The following criteria should be considered while evaluating vendors and technologies for meeting data quality requirements in an organization:
1. Enterprise Data Quality Framework – Evaluate the degree of effort required to integrate the wide range of functional capabilities into a single architecture and product, so that end users will ideally have a single point of access and integration point for the Data Quality domain. This requirement may be a long term goal as it is almost impossible to achieve enterprise wide data quality platform from a single initiative. A multi-year program consisting of small subject-oriented projects is a more feasible approach. However, the vendor plays an important role throughout the implementation of the program, and therefore should be evaluated against this criteria. Of course, it is critical to analyze if a true enterprise wide implementation is required, as a localized departmental one may suffice; it really depends on the organization’s mandates.
2. Functionality Coverage – Evaluate the extent of functionalities an capabilities of the data quality platform/solution. The usual capabilities that are required from a robust data quality platform include data profiling, reference data standardization, value matching, data rule analysis and rules discovery, data enrichment using supplemental data feeds and data quality monitoring. Depending on the specific requirements, evaluate a platform based on the depth of coverage for these and other capabilities. Of course, a solution that offers all of these may not be a best fit for the organization’s requirements, therefore a one size fits all approach may not be ideal either.
3. Adaptability and Usage by non-IT Resources – Recent economic and organizational evolution has increased the focus on the business, non-IT resources to own and to some extent manage domains such as data quality and master data management. To meet this requirement, it becomes necessary for the system and the tools to be more business user friendly. They should be able to be managed easily and effectively with minimum confusing clutter in the presentation layer, and a seamless workflow that is transparent to the end users. The extent and degree of these requirements should be properly evaluated.
4. Operational Characteristics of the Data quality Platform – Data Quality solutions can be deployed in multiple ways depending on an organization’s budgetary and organizational requirement. The three main options are on-premises software deployment, hosted solutions and software as a service [SaaS]. The technical capability of a product or solution to be suitable for the appropriate deployment option, the capability of a vendor to support and provide partnership for each of these options as well as the technology requirement to manage the data and code that will reside in each layer should be analyzed.
5. Centralized Data Quality Management Governance Framework – It is always a good idea to think about managing the data quality aspects similar to the fashion in which a master data or a metadata repository will be owned and operated by a dedicated team. To that effect, establishing the roots of a centralized data governance framework will reap benefits when the data quality initiative is operational. The capability of a vendor to support seeding such an initiative, least of all by sharing best practices, proposed organizational models and workflows will go a long way in helping out setting up the direction for data quality success. Rolled into this may be the task of increasing awareness across the business and technology community of the importance of data quality and therefore the concept of design for quality so that quality is built into future systems and not treated as an after the fact exercise.
6. Non-technology characteristics, such as cost, maintenance/support, upgrades, pricing models, viability and lateral partnerships of the vendors are critical factors since data quality is a program rather than a technology and the relationship with the vendor is a long term one.
