers to be their own check-out clerks.
Another example can be found in the long-forgotten steno pool. It used to be that very few businesspeople wrote their own memos or papers; instead, they would dictate and a trained professional would transcribe. The executive would then read over the transcription and handwrite corrections; and after a number of cycles, the finished product would be mailed. The advent of the personal computer and word-processing software made this process a perfect candidate for self-service.
Sometimes self-service destroys value
But sometimes self-service can be a bad idea, especially if it ignores the cost differentials among roles in a company. If converting a task to self-service makes it require more of an executive's time than it did before, or if it blocks someone from proceeding until the task is complete, self-service may be a mistake.
One of the common motivations for making a task self-service is to reduce costs - after all, the organization no longer has to pay for the experts or assistants who used to perform the task. But the costs of the process don't go away - they just get diffused out to all of the people who are now forced to do the job, and those people may well be forced to spend time on trivial tasks instead of doing their real jobs. Or they may be unhappy at being forced to fend for themselves, which can be problematic if a company pushes a task onto customers without giving them a clear benefit in exchange.
Moving away from the steno pool model for creating documents, for instance, is one example of how self-service can sometimes work against the best interests of the company. While most businesspeople have learned to type and can create their own letters and reports, they often do not know how to effectively use the graphic capabilities of modern office suites to convey their key messages - instead, they create pages of bulleted text, small fonts, ill-chosen colors and busy pictures, all of which detract from the message they are trying to deliver. Because a document stays on the executive's desk until he or she is finished with it, there's the danger of over-improvement - polishing a document far beyond its value. And there's also the danger of bad quality - misspellings and grammatical mistakes may not be caught before delivery because there's only one set of eyes looking at the document (automated spelling and grammar checks help, but software is not perfect).
Amateur Experts
But, of course, employees are asked to perform duties that are much more complicated than creating their own documents and presentations - and much more fraught with potential problems and additional costs. For example, many companies are asking their employees to use online travel reservation systems instead of calling an agency or using an internal travel department. The stated goals of these systems include enforcing travel policy, encouraging cheaper flights and eliminating travel agency fees, but shifting this work to high-priced employees doesn't lead to cost-savings. It leads to diffused complexity.
This process decision seems like a simple way to save money, but making travel reservations can be a fairly complex task. There are, of course, some situations where everything is clear and the most convenient itinerary is also the cheapest - and online systems are ideal for such trips. However, many trips fall far below this ideal, and the traveler is required to juggle factors like these:
- Knowing when the system's suggested itinerary is unacceptable (whether because the price is too high or the connections are too uncertain), and how to override it to get a feasible itinerary (such as when the times fail to meet business requirements)
- Figuring how to perform a ticket exchange for a cancelled or rebooked trip
- Choosing between alternative airports or routings
In one case that IBM Research's services consultants studied, the system offered only one choice for a trip between Phoenix and London - a $9,000 fare that required two plane changes, even though there was a $4,000 fare available with a shorter travel time. This trip required over an hour's work, including calls to the travel agency's help desk, to figure out how to book the cheaper fare.
The problem is that these systems are often deployed with very limited employee training - and that training is usually on how to navigate the system's menus, rather than on how to input the necessary information so the system can provide good options. And employees, who used to be able to rely on the expertise of the travel agent (both in eliciting the right information from the employee and in knowing what was reasonable and what wasn't), now have to spend time trying to coax the system into giving them acceptable itineraries and pricing.
Often, because employees do not access fares and other details on a daily basis, they may not understand or trust the system options. One common workaround IBM's services consultants have seen used by employees is to access a commercial travel site, such as Expedia or Travelocity, to investigate alternatives and determine if the corporate system is providing the best options. And, if an employee only travels a few times a year, then he or she has to relearn the system each time. Obviously, these are extra steps that add time and cost to performing the task.
IBM Research's experience indicates that companies that convert complex tasks like travel arrangements to self-service wind up with a large number of annoyed and unhappy employees - and this continues even after they have had time to learn the system. This is often due to the fact that the company only provides training aimed at the mechanics of the system, neglecting the additional information and knowledge people need to know to effectively perform the task. In the travel reservation example, this information could be alternate airport and hotel options, optimal routings and connection times through particular airports, and airline policies about changes and refunds. Companies should listen to the stories their employees are telling about such self-service systems.
Solutions
The fundamental law of diffused complexity is this: spreading out costs doesn't make them go away. The second law is also simple: untrained people are expensive.
These laws don't mean that distributing a process (and its costs) is always wrong; they do mean that companies have to be conscious of the burden they're putting on their employees and must communicate the value gained with them.
In light of these laws: In the supermarket example, the customers gain time because they don't have to wait for the clerk to help them pick out items, and they have the option of self-checkout or the traditional system. In contrast, the users in the online travel system example have to spend extra time and the trade-offs may make it possible that the organization has actually lost value by implementing the system.
IBM Research's advice is simple: When designing a process, a company must consider the costs and value across the entire organization, not just the group that owns it.
Some of the factors to be considered include:
- Which employees would have to perform the process?
- What is the cost of their time (both direct cost and opportunity cost) compared to the cost of performing the procedure centrally or with specialists?
- What training will these employees need? Remember that they'll need to be able to understand the results the process produces, not just mechanically traverse the system menus. As a rule of thumb, employees will need to know the same types of things that a centralized group or specialists know today.
- Evaluate the new process after it has been deployed for three to six months. Listen to employees' stories with an open mind. Seek to understand the good and the bad. And be sure to ask how much time is now being consumed in comparison to what was consumed before. Don't be surprised if the cost equation has tipped in an unintended direction.
- Be prepared to take additional actions to remedy any ill effects from diffused complexity. Companies may need to train people on the underlying content, change policies, change the system, or rethink the entire process. If the company remains open to this kind of follow-up action, they are much more likely to reap true benefits.
If a leader must deal with diffused complexity, yet has no immediate control over it, one action to take is to identify specific people within his or her group who will become experts in that topic. Then, diffused complexity can be funneled to these people, or others can go to them for support when learning the process or if they are running into difficult challenges. To make this type of strategy work, the company will need to invest extra education and training into the designated subject matter experts (SMEs), and it may have to change some work responsibilities around to enable them to perform these tasks, but it can help reduce the overall cost and delays that diffused complexity is bringing to the business.
Contact one of the experts below to discuss the ideas in the article or to see how IBM Research can help your organization:
David Singer
David Singer, an IBM Distinguished Engineer and member of the IBM Academy of Technology, joined the Services Research group at the IBM Almaden Research Center in San Jose, California in 2003. His primary research interests are in the effects of technology on people (both inside and outside of work); currently, he is developing a model of IT's effect on knowledge-worker productivity and pioneering research into combating needless complexity within organizations.
Singer received his bachelor's and master's degrees at Rensselaer Polytechnic Institute and joined IBM in 1976. He was heavily involved in IBM's early work with the Web and was an original member of the Advisory Board of the World Wide Web Consortium. He is the author of the OS/2 Gopher client, ran IBM's first Internet Gopher and Web sites, and built an innovative gateway to allow IBM employees to access Usenet through IBM's mainframe-based computer conferencing system. He can be reached at singer@almaden.ibm.com.
Sara Moulton Reger
Sara Moulton Reger joined the Services Research group at the IBM Almaden Research Center in San Jose, California in 2003 as one of the first subject matter experts brought over from IBM's Business Consulting Services to initiate this new area of study. She has been a management consultant for over 16 years, specializing in organizational change, culture transformation, governance and leadership, both at IBM and at other leading consulting firms.
Moulton Reger, a Certified Management Accountant who received her bachelor's degree at Colorado State University and her MBA at the University of Northern Colorado, held financial leadership positions during the early years of her career. Within the Services Research group, her efforts have resulted in an innovative, patent-pending method for making corporate culture tangible and workable; she is also pioneering research into combating needless complexity within organizations. She can be reached at moulton@us.ibm.com.
