The Council That Approves but Does Not Decide
Helix Mutual, a 4,000-employee insurer, launched a data governance council 14 months ago. It has 22 stewards, meets monthly, and average attendance is 9. The council has approved 31 policies; internal audit found 19 are not enforced. Finance and Marketing each maintain a certified definition of 'active policyholder' that differ by about 6% of the book, and customer analytics has quietly built a third. Escalations to the council average 37 days, and two decisions were reversed after a vice president objected post-meeting. The CDO wants to strengthen governance by adding a second council layer and more policy templates. You have one quarter and no additional headcount.
1. Diagnose why the governance program is not changing behavior, and state the first intervention you would make in the quarter.
Read the worked answer
The symptom pattern points to weak decision rights and unowned policies, not a shortage of governance artifacts. Attendance of 9 of 22 means the council is not a body of decision-makers; 19 of 31 unenforced policies means approvals were decoupled from operational accountability; 37-day escalations and post-meeting reversals mean decisions lack sponsor-backed authority. Adding templates or a second layer would add latency without adding authority.
First intervention: convert the council into a small decision forum of 8-10 accountable executives, move the 22 stewards into a working group that prepares options, publish a decision log with named decision owners, and set a 10-business-day decision SLA. Retire or assign owners to the 19 unenforced policies rather than writing new ones. Measure cycle time and reversal rate, not policy count.
2. Finance and Marketing disagree on 'active policyholder.' Design the decision rights and resolution path. What determines whether one definition can serve both?
Read the worked answer
Separate accountability from consultation. Name one accountable owner for the enterprise term based on the decision it drives: for statutory and financial reporting, Finance/Actuarial is accountable; Marketing and Customer Analytics are consulted; the CDO is informed and owns the process. The owner must publish a definition with explicit inclusion and exclusion rules, an effective date, and quality thresholds. If Marketing's need is genuinely different (campaign-eligible versus in-force), do not force one definition: create a governed enterprise definition plus labeled variants, with each variant linked to the conformed term in the glossary.
Resolution path: quantify the 6% gap in records and revenue at stake, test both definitions against three downstream use cases, decide within the SLA, and record any exception with a review date. The third shadow definition is retired or registered as a variant; unregistered copies are treated as defects.
3. The CDO proposes a second council layer to handle volume. Assess this proposal.
Read the worked answer
A second layer is likely to worsen the core problem. The council already has low attendance and slow decisions; adding a layer inserts another queue and makes accountability ambiguous. Only add a tier if it has a distinct scope, its own accountable chair, and a clear delegation rule, for example domain councils decide within standards while the enterprise council decides cross-domain conflicts.
Given one quarter and no headcount, run a 30-day pilot on the active-policyholder decision with the redesigned forum. If throughput improves, consider domain-level forums later. Success criteria: decision cycle time under 10 days, zero reversals without new evidence, and at least 80% of policies with named operational owners.