Services /
Data governance
Data
governance
implementation.
Establishing clear accountability for your most critical data so that quality problems have named owners, definitions stop causing disputes and AI or analytics investments can actually deliver what they promise.
The right conversation to have
Data governance is an organisational discipline, not a technology project.
The most common failure mode in data governance is not poor framework design: it's insufficient executive sponsorship combined with a programme that gets framed as a technology or data platform project from the start.
When governance is owned by IT, recommendations get ignored outside IT. When it is owned by risk and compliance, business units disengage and the framework becomes bureaucracy. When there is no clear owner at all, the engagement stalls at diagnosis and never recovers.
The conversation that matters is about accountability and decision rights — who is responsible for the accuracy of your most critical data and what happens when that data is wrong. Technology is an enabler of that accountability: it is not a substitute for it.
A Medasi data governance engagement lands governance as a business capability, owned at executive level, with a practical framework that an organisation of your scale can realistically implement and sustain.
How we frame the conversation
Not this …
But this …
We need a data governance programme
You need clear accountability for your most critical data so you can stop burning resource on rework.
Your data maturity is low.
Two senior people can give the board different numbers for the same metric. That is the problem we are solving.
We will implement a framework.
In 90 days, you will have named owners for your ten most critical datasets and a process to resolve quality disputes.
The framework
Six components.
Practical by design. Built to last.
Most scaling organisations do not need an overly sophisticated governance architecture. They need clear answers to three questions — who owns it, how are quality problems resolved and who defines what the data means — for their ten to twenty most critical datasets. This framework delivers that.
Data domains
Logical groupings of data — customer, finance, operations, product, people, supplier — that form the boundaries of ownership and make accountability conversations tractable. Designed around business accountability, not systems.
Ownership model
Named domain owners (director or above) and data stewards for every domain. Without names, accountability belongs to nobody. Domain owners must have real authority: if they cannot instruct a change, they are not actually owners.
Data policies
Rules governing how data is created, used, changed and retired. Gives day-to-day decisions a reference point without requiring escalation, and makes the governance framework operational rather than aspirational.
Data standards and business glossary
Agreed definitions, formats and business rules for critical data elements, resolving the most common source of management information conflict. Two people using the same word to mean different things is a structural problem, not a communication one.
Data quality framework
The six dimensions, metrics and thresholds used to assess and report quality across priority datasets. Creates a shared, measurable language for data health — replacing subjective complaints with objective, trackable baselines.
Governance processes
How issues are triaged, disputes resolved, policies updated and the governance council operates. A framework without defined processes decays within six months. These are the mechanisms that keep governance alive after delivery ends.
the 90-day approach
Risk reduction before speed.
Foundations before features.
The 90-day horizon is the primary planning unit. Organisations at this scale lose focus on multi-year plans. We sequence by what reduces risk first, not what is easiest to deliver first.
Days 1-30
Establish foundations
Build the structural conditions for governance to take hold — including the executive conversations that determine whether it will survive.
Domain model agreed with senior leadership
Domain owners and stewards nominated
Governance lead identified
Priority datasets scoped for quality analysis
Framework documentation in client hands
Days 31-60
Build and activate
The framework goes live. The governance council meets. Quality data starts to inform decisions rather than frustrate them.
First governance council meeting held
Quality scorecards live for priority domains
Business glossary — first entries agreed
Data quality profiling complete
Remediation backlog populated and prioritised
Two to three quick wins identified and underway
Days 61-90
Stabilise and hand over
The programme transitions from consultant-led to client-led. Handover is planned from day one — not added at the end.
Council operating without consultant facilitation
Quick wins delivered and reported to sponsor
Executive sponsor quarterly review held
Handover readiness criteria confirmed
Maintenance regime in place and owned
All documentation in client-owned repository
The ownership model
Named accountability at every level.
No committees without owners.
Executive level
Executive sponsor
Typically the CFO, COO or Chief Digital Officer. Provides the mandate and authority for the programme. Makes final decisions on unresolved cross-domain disputes. Chairs the quarterly governance review.
Programme level
Data governance lead
The day-to-day owner of the governance programme. Maintains the framework, facilitates the council, tracks remediation and manages the business glossary. A relationship management and organisational role — not a technical data role.
Domain level
Data domain owner
A senior leader (director or above) accountable for the quality, availability and appropriate use of data within a defined domain. One owner per domain. Makes final calls on intra-domain disputes and represents the domain in the governance council.
Operational level
Data steward
An operational manager or senior analyst responsible for day-to-day governance within the domain — monitoring quality, triaging issues, maintaining glossary entries and enforcing standards. The first escalation point for quality disputes.
Governance forum
Data governance council
The cross-functional forum where domain owners convene to make shared decisions — not to receive status updates. Typically meets monthly. Agenda items include cross-domain quality issues, definition disputes, policy updates and programme progress. The governance lead chairs; the executive sponsor attends quarterly.
Data quality analysis
Six dimensions.
Assessed against your most critical datasets.
Before designing the framework, we establish an evidence-based picture of where quality fails, why it fails and what the business consequence is. Not every dimension matters equally for every dataset — we focus effort where it matters most.
Completeness
The degree to which required fields are populated across records in the dataset.
Mandatory fields blank; records created with placeholder or default values to satisfy system requirements.
Accuracy
The degree to which data correctly describes the real-world entity it represents.
Contact details out of date; financial figures that do not reconcile across systems or to source.
Consistency
The degree to which data is the same across systems, reports and representations.
Different values for the same metric appearing in different reports — the most common cause of MI disputes.
Timeliness
The degree to which data is available when needed and reflects current reality.
Reports based on data days or weeks old; batch refresh failures; records that are technically populated but practically stale.
Validity
The degree to which data conforms to defined formats, types and business rules.
Free text in structured fields; invalid codes, reference values or formats that bypass system validation.
Uniqueness
The degree to which each entity appears exactly once within the dataset.
Duplicate customer, product or supplier records that create downstream errors in reporting, billing and service.
What you get
Board-ready outputs at every stage of the engagement.
Board-level one-pager
A standalone summary of governance position, maturity across all six components, the top three risks of inaction, and a clear recommendation — designed to be read independently of any supporting document.
Maturity scorecard
A traffic-light assessment across all six framework components — scored against evidence, not aspiration. Updated at each major programme milestone and included in board and executive reporting.
Data quality report
A structured assessment of priority datasets across all six quality dimensions — with root cause analysis, business impact ratings and a prioritised remediation list ranked by impact and effort.
Governance framework and operating model
The fully designed framework — domain model, ownership structure, policies, business glossary, quality framework and governance processes — with role profiles for every accountability.
90-day roadmap
A sequenced, phased delivery plan across the three horizons — sequenced by risk reduction, with named owners, realistic milestones and clear handover criteria for each phase.
Maintenance regime
The recurring activities, cadences and ownership needed to keep the framework current after delivery — including the quality scorecard, remediation backlog, governance council rhythm and annual maturity review.
Who this is for
Leaders whose decisions are being undermined by data they cannot trust.
Data governance engagements are most valuable when data quality is visibly hurting the business — in MI disputes, AI initiatives that are underdelivering, operational rework or regulatory pressure. They are also the right investment before committing significant capital to data platforms or AI tooling, where the absence of governance is the most likely reason the investment will disappoint.
Chief Financial Officer
Motivated by financial accuracy and eliminating MI conflict at board level
Chief Operating Officer
Dealing with operational rework and reconciliation caused by data quality failures
Chief Digital or Technology Officer
Whose AI or analytics initiatives are stalled or underperforming due to data quality
Chief Executive Officer
Whose strategic decisions are being undermined by data they cannot rely on
When organisations engage us
MI conflict at board or ExCo level
Two senior people can give the board different numbers for the same metric, and nobody knows which is correct or who is accountable for resolving it.
AI or analytics investment stalled
A digital or AI initiative is underperforming and the team cites data quality as the primary cause. The programme is sound; the foundations are not.
Regulatory or audit pressure
Audit findings or regulatory requirements demand traceable accountability for data accuracy, and the organisation does not currently have the structures to provide it.
Pre-platform investment
A significant investment in data infrastructure, a data warehouse or an AI platform is being considered — and governance is the prerequisite that needs to be established first.
Related services
Start with clarity
Determine who is accountable for your data — and what to do when it's wrong.
A conversation about your data quality challenges and what a governance engagement would realistically involve for an organisation of your scale.