Abstract
Current U.S. postings for data quality, data governance and data stewardship work show that trustworthy data is made through recurring operational decisions. Analysts and stewards profile data, define and maintain rules, watch scorecards, document meaning and lineage, coordinate issue resolution, and make ownership and escalation visible. More senior roles establish decision rights, set standards, arbitrate definitions and lead cross-functional remediation. A catalog or monitoring platform may support this work, but the posting evidence does not make any one vendor product a universal requirement.
The study uses a structured purposive sample of 100 directly accessible U.S. employer and employer-hosted application pages from 90 employers, reviewed on 29 September 2026. It is a point-in-time picture of advertised work, not a representative estimate of all U.S. vacancies or a forecast. The aggregate findings below must be read alongside the role levels and specialized contexts in the sample. Requirements absent from a posting are unobserved, rather than negative evidence.
Scope and method
The research geography was declared as the United States before collection. Eligible records had a U.S. job location, an identifiable employer and role, a substantive live job description, and an application or interest control visible on an employer or employer-hosted applicant-tracking page. The final sample draws from seven direct public source families: Greenhouse (45 postings), employer career sites (24), GovernmentJobs employer listings (13), Lever (6), SmartRecruiters (6), Amazon Jobs (4) and iCIMS (2). No Indeed or Workday record was counted. The sample deliberately sought data quality analyst and manager, data governance analyst and manager, data steward, master data, catalog and related operational positions. It contains 69 core operating roles, 3 specialized core roles, 22 adjacent specialist roles and 6 adjacent context roles. Narrower engineering, architecture, public-sector and regulated-sector roles inform boundaries or progression where they explicitly describe quality or governance tasks; they do not define the core role by themselves.
Each unique posting was coded from the employer's own description for responsibilities; named work products; hard methods and skills; soft skills expressed as observable behavior; tools; stated required and preferred qualifications; role level; work cadence; interfaces; and decision or escalation boundaries. A code required a short supporting span tied to the posting URL. A missing code was treated as not observed in the retained source, not as proof that the employer did not require it. Employer names, requisitions and near-identical descriptions were reviewed for duplicates. Search results, inaccessible or expired pages, and other unverified snippets were excluded from the accepted sample.
This is a purposive, cross-role sample rather than a probability sample. Its purpose is to identify recurring work patterns and their variation, not estimate the share of all U.S. employers using a tool or method. A posted responsibility also does not prove how consistently an employer performs it after hiring. Specialized sector postings can show a boundary or progression task, but their regulatory detail is not generalized to all data teams.
Explicit signals in the sampled postings
The table reports conservative minimum counts of explicit coded mentions in the 100 reviewed postings. Several postings mention more than one category, so the rows cannot be added. A posting without a retained mention is unknown for that category. These are not estimates of national prevalence.
| Coded requirement | Postings with an explicit retained signal, minimum of 100 |
|---|---|
| Governance operating model | 50 |
| Data ownership and stewardship | 42 |
| Master and reference data | 43 |
| Issue intake, root cause and remediation | 41 |
| Metadata, glossary and catalog | 39 |
| Data quality rules and controls | 39 |
| Lineage and traceability | 29 |
| Quality and governance measurement | 20 |
| Profiling, audit and reconciliation | 16 |
| Classification, access and privacy | 14 |
The lower count for profiling, for example, means only that fewer retained posting spans explicitly used that signal. It cannot establish that profiling is rare in real data-quality work. The examples below show what the work looks like where postings describe it in enough detail.
Findings: a quality rule is an operated control
The postings describe quality rules as more than a list of desired attributes. An iCapital master-data analyst is asked to help create and maintain business rules, monitor quality across systems and coordinate resolution of gaps with stakeholders. A Wellmark data-governance analyst profiles warehouse data, reviews source-to-target mappings, produces data-quality scorecards and supports triage and impact analysis. A Caterpillar data steward specialist builds statistical scorecards for product attributes, profiles with SQL and Python, and uses automated rules to prioritize exceptions. These examples span different sectors and levels, but they all connect the rule to a tested dataset and a response when the data fails. iCapital posting; Wellmark posting; Caterpillar posting.
The question for an operational team is therefore specific: which records and fields are in scope, which condition constitutes failure, who reviews the result, and what evidence closes the issue? A score is less useful when the tested population, threshold, owner or follow-up action is unclear. The postings support this workflow interpretation; they do not establish a single universal rule formula or threshold.
Findings: ownership is a decision path
Ownership appears in the practical handling of data assets and disputes. Justworks' senior governance manager role owns the catalog, certification standards, ongoing quality observability and review of new assets; its posting explicitly gives governance the right to withhold or revoke certification. Impact.com's manager/strategist role develops an enterprise framework defining ownership, stewardship and accountability, facilitates governance council meetings, and aligns business leaders with Salesforce administrators and architects. Phoenix Contact's regional master-data role is a primary escalation point for complex cross-system issues and validates or approves master-data requirements for changes. Justworks posting; impact.com posting; Phoenix Contact posting.
These examples should not be flattened into one job description. An analyst may identify and recommend; a steward may maintain a defined data domain; a manager may enforce certification or resolve competing definitions. The evidence supports teaching the distinction between proposing a rule, approving a decision, implementing it and verifying its effect. The local employer's policy, not a generic course template, determines the actual authority.
Findings: metadata and lineage serve investigations and change
The postings connect metadata to work that a colleague must perform. Wellmark describes updating business names, definitions and valid values, and keeping technical source-to-target mappings current. Justworks ties catalog hygiene, glossary terms and metrics-catalog freshness to asset certification. Veeam's director role links a technical catalog with a business-facing catalog and asks for domain-by-domain certification, lineage, continuous scoring and a named owner. Wellmark posting; Justworks posting; Veeam posting.
The operational test is whether someone can use the metadata to answer a real question: what does this field mean, where did it come from, which downstream product uses it, and whom should a proposed change reach? Vendor-specific catalog experience appears in individual roles, but the transferable requirement is to maintain usable definitions and traceable relationships, then identify gaps rather than assume automatic capture is complete.
Findings: issue management has a cadence and a handoff
Several postings make quality work a repeatable rhythm. iCapital asks an analyst to follow up with teams so data requirements are completed accurately and on time. Harbor's data-quality professional reviews and corrects client records, resolves duplicates, examines weekly audit feedback and proactively communicates blockers. Caterpillar tracks product-data accuracy, completeness and consistency over time. Phoenix Contact's senior role monitors metrics, identifies risks and drives corrective and preventive action with business and IT stakeholders. iCapital posting; Harbor posting; Caterpillar posting; Phoenix Contact posting.
This is a material distinction between finding an anomaly and resolving it. The work includes an intake or observation, evidence about affected records, a likely cause, an accountable decision, remediation, retesting and communication to the data consumer. The exact daily or weekly rhythm varies by role and posting; the study does not imply that every employer uses the same service-level target or ticketing platform.
Levels and boundaries
The same field contains entry, individual contributor, specialist and leadership work. An Avride analyst validates annotated autonomous-vehicle data, reviews recurring quality metrics and works with production and engineering. An iCapital analyst follows up on governance requirements and coordinates resolution. Justworks' senior manager leads a team and may withhold certification. Phoenix Contact's manager title refers to a senior individual contributor with escalation responsibility rather than direct people management. Avride posting; iCapital posting; Justworks posting; Phoenix Contact posting.
The practical capability common to these levels is to turn an unclear data concern into a reviewable work item: define the affected asset and business use, collect evidence, distinguish a proposed fix from an approved change, record the owner, and verify the result. Specialized postings add narrower obligations. Formation Bio, for example, assigns R&D data-integrity ownership and AI risk-assessment approval in a regulated pharmaceutical setting. That is evidence of a sector-specific senior boundary, not a general legal duty for every data-governance practitioner. Formation Bio posting.
Geographic coherence and independent context
Every vacancy counted in this study is tied to a U.S. job location. U.S. occupational and public-sector sources provide context only: O*NET's Data Warehousing Specialists profile identifies data quality, source mapping and metadata tasks in an adjacent occupation, while a 2026 U.S. GAO report on federal eligibility data documents a concrete data-quality and interoperability problem in one government setting. Neither source is counted as a vacancy, and neither is used to estimate the frequency of an advertised requirement. O*NET occupational profile; GAO report.
MTF Institute's separate current-changes review examines dated product and public-sector releases from the latest 90 days. It helps explain why connected metadata, incident visibility and explicit access decisions are timely topics. Its source corpus is independent of this vacancy sample; vendor releases are not counted as employer requirements, and the vacancy sample is not used to prove product adoption.
Limitations and interpretation
The sample is structured and purposive, with deliberate coverage of related occupational labels and employer types. It is not statistically representative of all U.S. vacancies. The retrieval date is a snapshot: a page may close, change or disappear later. A live application control is a source-assurance signal at retrieval time, not proof of a future opening or hire. Employer-written descriptions vary in specificity. Some omit cadence, tools, authority or preferred qualifications even when those matters exist in practice. Absence of a code cannot be interpreted as employer rejection of the skill.
The sample also spans experience levels and sectors. Banking, healthcare, public-sector and defense postings may have local constraints that do not apply elsewhere. Vendor names are recorded only when stated; this research does not endorse products, require paid software, reproduce proprietary frameworks or offer legal or regulated-sector advice. The evidence supports a practical U.S. professional learning focus on measurable quality control, usable metadata and lineage, issue resolution, and an explicit governance decision path. It does not support an employment guarantee or a claim that one certificate is required by employers.
Conclusion
The U.S. vacancy evidence points to data quality and governance as connected operating work. The recurring task is to define what trusted data means for a specific use, test it, make its origin and owner findable, investigate failures, and document decisions until the issue is resolved. A useful practitioner can work across business and technical teams while recognizing where approval belongs to a steward, manager, security function or local policy owner.
Continue learning
Apply the evidence from this report through MTF Institute's Professional Certificate in Data Quality & Data Governance. The programme turns the identified capabilities into structured theory, guided AI practice and reusable workplace artifacts.