Data governance that keeps running

In the Japan organization of a multinational beverage company, data had piled up by department, with overlapping sources and conflicting definitions. We mapped the landscape, set roles and a council, and left a practice the teams can keep.

  • Multinational beverage company, Japan
  • Data Intelligence
  1. Sharedtaxonomy and a data dictionary
  2. A councilcross-functional Data Governance Council
  3. Ongoinga practice, not a one-off project

The challenge

The data was there. A trustworthy view of the whole business was not.

A multinational beverage company wanted to move from pockets of local reporting to a consistent, enterprise-wide approach to data.

In Japan, data had grown in departmental silos, with overlapping sources and conflicting definitions. A reliable end-to-end view of the business was hard to get.

Policies were ad hoc, access controls were uneven, and core data domains had no single owner.

There was no local MDM, no documented cartography of systems and data, and uneven awareness of governance. The company needed a program that would stabilize today's reporting and prepare for later analytics and AI.

  • Data sat in departments, and definitions conflicted.
  • Core data domains had no owner.
  • No MDM and no map of the data, so the foundation was missing.

What we did

How the work ran

A data-governance program built so it can scale.

  1. 1

    Surface the pain.

    Executive and departmental workshops covered the data lifecycle, from acquisition to maintenance, dissemination and archival, and put the issues in order.

  2. 2

    Map the landscape.

    We mapped sources, systems and flows, and created a shared taxonomy for domains, families, KPIs and key use cases.

  3. 3

    Name the roles and start the council.

    Sponsors, owners and stewards were defined. A cross-functional Data Governance Council gave the program accountability and a cadence.

  4. 4

    Leave the assets, and the habit.

    A data dictionary and catalog, classification standards, business rules and integration guidelines, plus preventative data-quality audits and a documentation hub. Communications, templates and a prioritized roadmap keep it from being a one-off.

Before and after

What changed

TopicBeforeAfter
OwnershipNo owner for core domainsOwners and stewards, and a council that runs the practice
DefinitionsDifferent in each departmentA shared taxonomy and a data dictionary
How the work happensAd hoc, one effort at a timeMethods, tools and templates, as one program
What comes nextNo foundation for analytics or AIA model ready for AI-enabled use cases

The results

Governance the business can keep running.

  • Four structural issues were named: redundant data, no exhaustive view of the data models, unclear ownership, and business processes that were neither understood nor documented.
  • Teams have methods, tools and templates to deploy and maintain governance across business units.
  • The Data Governance Council and the steward model make this a business-led practice, with a data-oriented culture and a path toward AI-enabled use cases.

Services on this project

The same work, for your team.

Technology used

  • Data governance
  • Data catalog
  • Data dictionary
  • Data-quality audits
  • Change management

Service: Data Intelligence

Facing a similar problem?
Tell us where you are today.

Discuss a similar project

We reply within 2 business days.