Visibility Data Repository
Finally, data collected in the visibility data repository is integrated into existing supply chain applications, data marts or a corporate data warehouse. The visibility repository provides data to execution and planning systems, thus improving performance in supply chain execution and operational planning. It can also accumulate supply chain execution results and historical data for longer-term tactical and strategic decisions. With all supply chain stakeholders working from a common data source, decision making and planning can realize significant improvements.
The visibility hub centralizes reliable information and business processes, thereby empowering a company to make informed operational, tactical and strategic decisions.
Where to Begin?
Clearly, this white paper describes a large, complex implementation. Knowing where to begin is often a significant challenge. In general, a crawl, walk, run strategy is the best method to implement a supply chain visibility project.
Start with a business process assessment and data strategy to determine the organization's preparedness to implement such a complex project. During this phase, it is important to develop high-level metrics to understand how well a process performs and to determine opportunities and objectives. This data can be used to develop a graph of processes that show returns-versus-risk, which can help establish implementation priorities. Low-hanging fruit can be harvested quickly and easily, while other opportunities may require careful planning and execution. Assessments should be scoped as eight- to 12 week projects, with full-time resources committed.
Implementation phases should be scoped in 12- to 16-week increments, although some complex processes may take longer. Large projects should be broken into small, incremental and easily managed components. This not only allows a more focused approach, it also enables an organization to realize incremental returns quicker.
Companies should develop key metrics for such a project, including critical success factors and key performance indicators for trading partners and internal processes. Doing so enables an organization to monitor its visibility implementation continuously, address problems and identify areas for improvement.
Sidebar: Defining the Value of Data Quality
Over the years, numerous systems have been implemented to improve supply chain performance. However, because of poor data quality, potential ROI has not been realized.
A focus on data quality is core to any visibility hub project. With superior data, the results described in this paper can be achieved through a continuous improvement campaign that incorporates all departments, and includes trading partners.
Data visibility becomes valuable when integrated with existing supply chain systems and other legacy applications. Study after study has shown that inferior data quality leads businesses to make poor decisions, hinders returns on systems investments and thwarts user adoption. A study (by the Data Warehouse Institute) determined that poor data quality cost businesses over $600 million annually.
Can Data Quality Be Defined and Measured?
It is useful to define data quality in terms of three key measures: completeness, accuracy and timeliness. Data is complete if all required elements are present to give value to a business transaction. Accuracy measures quality of content. In other words, are the data elements present in the transaction correct? Timeliness can be calculated by whether a transaction arrived in time for it to be actionable. Trading partner data becomes valuable in a typical visibility application when data meets these criteria more than 90 percent of the time.
Where Does Data Go Bad?
There are numerous ways data can be corrupted or misinterpreted. Data that comes from external organizations and various other sources will often turn up in a variety of formats.
For example, data can arrive from many external organizations, such as outsource manufacturers, logistics service providers and transportation companies, customers, and others. Much of this data may be created manually, or feature semantic differences that can cause confusion.
Transaction sequence errors can occur as well. For example, a shipment update is received before an advance shipment notice is processed. Change control in trading partner systems also impacts data, as does use of improper code tables.
In addition, standard operating procedures may not be communicated or understood well by trading partners, which can also lead to incomplete or wrong data. In short, supply chain data emanating from another organization is always suspect.