This chapter excerpt of Smart (Enough) Systems will reveal common characteristics of smart systems including operational
and agile functionality and real-time capabilities. You'll also learn definitions for agile compliance, loose coupling and human latency and find out about Web 2.0 features that companies can leverage.
Table of contents:
- The importance and benefits of operational decision making
- How to make operational decisions and data corporate assets
- Benefits of operational, real-time capabilities in smart systems
- How to create automated operational decisions
Introducing Smart Enough Systems
What kind of systems would deliver this vision of operational decisions? The term "smart enough systems" is used in this book to describe them. A smart enough system is not some kind of artificial intelligence device like HAL 9000 (from 2001: A Space Odyssey). Equally, a smart enough system can't be developed the same way you build traditional "dumb" information systems.
Smart (Enough) Systems
by James Taylor and Neil Raden ISBN 0132347962
First printing June 2007
Prentice Hall Professional
Building smart enough systems means taking a new approach to bringing automation to operational decisions. Instead of hard-coding decision rules into systems, it means using separate tools to build, manage, and carry out decisions in concert with other operating processes. It means developing new services that can deliver operational decisions that perform well enough to be used in real-time front-line systems and processes. It means developing services that are agile enough to keep up with a changing world and, indeed, learn from it. It means services that make customer-centered (associate-centered) decisions and services that can support an extended enterprise.
Characteristics of Smart Enough Systems
Smart enough systems have some key characteristics. In particular, they are operational and capable of real-time performance. They are agile, capable of learning, and customer- (associate) centered as well as compliant and supportive of an increasingly extended enterprise.
The increasingly distributed and always-on nature of organizations puts a premium on high-performance execution, which means operating quickly, flawlessly, legally, and profitably at every level. It's about more than how your employees perform; it's about how your systems perform. For organizations to exhibit high-performance execution in day-to-day operations, their top performers' expert judgment must be made available everywhere. This means making sure the systems everyone uses embody that expertise—not in an expert system, ask-if-you-get-stuck way, but with systems that are embedded in the operational processes necessary to the business. This expertise, however, must be balanced by analytic insight developed by a careful analysis of the organization and its operational history. As Malcolm Gladwell said:
"Truly successful decision making relies on a balance between deliberate and instinctive thinking."
Not only must the organization's expertise be balanced with an effort to run the organization "by the numbers," but the interaction skills of those who serve as the point of contact with associates also must be considered. For the moment, no system can replace human interaction. Ensuring that these interactions make use of decisions informed with an analysis of past success and experts' judgment can ensure that customers get the best possible experience and organizations can get the best possible results. Smart enough systems make organizational knowledge "explicit, executable, actionable, and adaptable.
Capable of Real-Time Performance
Smart enough systems must operate in a no-wait, multichannel world where customers and other associates expect responses, actions, and decisions immediately. Suppliers expect immediate updates on the demand chain, and retailers and distributors expect to know about problems in the supply chain instantly. Real-time connections between organizations are also essential, because organizations must become more loosely coupled. They must deliver their products and services by coordinating and orchestrating many distinct organizations, both internal and external.
In the past it was sufficient to coordinate operations within an enterprise, but today successful organizations must be able to operate with both known and unknown entities without delays. Systems can't wait for someone to wake up before acting, and people want to be told what has been done to make their life easier, not asked for decisions. Smart enough systems must make decisions fast enough to be used in operational, real-time systems.
In the heavily regulated environment in which many organizations must operate, agility can't come at the expense of compliance. Every time an organization shifts its strategy and changes its operations, it needs to be sure that the new approach is compliant and can be demonstrated to be so. Compliance, then, can act as a drag on an agile organization by preventing it from making changes as quickly as needed unless the approach it uses to achieve agility allows it to remain compliant—which can be called "agile compliance."
An agile organization can effectively change the way it operates when it needs to, but only if it has a good understanding of how it's operating and why it operates that way. Smart enough systems support this agility by making how they operate explicit, easy to understand, and easy to modify. Agility is a measurement of the total time and cost in getting from having the data that means you should change your business to actually making the change.
Gartner Group Inc. defines agility as "the ability of an organization to sense environmental change and to respond efficiently and effectively to that change." Gartner uses an agility cycle, shown in Figure 1.3, to show how agility is achieved and to indicate that it's ongoing. The basic steps are sensing a threat or opportunity, strategizing about options, deciding on the most appropriate action, and then communicating it before acting. This cycle must be continuous, because each change must be monitored for subsequent changes.
People will always be a bottleneck when it comes to change. Gartner defines "human latency" as something that reduces agility, for instance. Some technology approaches make it harder to be agile, and some make it easier. Some help you persuade people to change; others help you make the changes after they have been agreed on. Organizations in the future will have to improve their agility organizationally and in terms of the technology they use and how they use it.
Capable of Learning
Smart enough systems need to "learn" as new data is collected. Organizations collect an enormous quantity of data, and the volume of data is increasing steadily. Generally, organizations don't have systems that are smart enough to take advantage of this data. For instance, a PricewaterhouseCoopers Barometer survey12 in late 2006 gave an accurate summary of how organizations think their data should give them a real competitive edge and why it currently doesn't:
- 71 percent of senior executives describe the data in their company's information systems as potentially very valuable.
- 68 percent of these executives expect this data will become even more valuable as a source of competitive advantage during the next 12 to 18 months.
- 84 percent cited their inability to mine and interpret data as the highest-ranked obstacle to achieving value.
- 75 percent said that an ability to mine and interpret data was key to getting value from data.
You have to do more than collect, organize, and report on your data. You, and your systems, have to learn from it, mine it for insights, and interpret what it means for the future. Doing so is a key to taking advantage of one of your last remaining areas of competitive advantage—knowledge of your associates—and is a way to improve performance by insisting on realism. If your business decisions are based on what your data tells you, you're more likely to get realism than if you rely on hunches and how you have always done something. Even Malcolm Gladwell, in his paean to instinctive reactions, noted that informed snap judgments outperform uninformed ones.
Competing on Analytics
In their book, Davenport and Harris describe competing on analytics as "the extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions and actions." They define an analytical competitor as an organization that uses analytics extensively and systematically to outthink and outperform its competition. They see a growing cadre of companies competing this way and identify a number of trends as a result:
- More automated decisions
- More real-time decisions
- Increased use of alerts
- More prediction and less reporting
- More mining of text
Companies wanting to compete with analytics will require far more than a few "PhDs with personality" using data-mining tools on existing enterprise data. They will need a systematic framework for using analytics to make their systems smarter.
You must also realize that a generation of workers, the baby boomers, is retiring. The new generation of workers is more technology-literate but is unlikely to take the kind of jobs their parents and grandparents took. Even if they did, they lack the depth of experience on which organizations have been relying. Those retiring baby boomers know all the tricks, exceptions, and workarounds that make your manual decisions work. Without them, you need some other way to get this knowledge to your workers, and these workers will look to information systems for that knowledge.
Customer- (Associate) Centered
"Customers can access more information about more vendors, negotiate more effectively with still more vendors, and switch from one vendor to another whenever they find greater value."
Much money and energy have been spent using technology to improve customer relationships, yet much of it has been used as a technological alternative to talking with customers, not to empower customers.
Many organizations fail to respond to customers in a consistent, focused, targeted way and have customer processes that are costly in terms of customer satisfaction, operating costs, and profits. As the world moves faster and gets more complex in terms of regulations and competition, this situation will get worse. Customers expect quicker decisions and are no longer willing to wait for them. With all the information about competitors and quick Web-based access to them, they can find an alternative easily.
With the many channels now available, the potential for annoying or ignoring customers unintentionally is rising. Competitors are constantly forcing reactions, because customers might find another supplier who offers them something more compelling. These customers want to self-serve, to actively manage their relationship with their suppliers, and the organization of the future must make it possible for them to do so. As interactive web applications get better, many people will prefer "self-service" over "customer service."
The information an organization has about its associates is widely regarded as one of the few advantages an "incumbent" has. The current frenzy for customer data integration (CDI) is clear evidence that more attention is being paid to managing the resource of customer data. However, it doesn't matter how well managed and integrated this information is unless it contains customer preferences, and unless their preferences and your insights are used to tailor interactions with them.
The information you have about associates is a critical advantage only if you can learn from it. This learning can't be static, either; you can't discover an interesting piece of information about your associates and then stop. This insight must make it into your systems. Smart enough systems focus on better decisions for how to treat associates.
Support for an Extended Enterprise
The growth of outsourcing and smartsourcing15 is leading to more loosely coupled organizations or groups of organizations. As Hagel and Seely-Brown said, "Loose coupling represents a more modular approach to process management." Loose coupling means creating independent activities with clear owners and interfaces and performance guidelines. These activities can then be assembled and disassembled more easily to meet changing needs. This kind of business structure parallels the more flexible approach to information systems represented by a service-oriented architecture (SOA). This approach implies trusting relationships.
Organizations adopting this approach need systems smart enough to work in this environment and smart enough to allow associates to change how decisions are made in the processes that span organizations—that is, processes that require multiple organizations to deliver. Business processes, which once belonged to a single organization, are now composed of agile mini-processes that must be configured dynamically across organizational boundaries. This is impossible without the handshake of industry standards, directory services, and orchestration—and, once again, loose coupling in a service-oriented architecture. The systems supporting these processes must also be smart enough to generate the kind of audit trails and decision outcome logs that build trust between companies and between companies and their regulators.
There was a time when trendy expressions were durable. "Groovy" lasted about five years before it was no longer "cool" to say it. In business, "impact" as a verb stuck around for a decade or more. During the Web Bubble, "disintermediation" was cool for a year or so before, as with other trendy words, using it indicated you were a little behind the curve. The problem is that technology moves so fast now that these terms fall out of favor long before they have a chance to prove themselves. This trend is already happening with "Web 2.0."
Web 2.0 has real merit and staying power, however. It might no longer be avant-garde because of overuse and overexposure, and by favoring it we may find ourselves a bit derriere. But the term is an intermediate point between the original "World Wide Web," a collection of pages and a protocol for using them, and "Web 3.0," the truly semantic Web, where the entire collection can be mined for meaning. Web 2.0 offers some fascinating features and capabilities that enable people, organizations, governments, and even machines to interact based on some simple principles:
- The Web as a platform— The Web itself becomes the place where computing happens, which is part of the growing interconnectedness or "flattening" of the world. Services on the Internet can be assembled and disassembled at will.
- Collective intelligence and wisdom of crowds— Some evidence exists that the collective behavior of large numbers of people is a better predictor than expert judgment.
- Using data, not just collecting it— Data is gathered from both internal and external sources with the intent to use it to act differently.
- The end of release cycles— Continuous, unnoticed software change without conflicting and overlapping release cycles replaces point releases and disruptive, crippling maintenance efforts that drain IT budgets.
- Designs for "mashing" and "hacking" applications— These application designs provide a rich user experience in the corporate IT environment by "mashing" multiple applications together or by "hacking" an application to alter its behavior. This is all made possible by adherence to standards precipitated by the dynamic nature of the Web.
Operational decisions being reflected in services, the use of predictive analytics to apply the implications of group behavior to transactions, the focus on getting insight from data rather than just collecting data, and the ability to refine decisions continually without affecting other systems are characteristics of smart enough systems and Web 2.0.
Service-oriented architecture (SOA) is one of those phrases that gets thrown around in everything from technical standards to business books. Thomas Erl makes four key points in his books:
- SOA can establish an abstraction of business logic and technology that allows a looser coupling between an organization's processes and its technology.
- SOA is an evolution of past approaches, preserving successful characteristics of traditional architectures and adding distinct new principles that foster service orientation.
- SOA is ideally standardized throughout an enterprise, but achieving this requires a planned transition and still-evolving technology.
- SOA is a technology architecture that supports and promotes service-oriented principles throughout an enterprise.
What SOA does, at a fundamental level, is allow the development of individual pieces of business functionality in a way that lets them be combined and modified effectively and without tightly coupling them to each other.
Organizations must also handle more jobs that aren't located in one building or even one country but are outsourced or "homesourced" by using the Internet and related technologies to connect workers. The systems these workers use must be smart enough to let them do their jobs effectively and to act on behalf of the organization yet ensure compliance with company policy and more.
Thomas Friedman says, "There are currently about 245,000 people in India answering phones from all over the world or dialing out to solicit people for credit card or cell phone bargains or overdue bills." He describes a series of trends and technologies that have, in his words, "flattened" the world by making it more interconnected. He explains how this flattening fits with globalization and how companies are reinventing themselves in the face of these changes and describes some of the problems, risks, and effects on political and public policy.
For example, deciding where to locate work is becoming more complex. More options, with advantages and disadvantages, are available, thanks to the overall increase in interconnectedness. Friedman explains that work will go where it can be done most effectively.
Another concept emphasized in the book is that of global, dynamic supply chains that "[coordinate] disruption-prone supply with hard-to-predict demand." For most of history, location has been critical for businesses of all kinds: where to open a store, where to put a factory, where to find customers. Improvements in connectivity and network bandwidth, however, mean that location is no longer a factor. Now the trend is work taking place where it can be done best and for the lowest cost. In addition, organizations find customers as well as suppliers and staff all over the world. They can reach out to new markets, take advantage of new opportunities, and collaborate with new partners worldwide. The parallel growth in information content of products and the overall shift from products to services in the world economy have forced organizations to consider their "digital supply chain." You can no longer consider just how and when physical goods are moved through your supply chain; you must also manage the knowledge and information that flow through it.
This ability to build a more distributed, electronically connected organization has consequences, however. In particular, how do you control it? When you outsource work to India or home-source it to Peoria, how do you make sure the work is done the way you want it done, following your policies? You need to be able to ensure that people working all over the world for you and your partners or suppliers treat your customers, your products, and your employees the way you want them to. You must equip them to act as though you were sitting in the next cubicle, even though they are geographically dispersed and perhaps brought together only temporarily to meet a business need.
Will you rely on just policy manuals and training? Will you assume that the people making decisions on your behalf can interpret data correctly from their reports and apply your business strategy to what the data tells them? With home-sourced booking agents, for example, you want to make sure they offer your best travelers upgrades when they can and know how to prioritize customers who need rerouting. Those 245,000 phone operators in India need an automated system for approving credit and recommending what kind of collections strategy will work. They need smart enough systems.
Each new scandal seems to result in a new piece of regulation. Government and nonprofit organizations struggle under their own burden of reporting and compliance, and the penalties for noncompliance grow for organizations and individuals. For these reasons, governance and compliance are popular topics on the conference circuit.
Not only do organizations face more restrictions, but also many restrictions now demand demonstrating compliance. Organizations must be able to show that they are compliant with regulations. No one has to sue them or demand the information; they must report it annually, quarterly, or more often. In this environment, allowing front-line workers to make critical decisions is risky. They are less likely to be well trained, more likely to have high turnover, and most likely to be employed by third parties in the form of outsourcing. They don't necessarily make the best decisions. More important, showing that they made legal, appropriate, compliant decisions isn't easy. If more of your decisions are embedded in your information systems, however, you risk pushing the enforcement of these rules onto programmers who don't understand them, not onto businesspeople who do.
Additionally, more organizations must contend with multiple layers of regulation. They are obliged to follow local and national regulations, as they always have, and doing business on the Internet or using outsourcers around the world increasingly involves new sets of national regulations. Many international organizations, from the European Union to the World Trade Organization, also have rules that must be followed. Even knowing which set of local, national, and international rules must be applied to a specific transaction becomes a problem, let alone actually enforcing and demonstrating compliance with those rules.
Formal regulations are not the only rules an organization might need to follow. Socially conscious consumers, activist shareholders, and nongovernmental organizations also play a role. An organization might need to enforce rules to show that it's "green" or to defuse an unpopular perception of it. These "rules" must be enforced just like regulations, but they will be truly valuable only if made public. Those who care about these rules want to know exactly what the organization is planning to enforce. They want accountability—knowing what you did with their money, goods, and so forth. Managed transparency becomes important for most, if not all, organizations. Demonstrated compliance with publicly auditable rules creates new demands on systems and people.
Regulations are also becoming more sophisticated. No longer are they simply a set of rules to be enforced; some are starting to embody best practices and statistical measures. Two examples are Basel II, with its enforcement of best practices in risk management, and court rulings forbidding personnel actions that might reasonably result in discrimination against a class of employee, even when no actions specifically do so. Being compliant won't get any easier.
The push toward compliance has a cost, however. As Taylor notes, "The biggest problem with SOX . . . and [other regulations] is that it assumes a relatively static mode of business operations, and today, to be static is to be dead."
Organizations must deliver agile compliance; they must maintain business agility despite the burden of increased compliance. The increase in regulation tends to slow the rate of change in organizations by making it more expensive to make changes, but it can't stop change. Some organizations will find a way to evolve and be agile despite the regulations they operate under, and their competitors will need to do likewise. Achieving agility despite regulatory burdens requires smart enough systems.
Dig deeper on Business intelligence best practices