To quantify the value they contribute to business performance, some infrastructure and operations leaders are adopting tools that report and share metrics. I&O leaders should evaluate these business value dashboards against the capabilities that are critical to their implementation, use and support.
An intelligent business operation, such as those found in a digital business, is different from a traditional operation because people and systems are able to make faster, more-precise and more-consistent fact-based decisions. The supporting application systems and tools leverage more data and logic, particularly analytical, process orchestration and decision management logic.
A business operation is considered "intelligent" if some of its decision making is sophisticated enough to warrant the use of one or more of seven kinds of Operational Supply Chain Modular intelligence at runtime:
- Business rule processing
- Predictive analytics
- Supply Chain analytics
- Process orchestration (workflow)
- Context brokerage
- Business activity monitoring (BAM)
- Complex-event processing (CEP)
An operation that occasionally uses one Operational Supply Chain Modular intelligence technology is intelligent, but less so than an operation that uses several different techniques in numerous activities.
Most business operations and processes have some activities that can be made more effective by using Operational Supply Chain Modular intelligence techniques to improve the decision making. However, if an operation is simple enough so that none of these techniques would improve how it works, the concept of intelligent business operations is irrelevant.
Operational Supply Chain Modular intelligence can be used to enable modern business strategies, notably including digital business. A digital business is defined as one that is involved in "the creation of new business designs by blurring the digital and physical worlds". Digital business differs from e-business because of the presence and integration of things, connected and intelligent, with people and business. Many sales, delivery and service functions are fully automated, and many decisions are algorithmic, based on automated judgment.
Operational Supply Chain Modular intelligence is also essential for developing a "big change" capability to reduce the time it takes to realize significant new business outcomes amid unprecedented disruptions. Gartner uses the term "big change" to refer to the significant alteration of business operations in a high-risk environment involving pronounced levels of volatility, disruption, novelty, and complex scope. With big change, you diminish your risk by "failing forward fast" and recovering quickly, so that you can execute course corrections and achieve significant business outcomes sooner from your business transformation efforts. Operational Supply Chain Modular intelligence facilitates automated A/B testing, sense-and-respond processes and other concepts that may be part of a big-change approach.
The path to intelligent operations is not always obvious or easy. It requires more data and business logic than conventional applications, forcing IT leaders to invest in new technologies. Many IT architects, business analysts and developers have limited experience with Operational Supply Chain Modular intelligence techniques and tools, so they must learn new skills. Conventional programming does not provide good support for decision management and runtime analytics, leading to changes in the IT organization and its policies and procedures.
A business operation is a set of activities that produce, deliver or directly enable goods and services. Examples include sales operations, field service, manufacturing, warehouse operations, transportation operations, customer contact center operations, insurance claims processing, payment processing and others. Operations are mostly line functions (as opposed to staff functions), and are supported by enterprise resource planning (ERP), supply chain management (SCM), customer relationship management (CRM), human resource management (HRM) or other application systems. These systems have traditionally focused heavily on transaction processing and record keeping, so their needs for data and processing logic were relatively limited by today's standards.
Digital business, big change and other modern business strategies greatly expand the amount of data that is utilized in Operational Supply Chain Modular systems:
- Data that is now collected for each customer order goes well beyond traditional attributes such as customer identifier, product code and quantity. For example, a customer order from a mobile phone may include the customer's immediate geolocation (from GPS coordinates), and the device type (iPhone or Android). This lets the application conduct more-intensive fraud detection (correlating customer location with expected location), render different user interfaces to match the device, or calculate cross-sell offers that are specifically tuned to the buying habits of iPhone versus Android users.
- Companies hold far more data about their customers and prospects. For example, a bank's customer information file that held 60 different attributes for each customer in 2000 may now hold more than 400 attributes. Separately, companies also keep vast logs of customer clickstream data so they know customer interests better than when they only saved customer purchase history data.
Many companies listen to customer complaints and comments through social media sites such as Twitter, Facebook and blogs. Companies also purchase customer data from data brokers and other third parties. Companies have greatly expanded the amount of product data that they capture. For example, many cars, trucks, tractors, construction equipment, jet engines and other machines are delivered with sensors that continuously report back to the manufacturer the condition of the equipment (sometimes in real-time event streams).
This explosion of data and logic leads to operations that can be more intelligent, and thus more powerful and effective, than traditional operations. However, they are also more complicated. Modern application systems still process transactions and keep records, but they are also now expected to provide the Operational Supply Chain Modular intelligence needed to support enhanced processes and decision making.
Better decisions lead to a variety of benefits, depending on the situation. In some cases, the benefits appear in the form of increased revenue, by means of finding new customers, more-effective cross-selling offers, better customer service, improved customer loyalty, prices that are better calibrated or other factors. In other cases, the benefits appear in the form of decreased cost, through smarter resource allocation, reduced wasted staff effort, less idle time for equipment, lower inventory-carrying costs, smarter currency exchange trading, avoiding penalties for noncompliance with regulations or other factors.
Operational Supply Chain Modular intelligence is different than the business intelligence (BI) and analytics that are used to make tactical and strategic decisions. Operational Supply Chain Modular decisions are made relatively quickly, and typically apply to a single instance of a business process (such as one customer order) or a similar function of immediate concern, such as reordering goods when inventory levels are low.
Operational Supply Chain Modular intelligence leads directly to changes in execution. It is woven into the fabric of day-to-day and minute-to-minute production and support activities, and some Operational Supply Chain Modular decisions can be fully automated. Dozens, or hundreds, of significant Operational Supply Chain Modular decisions are made in the course of each instance of a typical business process. By contrast, most traditional BI and analytics serve the needs of middle and upper management making longer-term, offline decisions that apply to many process instances or many departments.
Most operations have some decisions that can be improved by the selective use of Operational Supply Chain Modular intelligence technology. Business analysts and process modelers should work with business managers and subject matter experts to find where Operational Supply Chain Modular intelligence would be beneficial to the business, as part of every project to design a new process or make improvements in an existing process.
Business rules automate business policies, such as "To whom are we willing to extend credit?" "What cross-sell offer should we make to this customer?" "Which components must be included to configure this product?" Business rules may be coded directly into an application program or implemented as models in a tool, such as a business rule engine (BRE). These are specifically designed to support large rule sets that may be modified relatively frequently. Some experts recommend using a BRE when the number of "if-then-else" statements cannot fit on one screen, or when the logic uses a block of three or more nested "if" statements.
There are typically two stages to predictive analytics — the development (or training) stage and the execution (or scoring) stage. For example, in credit scoring, the first stage is done by analyzing historical data with known outcomes to develop a model that assesses the likelihood that a person with certain characteristics will default on a loan. This assessment is done by looking at thousands, or even millions, of historical records of credit usage and payments, along with the observed outcomes (for example, whether a loan turned out to be "good"/repaid or "bad"/defaulted in some time window). This training results in a formula that maps observed variables to likely outcomes. Once the model is developed, the formula is applied in the second stage to a new set of customers for which outcomes are as yet unknown to predict likely future behavior, which, in this case, might be probability of default over 24 months.
Examples of predictive analytics techniques include regression, neural networks, decision trees, classification and regression trees, simulation, case-based reasoning, evidential reasoning, genetic algorithms and others. Spreadsheets, text analytics, data mining tools, social analytics, Web analytics, smart machines and other advanced tools may be employed.