As organisations mature they invariably wish to extend the reach of their data quality and data governance initiatives but how can they do this effectively? How can they ensure that there is no overlap and that everyone is focused on the same goal?
In this interview with expert Mark Allen we explore how various business and IT alignment issues impact ability to successfully execute enterprise data governance and data quality programs. Mark also provides insights for how to address these alignment issues.
Dylan Jones: In your experience, what are some of the common alignment issues most organisations face?
Mark Allen: The big one is failure to align data governance with other governance initiatives in the organization. For example, there may be a corporate IT governance strategy that steams ahead without any consultation with the data governance team. Project governance is another area where alignment with data governance can often be lacking.
Most organizations go through some fairly substantial structural or strategic changes from time to time and this can really slow the momentum of a data governance or data quality management program. Quite often structural changes in the organization can cause a splintering effect so these can be very damaging and all the groundwork has to be started again.
A common alignment issue relates to root-cause analysis. For example, a data warehousing team may spot a defect but there is no alignment between different departments to resolve the actual cause. Random point fixes may be implemented along the information chain but there is no common alignment to fix the problem permanently.
Another alignment common issue is where there are no cross-functional standards regarding use of reference data such as country codes, product codes, service codes, reason codes, language codes, etc.
Dylan Jones: How do you structure Data Quality Management within a Data Governance framework? What does a typical roadmap look like?
Mark Allen: Ultimately this structure should be tied to how data governance domains are defined.
Data governance should establish the domains and what data that domain is responsible for. Data quality management and improvement efforts should be coordinated within and across these domain structures.
An executive steering committee should oversee the enterprise data governance program and these domains. That creates a top down strategy for data governance and quality management roadmaps to be rationalized, prioritized, and appropriately assigned to the program and domains.
Each domain should be responsible for driving their quality initiatives and to participate as needed in cross-domain initiatives.
Dylan Jones: How do you staff a data governance team structure? What has worked well for you in the past?
Mark Allen: Staffing a data governance team sufficiently is often a challenge and can certainly be a recurring challenge in a multi-domain model.
Conceptually most data governance team structures will be aiming at having a multi-tier team structure involving an executive sponsor layer, a decision making layer, a data steward layer, and may also include some IT support roles such as a business analyst, data analyst, or data architect.
However, in practice it may be difficult to fully enlist all these types of team members because these people will usually have other full time job roles and responsibilities that can create commitment problems.
Based on my experiences, the best approach is where a data governance program office can work with the domain area to provide various services and facilitation support to help implement an initial level of data governance structure sufficient to start demonstrating value.
Over time as the value proposition and visibility of this data governance process grows, so will the level participation and support.
Dylan Jones: How have you aligned Data Quality Management with the project governance model?
Mark Allen: The alignment can occur through various relationships. A data governance domain team should have members representing various projects and initiatives within that domain. In this case data quality management needs can be reviewed and coordinated between the data governance and project teams.
Another scenario is where a solution design process, such as an SDLC (Software Development Life Cycle) process, and a project management office (PMO) have process checkpoints or gates in place where data quality requirements and data governance signoffs are required.
Having these type of protocols will maintain alignment between project governance, data governance and quality management.
Dylan Jones: A lot of companies appear to cope well with localised, departmental alignments of data quality and data governance but struggle to make the move to an enterprise initiative. What are some of the tips you can recommend for making that leap?
Mark Allen: Movement to an enterprise model needs to be driven by executable enterprise strategies and initiatives acting as forcing functions to migrate localized functions. Without these enterprise drivers there is little incentive and opportunity for localized departmental functions to move into an enterprise model. After all, localized departmental functions are just that -- local functions that are supporting either a local model or an enterprise model.
During a transition period there may be support for both local and enterprise models until the local model is decommissioned. But generally speaking, localized data quality and governance functions are not going to make the leap to an enterprise level unless opportunity and business strategies are enabling this.
Dylan Jones: What technology and associated infrastructure have you had to implement to manage all the extra processes and functionality associated with running an enterprise data governance and data quality program?
Mark Allen: To effectively execute an enterprise data governance and data quality program it is critical to define enterprise standard solutions and platforms to support the program. This should include enterprise standard methodologies and tools for metadata management, reference data, data profiling and analysis.
This will also likely require elimination of overlapping or competing functions in order to achieve standard, scalable solutions.
In fact, one of the very first tasks for any enterprise program office should be to inventory all existing data governance and data quality functions across the business model. Typically this will reveal overlapping processes, tools, and gatekeepers that are creating process and resource fragmentation.
Once this inventory is completed the program office should work toward defining a more standardized and consolidated approach.
Dylan Jones: How have you measured the impact of aligning Data Quality Management and Data Governance? For example, who do you report to and what kind of measures do you submit?
Mark Allen: This is a great question.
Level of impact will depend on how this alignment is approached. If there are corporate strategies and imperatives to transition from legacy processes and applications to enterprise solutions, then realignment of existing data governance and data quality efforts and resources is likely to be a natural bi-product of that strategy which is probably also driving some organizational change. Without such corporate imperatives, gaining alignment will likely require more tactical approaches such as a domain-by-domain assessment of how the data governance and data quality initiatives can be more effectively executed through better collaboration between the programs and projects. In this case, there may not be any impact to job roles or reporting relationships, just more cooperation between the programs resources.
Many good quality improvement proposals never get executed because of inability to gain agreement or commitment for just one or two extra resources needed in the plan.
An aligned data governance and data quality function is much better positioned to address these type of resource issues.
About Mark Allen
Mark Allen is co-author of the book Master Data Management in Practice: Achieving True Customer MDM (John Wiley & Sons, 2011) and is a senior IT consultant and enterprise data governance lead at WellPoint, Inc.
Prior to WellPoint, Mark was a senior program manager in customer operation groups at both Sun Microsystems and Oracle Corporation. At Sun Microsystems, Mark served as the lead data steward for the customer data domain throughout the planning and implementation of Sun’s customer data hub.
Mark has more than 20 years of data management and project management experience including extensive planning and deployment experience with customer master initiatives, data governance, and leading data quality management practices.
Mark has been a speaker at Data Governance and Information Quality (DGIQ) conferences and has served on various customer advisory boards focused on sharing and enhancing MDM and data governance practices.