Data Governance Proposal Deck

About this template

Twelve slides for a governance proposal

The deck you take to the meeting that funds and staffs data governance. It argues from a specific symptom rather than from principles: the same metric name returns a different value in every team, and that gets fixed by naming owners, writing definitions down and closing off raw access. Because it is a proposal, the last thing on screen is a phased plan with a duration on it, not a manifesto.

Six agenda lines, three body pairs

Slide 2 runs the problem today, ownership and accountability, standards and dictionary, today and once governance is in place, key metrics, trend analysis, and the theme prints an 01 / 06 counter beside each line. The first three carry a divider and a body slide across slides 3 to 8, line 04 is the comparison on slide 9, and the cards and bars are slides 10 and 11. Quality checks and access rights are not built as their own pairs here, so duplicate slides 7 and 8 for each and extend the agenda.

The three body slides

1. The problem today: “Three rival definitions of an active customer”, “Month-end close numbers reconciled all over again”, “42 people can reach the raw source”. Keep all three as measured facts. A count of rival definitions beats an adjective.
2. Ownership and accountability: “Owners named for 6 core domains”, “Definition changes take the owner’s approval”, “Ownership refreshed every quarter”. Owner means a named person with authority over the definition, not a team mailbox.
3. Standards and dictionary: “Definition, formula and source for every metric”, “Naming rules and code values standardized”, “Change history kept as versions”. These three are the minimum row of a metric dictionary; drop the source and the disputes come back.

The metric cards and the trend bars

Slide 10 holds “Domains assigned” 6, “People with raw access” 42, “Conflicting definitions” 3 and “Phase 1 duration (wks)” 12. Two of these are damage and two are plan, which is the shape that gets a proposal approved. The 42 and the 3 repeat slide 4, so edit them together. Slide 11 falls from “Before start” 100 through “Phase 1” 64, “Phase 2” 35 and “Phase 3” 12; it is an index of the problem shrinking, not a volume, so relabel it with what you are actually counting – open definition conflicts, say – or a falling bar reads as decline. Replace the “Unit: index” line accordingly.

The comparison slide

Slide 9 sets “Today” against “With governance”. “Metric definitions differ by department” faces “One registered definition, used everywhere”, “Queries run straight against the source” faces “Role-based access with approval steps”, and “Errors found only after someone reports them” faces “Automated checks block it before release”. Each right-hand line is something you are asking to be funded, so keep them concrete.

Access, personal data and Korean law

The access slides in this deck are written for organizations under Korea’s Personal Information Protection Act, so an access pair should describe the controls that law and the ISMS-P certification scheme ask for: approval-bound, time-limited access to personal data, pseudonymized datasets by default, retained access logs, and a privacy impact assessment where required. Do not swap in a GDPR frame – the lawful bases, the pseudonymization rules and the retention duties differ, and a reviewer working to the Korean act will stop reading.

What goes wrong

• Proposing a dictionary with no owners. A definition nobody can approve a change to is a wiki page, and it goes stale in a quarter.
• Asking for tooling money before naming the six domains. The tool is cheap; the ownership argument is the expensive part.
• Leaving the 12-week Phase 1 card in while describing a two-year program. The closing line is “Making the numbers trustworthy” and the plan under it has to be one someone can commit to.