In 2022, a $35 billion-asset Midwest bank launched a data stewardship program with great fanfare. Forty-five stewards were appointed across business lines. Training was delivered. A stewardship council was formed. Eighteen months later, 31 of the 45 stewards had either left the role or stopped participating. The council had not met in six months. The program was functionally dead, and the bank's CDO was back at square one explaining to the board why governance was still "maturing."
This is not an unusual story. It is the typical story. Data stewardship programs fail at a rate that would be unacceptable in any other operational function. They fail because they treat stewardship as a volunteer activity layered on top of someone's real job, rather than as a governed operational role with real authority, real accountability, and real consequences for inaction.
The Volunteer Trap
Here is how most stewardship programs are structured: a business analyst in commercial lending is nominated as a data steward. They attend a two-day training session. They receive a stewardship charter that describes their responsibilities. Then they go back to their actual job, which already consumes 110 percent of their time.
The stewardship responsibilities get done when there is time, which means they get done when there is pressure, which means they get done right before an audit or an exam. The rest of the time, the stewardship role exists in name only. The steward does not have authority over data decisions. They do not have time for data governance tasks. They do not have performance objectives tied to stewardship outcomes. They are volunteers with a title.
This is not a people problem. It is a design problem. The program was set up to fail by and in the structure itself.
What Sticking Looks Like
The few stewardship programs that survive share three characteristics that failing programs lack:
Authority that matches accountability. Stewards in successful programs can actually enforce data standards. If a business line is entering data that does not meet quality thresholds, the steward can escalate and the escalation has teeth. Stewards without authority are advisors. Stewards with authority are governors. The difference is not semantic. It is operational.
Governance as a workflow, not a meeting. Programs that survive embed stewardship into daily operational workflows. When a data quality issue is detected, the steward receives an actionable task, not an agenda item for next month's council. The work is pushed to them. They do not have to remember to pull it.
Consequences for inaction. This is the uncomfortable one. Stewardship roles in successful programs are tied to performance evaluation. If a steward consistently fails to act on quality issues, there is a consequence. Not punitive. But real. The same way a compliance officer who fails to file a SAR faces consequences. Stewardship is an operational role. Operational roles need operational accountability.
The Structural Inhibitor in Tier 2 Banks
Tier 2 banks face a specific inhibitor here: headcount pressure. You cannot create a dedicated data stewardship layer when you are already running lean. Every governance role is a role taken from somewhere else.
The answer is not to add headcount. It is to reduce the manual burden of stewardship. Most of what stewards spend time doing is manual: chasing data quality issues across spreadsheets, assembling attestation evidence, tracking down lineage gaps. If those tasks were automated, the steward's role shifts from administrator to decision-maker. They spend their limited capacity on the governance decisions that only humans can make, not on the clerical work that consumes 80 percent of their stewardship time.
The CoComply Angle
CoComply automates the clerical burden of stewardship. Data quality issues surface automatically. Attestation evidence assembles itself. Lineage gaps are flagged without manual tracing. Stewards spend their capacity on decisions and escalation, not on assembling proof packs. The program sticks because the role is sustainable.
Run This Diagnostic
List your current data stewards. For each one, answer: (1) Do they have authority to enforce data standards in their domain? (2) Is stewardship part of their performance evaluation? (3) Could they tell you, right now, the top three quality issues in their domain without checking a dashboard first? If more than half your stewards fail this test, your program is already in decline. The question is whether you fix the design or wait for the next exam to reveal it.