The Many Moving Parts of Governance
Originally published August 22, 2006
Earlier this year, a top-ten financial services firm asked Knightsbridge Solutions for assistance in assessing their enterprise-wide reporting, metrics, and analytic capabilities. The business units were struggling to retrieve required data, and the technology organization was too understaffed to satisfy business data demands. Not an atypical situation in the financial services industry. However, the results of our initial assessment revealed an interesting twist.
As part of the assessment, management wanted to know what could be done immediately to provide relief for pent-up data demands while longer-term solutions were being implemented. To meet that challenge, we conducted more than 70 interviews throughout the enterprise with both business and technology staff. We learned that while changes in architecture would result in long-term gains, significant short-term wins could be achieved by addressing process improvements.
As shown in Figure 1, most of the comments and critiques that surfaced in the interviews centered on business processes, prioritization of projects, data quality and access. The technical issues raised were serious, and solving them would take months. We were looking for quick wins that could be realized in weeks or, at most, a couple of months. Shortfalls such as the following intrigued us:
As we explored these in more detail, it became apparent that no coordinated management structure existed to comprehensively see all of the issues and respond with a unified action plan. Reporting and analytics were highly siloed by business unit – again, not atypical in this environment.
We had a choice of either attacking each symptom within each silo or trying to address the issues from an enterprise perspective. We chose the latter course and selected governance as the vehicle to capture meaningful results.
As we ventured down the governance path at the client, it proved helpful to outline our perspective on governance and how to maximize its impact.
The true function of governance is to actively link integrated business and technology teams with corporate and strategic initiatives. Within this context, governance becomes an integral part of enterprise line management. Executed properly, the governance function can actively and effectively reallocate business, technology, reporting and analytic resources to align with rapidly changing market demands.
As seen in Figure 2, a viable approach might be to divide governance activities across the major technology activities. Each area has unique governance needs and standards. This approach eliminates trying to fit one size to all. It also enables each technology area to have a different organizational structure – all linked to the enterprise.
Figure 2: Division of Governance Activities
Figure 3: Tactical Infrastructure Framework
But first, we will explore our data and reporting governance infrastructure. One of the shortcomings of many governance functions is the lack of “organizational accountability.” Many governance groups are comprised of disparate committees – for example, a “metrics committee” that works on defining performance metrics for a given silo and a “standards committee” that publishes standards but has no authority to enforce them. At the same time, a separate “demand management committee” could be prioritizing projects against limited resources based on criteria used to maximize resources rather than maximizing achievement of corporate/strategic initiatives. Finally, one of the most common committees in most organizations is the “technology steering committee” that must approve all technology project requests.
It is this type of uncoordinated committee structure that has rendered many governance organizations unproductive and confusing.
The important point is that these formerly disparate committees in most institutions should report into an overarching governance function that can organize their activities, align their efforts with corporate and strategic initiatives, and hold individuals accountable for performance.
In our Figure 3 example, these tools could easily have been a development center of excellence or a network services center. The key point is that the governance organization must have accountability for its “tool set” and must be charged with linking the activities of the business and technology teams with corporate objectives and strategic initiatives.
Governance Organization – Alternatives
Let’s look at the four alternatives.
The Centralize Everything Model
Figure 4: Centralize Everything Model
In this model, the C-level executives play a guidance role. They provide active support, serve as the “court of last resort” for line of business appeals, approve all standards and, most importantly, agree to accept only reports/analyses where certified data was used. This last point is critical, for it cements a commitment to an advanced data quality program referred to as data certification.
In this model, the enterprise governance executive is a full-time position, not a part-time function of the CIO. This executive is held accountable for governance across the enterprise. To conduct that role in a “centralized everything” framework, all of the tactical governance leaders (data and reporting, network and infrastructure, etc.) would report into this individual.
As shown in the data and reporting tactical governance arena, there are three groups interacting under the data and reporting tactical area. The governance business partners are aligned with each line of business (LOB). Under this model, areas such as finance, risk, credit administration, audit, etc., would be considered as LOBs. A good way to think about the governance business partners would be as data and reporting stewards. There are also pillar managers (see Figure 2) who interact with the LOB representatives. When we speak of interactions at this level, we are referring to things like support for activities such as self-service/parameterized reporting, project planning, estimation, prioritization and management, along with day-to-day interactions around data and reporting.
At the governance execution layer, we have the tight-knit integration of business and technology staff working to build new reports, perform complex data extracts, add data attributes to the data stores, create marts, etc. At this layer, there is extensive interaction with the pillar managers and the governance/LOB teams.
Again, with this model there is tight, centralized control that forces an adherence to standards, accountability, etc. This can translate into less flexibility, greater overhead and slower response times. Also, as priorities cross LOBs, there may be difficulty in getting buy-in and delivering. Execution is tightly aligned with LOB silos, which can be both an advantage and a deterrent.
The Centralized Standards Model
Figure 5: Centralized Standards Model
It is interesting to note that tactical governance across all three arenas could be managed through a single group of pillar managers who manage an expanded number of functional pillars that would cover data and reporting, application/system development, and network and infrastructure. Figure 6 illustrates this concept.
Figure 6: Functional Pillar Managers
Enforcing standards for the “enterprise good” would be challenging since the governance business partners could have split loyalties due to the direct reporting to the LOBs and indirect reporting to the governance executive.
The Centralized Management by Subject Matter Model
As seen in Figure 7, the role of the governance executive has been replaced with a committee consisting of direct reports to the C-level positions. In turn, this group rotates the chairmanship of the committee. This is another way of isolating the C-level executives from being put in the position of making every decision related to IT and governance. This committee will manage strategic planning and governance standards across the enterprise.
Subject area governance bodies work closely with SIGs versus lines of business. The SIGs can cross LOBs just as subjects do – e.g., finance, risk, operations, lending, etc. Subject area governance bodies and SIGs manage and enforce the governance pillars as well as provide the execution teams with priorities, standards and direction.
Management and enforcement of the functional pillars is performed by each of the subject area governance bodies in conformance with enterprise standards set by the tier-two governance committee.
This is a point solution. To be effective, enterprise transactional-level data needs to be loaded in the enterprise data warehouse (EDW). All major areas will abandon data marts and manage their lines of business utilizing the EDW.
Figure 7: Centralized Management by Subject Matter Model
The Centralized Governance Council Model
Figure 8: Centralized Governance Council Model
Recent articles by Duffie Brunson
Copyright 2004 — 2020. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC