How to build a Data Governance Program by Stealth: Introducing the Lighthouse Projects Concept

Blue Down Arrow

Chances are, you’re reading this article because you’ve been asked to launch, or support, some form of data governance initiative.

Firstly, congratulations. You’re embarking on a worthy endeavour. 

Here at Oakland, our Data Governance Team have trodden a similar path as we guided our clients across multiple data governance implementations. The results have been profoundly beneficial; often in ways the organisation had never considered at the outset.

Secondly, having delivered a number of these programs, we’re in the fortunate position to look in the rear-view mirror and identify those critical-path activities that were the keys to success. We’ve curated several of these learnings in this article; the first in a series of best-practices we’re keen to share with our clients and community.

Finally, we wanted to provide a ‘warts and all’ account of what you’re about to experience as you embark on a data governance journey. As you’ve no doubt experienced, every project is unique, so if you would like to discuss how our approach and experience could help within the context of your data initiative, please book a discovery call to discuss further.

Building the foundation for a successful data governance program


What motivates the next generation of Data Governance stakeholders?

We typically see the demand for an enterprise-wide data governance process initiated as part of a broader data foundation or enterprise-wide digital transformation. Many clients we work with are keen to embark on their data journey in a quest for monetisation and the creation of new opportunities through innovation.

There is a growing realisation that better enterprise data leads to increased business performance, smarter decision-making and significantly reduced risk. As a result of these strategic triggers, we increasingly find executive stakeholders reaching out for guidance.

Making enterprise data management changes centrally in these organisations can be a challenge because they are often organised around different segments or geographies. There is also a lot of autonomy in large organisations, so to make changes centrally you’ve got to be business-led.

People have to see the value in the changes you want to make.

The Data Governance Process: ‘What vs How’

If you’ve spent time sifting through the vast amount of books, articles, webinars and frameworks on the topic of data governance, you’ll soon realise that the ‘whats’ of data governance are reasonably well understood and readily available. Most organisations we speak to know what is required to get the basics of data governance in place.

However, particularly with larger, federated businesses, the translation from the ‘what’ into the ‘how’ is where many data governance programs begin to struggle.

This question of how to deliver data governance in a way that engages and motivates the business, whilst simultaneously laying a future data governance foundation, is a challenge we’ve obsessed over for many years at Oakland.

It’s through this obsession that led us to the realisation that many business cases are not positioned correctly to launch a successful platform for data governance.

Re-thinking the data governance business case

In the early beginnings of our data governance practice, we would collaborate with clients to build defensible business cases for data governance, often as as a request to help construct a large data transformation program.

The Data Management fundamentals

Large-scale data and digital transformation programs typically incorporate a broad swathe of data fundamentals such as:

  • Data Strategy
  • Master Data Management 
  • Data Analytics (Business Intelligence, Data Warehouse, Data Lakes)
  • Data Quality Management
  • Metadata Management and Data Lineage
  • Data Stewardship
  • Data Protection and Regulatory Compliance
  • Big Data

If you’re embarking on a similar data transformation, there are comprehensive frameworks such as DAMA DMBOKv2 that provide a comprehensive ‘what’ list of data elements required for the ‘mega-transformation’ projects.

Whilst it seems logical to combine data governance with all the other elements of a large-scale data transformation journey, you will often bump into challenges when constructing the data governance business case and getting it approved.

Challenges with the Data Governance business case

Some of the issues you’ll face when constructing a data governance business case may include:

  • Cost Quantification: The costs associated with poor data quality are difficult to quantify (they are often not even recognised, let alone measured)
  • Limited Interest: Things that could easily be measured (e.g. cost of excessive manual cleansing, customer complaints leading to administrative overhead) may not generate a sufficiently compelling business case
  • Lack of Attribution: It can be challenging to attribute any savings realised to a particular business unit (e.g where data quality improvements are made spanning multiple central and local departments)
  • Excessive Complexity: Identifying broader use cases for new revenue opportunities from data exploitation requires detailed technical and business understanding (e.g. building a single customer view or Master Data Management (MDM) capability will provide benefits, but they’re a non-trivial exercise) 
The 3 pillars of a successful Data Governance business case

To address the previous challenges, we discovered that a more pragmatic approach is to build a data governance business case with a simpler, three pillar approach:

  1. Limit the early investment – plan a modest level of investment, resisting the urge to build an ‘all-encompassing’ data governance program
  2. Focus on value creation – deliver a series of tactical ‘Lighthouse Projects’ that create value for the business
  3. Fly under the radar – gradually establish your data governance program and enabling activity in ‘stealth mode’, locking in the gains of each Lighthouse Project as you progress


Launching your data governance framework with stealth: Applying the ‘Lighthouse Project’ approach

When we switched to helping clients implement data governance through ‘Lighthouse Projects‘, we found that executives were more likely to get on board when they were investing in quick wins that delivered faster returns, as opposed to much larger data governance programs that often grapple for attention amongst so many other transformation programs.

Larger data governance programs often run into additional problems once they get past the business case phase, such as:

  • Interest stagnation: Too much emphasis is placed on building the data governance foundation, at the expense of solving problems that are attractive to the business.
  • Lack of ownership: Data cuts across organisational boundaries, so this requires new data ownership and accountability structures that are often lacking from data governance initiatives.
  • Hidden data risks: It can be challenging figuring out where the risks are within your present data landscape, and which ones require attention.
  • Lack of adaptation: You can’t copy and paste a data governance framework in the hope that ‘one-size-fits-all’. Instead, you need to adapt your implementation with a mix of different tactics over the long-term.

Through the delivery of high-value Lighthouse Projects, the foundations of your data governance framework can start to be deployed, but without the burden and resistance associated with a much larger program.


Setting up your data governance strategy for success

We typically aim to apply a 3-stage process when our data governance team works with clients:

Stage 1: Set the Direction (Defining the Data Governance Strategy)

The best approaches to data governance and the ‘exploitation’ of data must first seek to capture ‘management intent’. Once this is understood, you can then ‘smuggle’ data governance into the business using projects that engage and excite people whilst laying the necessary real-world data governance policies and supporting structures, such as data stewardship and accountability frameworks for data stakeholders.

Technical skills and tools are important but should support, not lead, your data governance strategy. For example, a data governance tool is meaningless if you don’t have a committed and trained group of data stewards capable of building out a corporate data stewardship capability.

Key Dimensions of Data Governance Strategy Readiness

There are three key dimensions that will provide the focus for defining the data governance strategy:

  1. Business Capabilities: Understand current business priorities for data and test coherence through policies, governance and leadership structures.
  2. Operational Capabilities: Sample key processes and data lifecycles to determine alignment with key data management principles. Gather evidence of operational behaviours, both positive and negative.
  3. Data Capabilities: Understand the data architecture and its footprint, together with any existing data management capabilities (e.g. tooling for data quality, master data management, metadata, data maturity metrics, data dictionary of data assets, data definitions completeness, data domains under governance, data owners assigned to existing data stewardship controls etc.)


Stage 2: Plan the Execution (Implement the Data Governance Strategy)

By appling some intentional planning in a multi-faceted approach, you will be able to:

  • Balance the various leadership (and political) drivers, whilst establishing clear ownership of the key levers of change (e.g. financial approvals)
  • Manage the narrative, particularly around the tendency for the inflated expectations that are so often associated large data initiatives
  • Curate and support your portfolio of ‘Lighthouse Projects’ (POCs) on two-fronts:
    • Business outcome success stories that everyone can get excited about
    • Targeted interventions to address areas of poor practice/high risk (e.g. regulatory compliance and data protection weaknesses)
  • Drive the necessary data governance policy, standards, frameworks and overall data governance process, that support the strategic direction of travel

Inputs to Stage 2

To prepare your Stage 2 planning activity, you will typically require various supporting documentation and insights such as:

  • Existing data strategy and data foundation plans
  • The known ‘whats’ of a preferred data governance process/strategy
  • Previous data maturity assessments, baseline reviews and data audits
  • Any other supporting documentation

 

Preparing the Data Governance Strategy Plan

During this phase, we would typically expect to carry out the following activities:

  1. In-depth interviews: Exploring possible root-causes of issues and areas of opportunity
  2. Landscape review: Investigate known data-related initiatives to identify focus areas and avenues for collaboration/harmonisation
  3. Deep dives and sampling: Reviewing the key data lifecycles and data management processes to identify good/poor practices as a reference point
  4. Data management maturity review: Provide a baseline data maturity assessment through quantitative and qualitative metrics, covering all aspects of people, process, systems, technology and data governance (at Oakland, our clients leverage our proprietary rapid data maturity assessment – to request details click here)
  5. Central capability review: You will need to review the available capacity and capability of your central data function against their ability to deliver on your vision for data.

Following this initial preparation work, you will have some clear indication of where the major gaps, opportunities and readiness exists within your current data landscape.

Creating the ‘Commissioning Document’

At Oakland, we recommend creating a ‘commissioning document’ that aligns senior management and various stakeholders around a set of common goals and plan to move forward.

This document is likely to include:

  • Components of the overall data governance strategy with an order and intensity in which they should be progressed
  • Assessment of the central support required (e.g. additional comms, policy development, standards)
  • Identification of any technical needs, beyond the POCs (e.g. Master Data Management (MDM), Data Quality, Metadata and Data Dictionary for Data Definitions and Data Lineage etc.)
  • Key delivery risks and issues

The plan needs to be resilient enough to cope with any organisational or leadership changes, and coordinated in a way that is most likely to succeed in the medium-term.

Typically, a core team will continue to support delivery of the plan as it rolls on into future phases.


Stage 3: Executing your Lighthouse Projects (Proofs of Concept):

For the remainder of this article, we’re going to focus on the process of launching ‘Lighthouse Projects’ because we find that when executed correctly, they form the driving force for an effective data governance strategy.

 

A deeper dive into the benefits of Lighthouse Projects

Lighthouse Projects (POCs) form a crucial component of your overall data governance strategy as they engage the business in tactical improvements that capture attention. These improvements create success stories that build momentum, reduce risk and establish confidence in your approach.

Overcoming Resistance to Data Management Change

Many organisations suffer from ‘change fatigue’. 

When starting out, you’re going to be ‘smuggling’ data governance into the organisation so make sure your project’s mission is 100% focused on the ‘pains and gains’ that those in the business already recognise.

The first rule of Data Governance – don’t talk about Data Governance

You don’t necessarily want to label your first Lighthouse Project as a ‘Data Governance Initiative’ because you’ll often face enough resistance as it is, without introducing a phrase that no-one in the business understands.

For example, with one client, we selected a project that addressed an asset data quality headache that had been frustrating the business for some time. 

We could have adopted a top-down approach and waited to build out accountability structures for data governance, create data quality assessment frameworks, train the business in data literacy and select a team of data stewards to manage the information chain.

Instead, we addressed the root-cause of why the data asset system was creating poor data quality.

We then introduced a system for onboarding data stewards, discovering metadata, defining data quality metrics and controls – all the elements of a data governance capability, but in a much shorter timeframe than if we had waited for a full data governance program to be rolled out.

Which data management improvements should you consider for a Lighthouse Project? 

When considering the monetisation opportunities of your Lighthouse Projects, we recommend ‘two waves of attack’.

Wave 1: Initial assessment against an agreed project selection criterion

This phase should identify a prioritised list of candidate POCs drawing on the selection criteria previously discussed. For example, you may collapse a list of candidate projects from 20 down to 5 possible options.

Sample criteria and considerations for Lighthouse Project selection

Every organisation has scope for data improvement, so your initial challenge will be determining which Lighthouse Projects should be candidates for consideration.

Having delivered many of these projects, we’ve found the following advice to be a good starting point for selecting the best data management/business improvement projects to kick off your strategy of ‘data governance by stealth’:

  1. Identify which projects are most likely to succeed – reduce the scope rather than fail an overly ambitious target
  2. Clearly define the scope – understand the specifics of the business problem being solved and its technical / data viability
  3. Focus on business-led projects – look for projects that rely on active contribution from the business
  4. Look for balance across the POCs – shortlist a ‘project portfolio’ that will nurture good ideas and opportunities, whilst tackling areas of poor data practice
  5. Consider what happens at the end of the POC – is the purpose to prove viability (and perhaps discard afterwards) or build a fully supported solution?

Wave 2: Investigation of each candidate POC

In this phase, you will start by validating project feasibility and scope, before arriving at an attractive set of options for more detailed investigation that will be completed at the end of Wave 2.

Investigation focus areas

Each investigation will typically focus on areas such as:

  • Use case definition
  • Business processes
  • Technology
  • Data availability
  • Data quality and associated metrics
  • Risks and challenges for delivery
  • Delivery planning

Upon completion, Wave 2 will select a focused set of projects that fully address several frustrating business problems. These projects will allow you to introduce key elements of your data governance roadmap as a side-benefit, but without the fanfare and resistance so often associated with a major data governance program.

Depending on the scale of each candidate project, it typically requires 10-20 days to complete the POC investigation. 

What team will you need for POC investigation?

Within our data strategy and governance consulting practice, we typically deploy a core investigation team consisting of:

  • Business translator (SME)
  • Business analyst
  • Technical data architect

The exact make-up of the investigation team needs to adapt to the specifics of your data, process, and technology landscape. 

For example, some projects may be deeply technical in nature, requiring expertise in data interface accessibility, workflow automation and Big Data platforms. 

Other times, the project requires ‘softer’ skills in the form of data literacy at the grassroots level, educating a team of stakeholders to take accountability for decision-making, or simply coordinating a team of data stewards to set more robust data definitions and assessment metrics for their data assets.

The key is to work with a data governance team who have the experience to address the complex and varied types of data challenges that typify the modern enterprise.

Next Steps

Building out a data governance program for the first time is a challenging undertaking. 

Our recommendation is to first gain an overview of the ‘whats’ of data governance and then seek an experienced implementation partner who can guide you through the process of delivering smaller, business-driven, ‘Lighthouse Projects’ that deliver incremental data governance milestones.

Don’t be taken in by the industry trend of building out all-encompassing data governance frameworks until you’ve at least demonstrated repeated value from lightweight data improvement projects that help you ‘smuggle’ in the fundamentals of data governance.

You’ll find that when you come to expand your reach with data governance, you already have a captive executive audience who are primed and eager to support your continued success.

If you want to discuss any of the ideas presented, and how they would benefit your data initiative, please contact us to arrange a call.

Previous Post: What is a Digital Twin? Definition and examples of next-level performance improvementNext Post: The Evolution of Project Analytics