Table of contents
At a glance
How do you validate Jira Service Management for its intended use?
Define the regulated operational decisions and records that Jira Service Management supports. Test the service and asset relationship, authorised business approvers, effective workflow rules and link to actual implementation. Include failure cases, traceable records and verification of the result. Assess Cloud, Data Center, apps and licence scope separately. An approval or CMDB status alone does not demonstrate a suitable operational state.
- Change release, technical implementation, and functional control are distinct forms of validation evidence.
- The correct authorizations matter; having a minimum number alone is insufficient.
- Assets provide context, whose assignment and currency must be verified.
- Cloud and data center guidelines, as well as integrations, remain specifically to be clearly defined.
1. Define the affected operational decision
Jira Service Management supports service requests, incidents, problems, and changes. Its GxP relevance arises from its actual usage. A ticket for a defective office mouse carries a different implication than the release of a change to a laboratory information system. Describe which decision is formally made or documented within the tool and which technical processes depend on it.
This article examines the journey from a reviewed operational change through to the executed action and its verified outcome. The use of general Jira workflows as test reports represents a distinct use case. A properly closed test ticket does not prove that the exact package was successfully deployed to the target system. Conversely, a successful deployment does not replace the absence of a technical assessment.
For supported human medicinal product GMP processes, regulated changes, incidents, and reliable records are particularly relevant. Annex 11 is a GMP guidance document, not a product certification. General ITSM usage does not automatically require the same level of evidence. Clearly define and trace critical services and dependencies. Neither a standard template nor an ITIL term determines whether a specific activity impacts quality or data integrity.
2. Evaluate cloud and data center separately
For Cloud, Atlassian describes, among other things, approval steps in company-managed service areas. For Data Center, the article is based on the documentation for version 11.3; individual pages have been updated earlier. For your project, use the output that matches your installed stand. Do not uncheck settings between cloud, data center and different project types.
Capture operating mode, project type, installed apps, workflow status, and actual license authorizations. The cloud deployment gating guide specifies Premium and other prerequisites for the connected tool. General function overviews and specific service documentation describe varying scopes. A blanket assurance that every asset or service function is included in every tier would not be defensible.
For existing Data Center environments, the product lifecycle must be incorporated into planning. Atlassian designates March 28, 2029, as the end-of-life date for affected products, including Jira Service Management Data Center; certain exceptions are separately provisioned. Review your contract status and evaluate potential migration options early. Successor operations represent only part of the effort: approval pathways, apps, permissions, and technical integrations must continue to align with intended use in the target environment.
3. Clearly link the approved and decision subject
Atlassian defines for cloud-configurable approval sources, required number, and target status for approval or rejection. The exclusion of individuals from specific fields is also planned. However, this does not yet result in a professionally aligned approval matrix. Two approvals may originate from two IT personnel, even though the process requires an IT and a QA decision.
Therefore, clearly define who evaluates which content. Should risk assessment, test scope, target service, package identifier, and rollback plan be part of the decision-making process? Then, at the time of decision, these information elements must be unambiguously assigned. If the target system or package is later modified, it must be assessed whether existing approval still applies to the same subject matter. Automatic re-evaluation or reset is not provided as a standard function.
In addition, verify who is authorized to modify approval fields, how delegation is governed, and whether an applicant can improperly self-assign. The configured minimum number of authorized parties alone does not adequately address these questions. Separate decision steps may be appropriate for your required workflow; the actual effectiveness of these steps must be demonstrated. Furthermore, an approval action is not automatically equivalent to an electronic signature possessing all required attributes within the specific regulatory domain.
4. Checking the transition to actual delivery
The data centre documentation identifies a specific workaround: if an approval status has more than the two predefined outgoing transitions specified for approval and rejection, agents may choose alternative transitions. In such cases, the approval is not enforced through the intended pathway. Therefore, the mere visibility of an approval step alone does not constitute sufficient evidence of an effective lock.
For Cloud, Atlassian defines Deployment Gating as a coupling between Jira Service Management and an established CI/CD tool. The guide covers Bitbucket and Jenkins. Proper setup on both sides, service integration, and appropriate environment configuration are essential. Approval of a ticket does not, by itself, prevent any manual modification, alternative pipeline, or direct administrative access.
Investigate the actually authorized implementation pathways. A normal case may proceed in the intended target after full decision-making; a rejected or pending case must not be authorized to follow this controlled pathway. Supplement a deliberately incorrect service linkage and an unbound pathway, if such a scenario is feasible during operation. The outcome may be a technical correction or a justified procedural measure. The key factor is the demonstrable protection of the specific process.
5. Leverage assets and services as a maintained decision-making context
Assets can link configuration elements to processes. In the Data Center, Atlassian further specifies how user- or group attributes can be assigned to selected objects to provide approvers. This ensures the quality of object maintenance becomes a prerequisite for decision-making: an outdated owner or incorrect object may direct a process to the wrong individual.
The object should unambiguously identify the actually affected system. Useful elements within the own procedure include a stable identifier, environment, responsible party, and relevant relationships. A version stored on the object is initially a statement of fact. Whether it corresponds to the installed state must be substantiated through appropriate observation or a reliable data source. Similarly, a CMDB status such as "stopped" alone does not prove that the system has been actually decommissioned.
The following work aid combines an example change to a LIMS with specific validation evidence. This is not a mandatory field list prescribed by Atlassian. Use existing data where its origin and timeliness are sufficient. Particularly verify how a wrong or missing asset is identified before any professional approval or technical action is taken.
| Link in the chain | Fixed subject | Required independent verification |
|---|---|---|
| Service | LIMS-17 Production | Assignment to the employed laboratory process |
| Asset | APP-17, Production Environment | Identity and observed starting material match |
| Change | Package L17-R3 supersedes L17-R2 | Verified content and package identification clearly defined |
| Decision | QA Role and IT Role | Both responsible roles have decision authority over the same batch. |
| Implementation | Linked pipeline for APP-17 | Actual objective, package, and result traceable |
| Closure | Defined registration and result inquiry | Functionality check upon implementation; deviations assessed |
6. Example: Intended Model – two roles require more than two clicks
The example is not an executed Jira test. For CHG-17, Person A is responsible for IT and Person B for QA. The applicant must not perform their own release. Each required valid decision from both roles is required for package L17-R3 for APP-17. A second IT personnel does not substitute for Person B. The authorized technical pathway is explicitly restricted to this target environment.
The role coverage is calculated as a quantity: From {IT}, coverage is also achieved through two IT approvals, yet still lacks {IT, QA}. Thus, zero, one, or two of the two required roles remain uncovered. Full role coverage permits release only additionally when the package, destination, and valid technical coupling are correct. Following execution, the defined functional check must be performed; "approval" does not constitute evidence of a successful operational state.
In all cases, the approval status, package identifier, and target reference will be reset to the described initial state. The table includes six separate required scenarios. As an additional error verification, a further workflow transition is examined within a controlled data center test configuration. A failed pipeline run must also remain visibly flagged as a failed execution, even if both roles have given approval. These additional scenarios evaluate other boundaries and are not resolved by counting approvals.
| Situation and Background | Roles covered | Expected result for controlled delivery |
|---|---|---|
| No approval; package and target correctly specified | 0 of 2 | Not for release |
| Only Person A consents; package and destination are correct | 1 of 2: IT | Not for release |
| Two IT personnel agree; QA is missing | 1 of 2: IT | Not for release |
| Persons A and B agree; all other conditions are met | 2 of 2: IT, QA | Release permitted; final release inspection remains open |
| Persons A and B agree; another asset selected | 2 of 2, incorrect item | Not to be used as a release for APP-17 |
| Persons A and B agree; the package is subsequently replaced by L17-R4 | 2 of 2 from the previous status | Reassess fit; no unaudited takeover |
7. Restoration, Root Cause Analysis, and Separation of CAPA
Atlassian defines Incident Management as a response to unplanned events or service outages and the restoration of service. Problem Management addresses root causes or potential root causes of such events. This distinction is valuable: resolving a single access outage does not confirm why it recurred or whether data was affected.
Within our GMP process, the technical impact is additionally assessed. If a laboratory service is once again accessible, it must be determined, for example, whether activities initiated during the disruption were fully and correctly continued. Where necessary, link incidents, problems, changes, and quality actions such that an open investigation is not obscured by closing a service ticket.
Annex 11 requires reporting and evaluation of incidents; for critical incidents, it addresses root causes and derived corrective and preventive actions. This does not automatically trigger a CAPA for every IT ticket. Similarly, creating a process type "CAPA" does not constitute a complete quality process. Responsibility, action, effectiveness evaluation, and closure must be clearly defined in accordance with the specific incident. For urgent interventions, a controlled exception pathway is required, ensuring that subsequent evaluation is traceable across all involved systems.
8. Maintain traceability of decisions and configuration changes
Atlassian displays data center approval information, including in Activity and Past Approvals. Review the relevant structure internally to determine whether the required relationships are clearly evident: Who assessed which status for which objective, and when? A historical status alone may be insufficient if equipment, object references, or approvers have been modified since.
For GMP-related records, consideration must be given to protection, accessibility, and controlled amendments. Determine which information is actually retained and how processes, equipment, and related documentation can be effectively reconciled and reviewed at a later stage. A universal retention period or falsification prevention cannot be inferred from the product specification. An export is only useful when its content enables clear and traceable support for the required decision.
The tool's configuration also impacts the process: a new automation rule, a different approver filter, or an additional transition step can alter previously effective controls. Evaluate such changes and specifically review the affected normal and error cases. Existing appropriate documentation may be utilized for this purpose. Neither a vendor package nor an app update justifies unverified claims that all aspects are covered or that complete revalidation is required.
9. Specifically track cloud modifications and service migrations
The current cloud documentation describes a transition from a services structure to an assets schema: Software Registry for software services and Business Portfolio for business services. Certain references are carried over. However, custom fields, automation rules, dashboards, or external scripts may require adjustments. Atlassian explicitly states that external script references will not be automatically tracked.
Therefore, for your change process, it must be verified whether the relevant service still specifies the same objective and the correct responsible parties. Use a well-known normal case and an intentionally outdated reference. Monitor approval determination, technical coupling, and closure substantiation. A completed relocation in the administration dialog does not, by itself, demonstrate the suitability of all existing integrations.
This source describes a specific migration scenario and does not establish a general rule applicable to every cloud update. During ongoing operations, associate relevant vendor changes with the actual functions currently in use. When transitioning from a data center to the cloud, additional configuration and application boundaries are introduced. Clearly justify which previous documentation remains valid, which data relationships require verification, and how open findings are to be managed prior to intended use.
10. Initiate introduction and inventory verification with the correct personnel
For a small and medium-sized enterprise (SME), clear decisions are often more valuable than numerous additional fields. Prior to launch, a representative and relevant change, an incident, and the associated services should be well understood. The service owner defines the required operational decision, QA provides the technical impact assessment, JSM administration outlines the workflow, and the operations team details the actual implementation pathway.
Effort arises from diverse project types, applications, automation rules, incomplete asset data, and multiple distribution channels. A complex exception process may require more clarification than many similar tickets. In the existing environment, it is first necessary to verify which rules actually apply; during implementation, the intended state is defined prior to configuration. Without access to appropriate testing capabilities, technical claims of impact remain unsubstantiated.
The preparation checklist focuses on verifiable relationships. Utilize existing descriptions, provided they are current and relevant. A closing assessment should indicate which functions are adequately supported, which limitations apply, and who is responsible for remaining actions.
- Document the cloud or data center, project type, version or configuration status, and applications in use.
- Select the relevant service with the correct asset, environment, and responsible parties.
- Describe precisely what decision is being made and which roles must approve it.
- List all possible implementation pathways including exceptions.
- Prepare initial state, reset procedures, and anticipated error responses.
- Combine records for evaluation, implementation, and functional control.
- Involves service responsibility, QA, administration, and operating personnel in a binding capacity.
Project experience and collaboration
Trace approval through to the operational change actually implemented
With service owners, QA and IT, we follow a prepared change from the correct asset to the verified implementation outcome. Who approves what, and what prevents a prohibited transition? Existing workflows may allow targeted tests. A gap between approval and implementation needs an additional link or control. We also examine rejection, errors and historical retrieval. An approved ticket alone does not demonstrate the correct operating state.
Frequently asked questions about Jira Service Management
When is Jira Service Management relevant for validation?
When a tool supports significant decisions or records for a regulated process, its intended use must be assessed accordingly. Examples include the evaluation of a laboratory service change or the management of a critical deviation. General service requests are not automatically considered critical. First, describe the function, impact, and dependencies, rather than responding solely to the product name.
What distinguishes this usage from regular Jira test tickets?
A test ticket can document a testing result. In contrast, Service Management integrates relevant service, change decision, technical implementation, and operational restoration activities. These records can be linked together but do not replace one another. For a change, it must remain evident that the approved package was actually implemented in the intended target environment and that the required functionality was subsequently verified.
Do two Approvals suffice as evidence of IT and QA approval?
Not automated. Two individuals with the same professional role may be assigned. Configure and verify that the actually required roles are correctly determining the appropriate object. Clarify representation, self-approval, and changes to the list of approvers. The number or selection described by Atlassian represents a technical component, not a pre-defined professional approval matrix.
Can a approval be bypassed in the workflow?
The data centre documentation explicitly addresses additional outgoing transitions alongside Approve and Decline. Therefore, verify all relevant transitions and authorized user roles. This specific statement is not automatically applied to every cloud configuration. In cloud environments, the responsibility remains to evaluate the actually implemented pathway—including automations—against the established requirements.
Does an unapproved change automatically prevent every deployment?
No. The cloud guidance requires established deployment gating in JSM and the associated CI/CD tool; it provides specific prerequisites for Bitbucket and Jenkins. Other pipelines or manual administration pathways are not automatically monitored. Document the actual implementation approaches and assess where decisions are actually made and which additional controls are necessary.
Sources: [4]
How can incorrect assets impact an approval process?
A misconfigured object may point to a different responsible party, service, or target state. In a Data Center, configured user and group attributes derived from assets can feed into authorization evaluation. Therefore, both object assignments and their maintained content must be reviewed. The mere presence of a required field does not constitute evidence that the correct production service has been selected.
Is a CAPA automatically generated from a problem ticket?
No. Incidents, problems, and quality actions serve distinct purposes. A problem can support root cause analysis; in a specific GMP context, the impacts, required actions, and effectiveness of those actions must be appropriately addressed. Assessment of critical incidents is particularly relevant. The name of a process type or its status does not replace professional justification or a suitable closure.
Are approvals automatically electronic signatures?
An approval in the workflow is initially the documented product action. Whether it constitutes the required electronic signature for your process depends on specific requirements and actual implementation. Verify identity, subject of decision, and traceability of the action. The use of JSM does not imply a universal signature obligation for all tickets nor a guarantee that every approval configuration meets these requirements.
What do we need prior to startup and who must participate?
Provide relevant services, assets, workflows, authorized roles, and implementation pathways. Service owners, QA, JSM administration, and technical operations must jointly be able to demonstrate the intended workflow. Effort depends particularly on applications, automations, data quality, and exceptions. Missing test access or unclear responsibilities represent concrete planning risks; generic cost and time estimates are of limited value here.
What is important regarding cloud changes or data center migrations?
Assess the alignment of your functions and integrations in the target state. For example, documented service migration demonstrates that existing fields or external script references may require follow-up work. For Data Center, the announced product lifecycle must also be considered. Leverage appropriate existing evidence with justification and supplement the affected test cases, rather than automatically assuming full equivalence or a complete rebuild.
Sources and review date
Source date: 6 October 2026. Cloud documentation without a fixed build; Data Center functions from the documented section 11.3. Project type, applications, license rights, and configuration are to be separately verified. Tables and decision matrix are standalone reference tools, not completed product testing.
- Atlassian: Main Jira Service Management features
- Atlassian: Add an approval step to a workflow
- Atlassian: Set up approval steps
- Atlassian: Set up deployment gating
- Atlassian: Migration of your services
- Atlassian: Setting up approvals
- Atlassian: Adding approvers from Assets to requests in Jira
- Atlassian: Managing changes with your IT service desk
- Atlassian: Data Center End of Life
- European Commission: EudraLex Volume 4
- European Commission: EU GMP Annex 11 Computerised Systems
- European Commission: EU GMP Chapter 4 Documentation
Sources checked: .
Combine operational decisions and supporting evidence
We support you in identifying relevant JSM processes, linking approvals, system context, and actual implementation in a traceable manner.
- Cloud and SaaS validation →
Maintain the validated state of cloud and SaaS systems as suppliers issue new versions. We define the boundary between your responsibilities and the supplier’s and agree how each release will be assessed.

