Module 08 Lesson 6 · 7 lessons in this module

Change Management

In brief: The organizational change that MDM governance introduces for AI development teams is more significant than it initially appears. It changes how data is sourced — from informal extraction to governed pipeline access. It changes how training datasets are documented — from undocumented to version-linked with provenance metadata. It changes how model…

Watch: How MDM governance changes the way AI teams operate — not just the tools they use — and why executive sponsor visibility in change management is one of the highest-leverage investments an MDM program can make.

Module support notes

The Organizational Change That AI Programs Require

The organizational change that MDM governance introduces for AI development teams is more significant than it initially appears. It changes how data is sourced — from informal extraction to governed pipeline access. It changes how training datasets are documented — from undocumented to version-linked with provenance metadata. It changes how model performance degradation is interpreted — from a model problem to potentially a data quality problem requiring MDM response. And it changes how retraining is triggered — from ad hoc on performance complaints to systematic on quality drift alerts.

Each of these changes requires not just training on new tools and processes but active support through the transition period when the old working patterns are still more familiar than the new ones.

Note

Adopting MDM governance requires AI teams to change how they work, not just what tools they use. The four operating model changes — governed data sourcing, version-linked training dataset documentation, quality-informed performance interpretation, and drift-triggered retraining — are behavioral changes that require change management support, not just a training session on the new MDM platform access process.

A before-and-after comparison of an AI team's operating model. Before MDM governance: informal data sourcing, no lineage documentation, retraining triggered on performance complaints. After MDM governance: certified governed pipeline access, training datasets linked to golden record versions, lineage documented automatically, retraining triggered by quality drift monitoring. A change management implication note states that AI teams need support transitioning between both models, not just training on the new one.

Adopting MDM governance requires AI teams to change how they work, not just what tools they use.

Change Management and Executive Sponsor Engagement

Executive sponsor engagement in change management is often treated as a program governance function — approving budgets, reviewing progress reports, resolving escalations. But the most effective sponsors treat change management as a leadership responsibility that requires personal visibility.

When a business unit leader resists MDM accountability, a message from the executive sponsor carries more behavioral change potential than any governance policy. When an AI team documents a significant model performance improvement on certified governed data, public recognition from the executive sponsor signals to the rest of the organization that this behavior is valued. Active sponsor visibility in the change management program — not just in governance meetings — is one of the highest-leverage investments an MDM program can make.

Tip

Give executive sponsors a specific change management engagement checklist: visibly champion the program in leadership communications, attend and speak at the go-live event, review adoption metrics monthly in the first six months, personally address resistance at the business unit leadership level, publicly recognize early adopters, and resolve resource conflicts threatening change management activities. Sponsors who do all six consistently produce faster adoption than those who do governance approval alone.

A sponsor engagement checklist for change management activities listing six specific actions: visibly championing the program in leadership communications, attending and speaking at the go-live event, reviewing adoption metrics monthly in the first six months, personally addressing resistance at the business unit leadership level, publicly recognizing early adopters, and resolving resource conflicts that threaten change management activities.

Executive sponsorship of change management requires active visibility, not passive approval.

Implementation delivers the technology. Change management delivers the behavior change that makes the technology valuable.

A four-phase change management timeline across implementation — Prepare, Launch, Embed, and Sustain — with key activities, stakeholder engagement actions, and AI team change milestones at each phase, showing how change management runs alongside and enables every implementation phase.

Implementation delivers the technology. Change management delivers the behavior change that makes the technology valuable.

Lesson progress

0% watched

← Previous Next lesson →