| about every field of work, there are quality | | | | requirements for quality of data. Once data |
| measures in place to ensure customer satisfaction | | | | management progresses toward an |
| and product/service effectiveness. Manufacturing | | | | enterprise-wide set of standards, there is often |
| companies rely on quality control processes to | | | | push back or hesitation by the business unit |
| minimize defects and reworks. Consultants | | | | managers to invest time and resources in |
| measure the quality of their services to ensure | | | | addressing issues that were not relevant at the |
| repeat business. Journalist rely on quality | | | | business unit level. |
| information and leads to maintain integrity and | | | | Choosing the Right Data Management Tools |
| credibility. But when it comes to corporate data, | | | | A frequent response by organizations with |
| many organizations fail to understand the | | | | respect to building a data quality management |
| significance and drawbacks of unreliable or | | | | program is to immediately begin to research the |
| inconsistent data. This article discusses five quality | | | | purchase of automated data cleansing or profiling |
| challenges many organizations face and ways | | | | tools. While some data quality tools do provide |
| they can be more proactive in managing data. | | | | some benefit right out of the box, without a |
| Among the primary reasons for inconsistent or | | | | well-defined understanding of the types and scope |
| unusable data are: | | | | of specific quality problems, and without a |
| * bad data from human data-entry error | | | | management plan for addressing discovered |
| * poorly-structured process | | | | problems, buying a tool will not have a significant |
| * lack of data standards across functional units or | | | | return on investment in achieving long-term |
| divisions | | | | strategic goals. |
| Ensuring the quality of data can become | | | | Placing Responsibility for the Quality of Data on |
| extremely difficult when you attempt to integrate | | | | the IT Department |
| data from across multiple sources. Before your | | | | Business units often assume that any issues |
| organization begins a data-driven initiative it is | | | | regarding the quality of data are IT issues, and |
| important that you address issues of data quality | | | | should be addressed by the technical teams. |
| within your existing data sources. Aside from the | | | | However, the business rules associated with |
| complexity of the actual process of ensuring the | | | | running the business is best managed by the |
| quality of your data, below are five challenges you | | | | business client? |
| may face when beginning this initiative: | | | | Reactive vs. Proactive Mentality |
| * Data ownership | | | | Most data quality programs are designed to react |
| * Non standard data requirements | | | | to data quality events instead of determining how |
| * Choosing the Right Data Management Tools | | | | to prevent problems from occurring in the first |
| * Placing Responsibility for the Quality of Data on | | | | place. A mature data quality program determines |
| the IT Department | | | | where the risks are, what the objective metrics |
| * Reactive vs. Proactive Mentality | | | | are for determining levels and impact of data |
| Data Ownership | | | | quality compliance, and approaches to ensure high |
| Data ownership, especially on the enterprise level, | | | | levels of quality. |
| is a very complicated transition, and can | | | | Ways your organization can be more proactive |
| contribute to significant pushback within an | | | | towards data quality management: |
| organization. Often the business unit managers or | | | | * Ask for data quality performance measures as |
| technicians entrusted with the implementation of | | | | part of your business requirements gathering and |
| an application assume ownership of the | | | | prioritizing process. |
| information used within that system. This | | | | * Determine, along with the business, how you |
| introduces potential conflicts when these individuals | | | | are going to handle data quality issues both during |
| must participate in enterprise-wide data initiatives | | | | the development process and when your |
| and expose the internals of their information | | | | processes are operational. |
| management to data quality audits and reviews. | | | | * Monitor data quality at every stage where data |
| Non Standard Data Requirements | | | | is touched |
| Traditionally data management is structured | | | | * Create a data quality management dashboard |
| where the business unit's management chain has | | | | to monitor the agreed upon data quality |
| authority over the information used within the | | | | performance measures. |
| business unit, and each business unit has its own | | | | |