Centers of Excellence (COEs) are your path to beating your competition

Part 3 – Justification, Funding, and Placement

In this third and last installment, I will focus on:

• Staffing and Funding Sources for the Analytics COEI
• Justifying and Funding the Analytics COEI
• Positioning of the Analytics COEI
• Types of Analytics COEI implementations.

McKinsey found that only 20% of companies have maximized their potential and achieved advanced analytics at scale, even while most companies understand the importance.

PwC discovered that 61% of companies say the key to reaching strategic goals is collaborating more across functions, paired with faster decision-making and innovation.

The next set of questions that we typically get have to do with staffing and funding sources.

The Analytics COEI needs to be staffed with key resources that understand business and data issues, the company’s business processes, plans, drivers and strategies, and what LOBs need to be supported, when, and how. In many implementations, some of these resources are not assigned to the COEI on a full-time basis but more on an as-needed basis, making funding the COEI and charging for its services more complex. Sources for these resources include:

• Data Management processes resources from the following DG disciplines:
  • Data Quality, Integrity, Lineage and Metadata Capture and Management
  • Data Governance (representing each LOB)
  • Security and access management
• IT:
  • Data Integration: ETL/CDC
  • Data and Solution Architecture and Data Modeling
  • Data Analysis and Profiling
• Business Groups/LOBs:
  • Business Analysts
  • Business Data Stewards (at least one per LOB)
  • Business operations and process SMEs (at least one per LOB)
  • Business Shadow IT groups (in many cases 25-50% of these groups became part of the COEI)
• Outside Consultants:
• SMEs to serve as SWAT team members and address specific staffing and skills gaps

The Analytics COEI is basically a cost center and needs to be funded. How companies pay for a COEI depends on many factors such as how the COEI is implemented, available chargeback models, its ownership and leadership, and how many LOBs it supports.

Common funding sources include the following organizations: LOBs, CDO, CIO/IT/CTO, CFO, COO, revenue from tools and license rationalization, and repurposing staff.

Some companies calculate the charge-back funding model of the COEI using criteria such as:

• Number of resources saved per LOB (especially from shadow IT organizations) and the number of resources saved in the IT organization
• Number of reports and analytics per LOB as well as per C-Level executive
• Specific LOB and enterprise metrics and KPIs that have improved over time and specific SLAs that were met and exceeded over time

In order to justify the Analytics COEI, here are the key measures used to calculate what each LOB will contribute to the Analytics COEI budget. Each company uses different budgeting calculations, but here are some general guidelines used (per LOB):

• Number of analytics the LOB will no longer produce on their own and new analytics that have been added YOY (reports, dashboards, heatmaps and scorecards)
• Percentage increase in reporting quality, reliability, and timeliness
• Percentage of LOB analysts that have self-service access to the data
• Percentage of reduced lead time (when IT produces existing and new analytics vs. when the Analytics COEI does it)
• LOB resources that have been repurposed, such as data wranglers that have gone back to being business data analysts, actuaries, and underwriters. Many of these LOB shadow IT resources become part of the Analytics COEI
• Percentage increase in SLAs met or exceeded when using the Analytics COEI vs. previously
• Number of complex analytics (AI/ML/Data Mining and Trending) required

As the Analytics COEI matures and more self-service capabilities are introduced, other ways to deliver value and charge for it can be introduced:

• Data-as-a-Service:  Sourcing and Provisioning, Modeling, Management, Lineage, Quality served via self-service analytics
• Analytics-as-a-Service:  Predictive Modeling and BI, Prescriptive analytics, Data Mining served via self-service analytics or via the delivery of preprocessed data sets
• Insights-as-a-Service:   ML/AI based analytics, self-learning and alerting analytics delivered to end-users directly or via their self-service portal
• Reporting-as-a-Service:  Reports, Dashboards, Heatmaps, Scorecards, KPIs, Drill Up/Down (hosted and accessed directly)

When multiple LOBs are supported, they are each usually charged according to the resources they consume. Initial budget allocations are based on:

• Current size of LOB analytics staff and their shadow IT staff
• Number of analytics the LOB currently produces (number of reports, dashboards, etc.)
• Estimates of what the Analytics COEI will require to support a LOB (staffing, tools, skillsets)
• Whether or not the LOB will require self-service capabilities
• Other factors that vary from company to company

Once budgets are calculated, they become part of the cost center expenses they will belong to (CDO, CIO, COO, etc.)

The question I always get focuses on the location of the Analytics COEI in the enterprise. Where does the Analytics COEI organization belong? The best place is under the CDO (at least in my opinion and based on my experience), but many organizations have placed the COEI under different groups. If the CDO reports to the CIO, place the COEI under the COO, the Chief Transformation officer or the CFO.

I have seen the Analytics COEI implemented under different executive/C-suite leaders, but the most successful implementations were when the Analytics COEI was part of the CDO’s organization, and the CDO reported directly to the CEO or president. Figure 1 (above) outlines the other potential placements of the Analytics COEI.

Another frequent question is about the Analytics COEI focuses on the COEI organizational type. Here are four different types with some pros and cons:

The Analytics COEI implemented as part of IT under the CIO: The Analytics COEI will provide support to multiple LOBs on an as needed basis, just as other IT resources and groups do.

This implementation is just an extension of the IT group’s responsibilities. It is not recommended because:

• Resources are not dedicated
• IT’s history of successfully implementing analytics or Data Management COEIs is very poor
• Resources do not get the appropriate training and cross-training
• Very little time is invested in developing standards, best practices, research, and innovation
• This is probablywhy the enterprise needs a dedicated Analytics COEI in the first place

The Analytics COEI implemented as part of operations under the COO: The Analytics COEI will provide support to multiple LOBs on an as needed basis, just as other IT resources do.

This implementation is better but not optimal because:

• COO experience with an Analytics COEI is limited
• Staffing and other investments are difficult to justify
• Resources do not get the appropriate training or cross-training
• Very little time is invested in developing standards, best practices, research, and innovation
• Operations organization does not usually focus on proactively supporting the LOB analytics functions

The distributed Analytics COEI implemented under the CDO: The Analytics COEI will provide embedded support to multiple LOBs on a fulltime, dedicated basis.

This implementation is highly recommended because:

• COEI staff are dedicated, and many of them are very experienced with and dedicated to a specific LOB requirements/needs
• Staffing and other investments are less difficult to justify
• Chargeback systems are easier to implement
• Resources get the appropriate training and cross training
• Time is invested in developing standards, best practices, research and innovation
• Certain core COEI resources are shared between LOBs (data architects, data modelers, data integration SMEs), while others are dedicated to specific LOBs
• CDOs understand DATA and treat and respect it as a corporate asset

The virtual, fully shared Analytics COEI implemented under the CDO: The Analytics COEI will provide dedicated, virtual, not embedded support to multiple LOBs on a as needed basis.

This implementation is my second favorite and is also recommended because:

• COEI resources are shared and more can be allocated to support a LOB when needed
• Staffing and other investments are easier to justify since usage/consumption is fully understood
• Resources get the appropriate training as well as required cross-training so they can support multiple LOBs
• Time is invested in developing standards, best practices, research and innovation
• CDOs understand DATA and treat and respect it as a corporate asset
• Staff is not dedicated to a specific LOB, so cross-training is very important.

In closing, I will try to address one last question: which COEI is more important, and which one should I start with? I mentioned quite a few different COEIs in my first installment of this blog, and I still recommend that you focus on your Analytics and Integration COEIs first, mostly because that’s where most of the everyday work (data movement, processing, integration and analytics/reporting is). However, I hope I emphasized the importance of Data Management and Governance enough to realize that they are vital to the successful implementation of both the Analytics and the Integration COEI.

Critical point:

The Analytics and Data Integration COEs are very powerful. Implementing them with provide your enterprise with key competitive differentiators.

For additional information on setting up an Integration COE, please take a look at John G. Smith’s e-book Integration Competency Center: An Implementation Methodology.

I hope you’ve found this series informative, helpful, and actionable.

Need help partnering with your business stakeholders to drive more value, build a roadmap, build a COE, or design your data strategy and data governance programs and processes?  Contact Informatica’s Advisory Services group for an initial consultation.