Microsoft PL-200, the Power Platform Functional Consultant Associate exam, retired on August 31, 2026. A useful study guide must now explain the practical consulting skills that continue to matter, rather than imply that new candidates can still schedule PL-200. The work includes requirements discovery, Dataverse modeling, apps and automation, security, solution deployment and user acceptance.
Treat the retired exam as historical context
Microsoft’s exam retirement register records PL-200’s retirement. Legacy objective lists can still help organize work in Power Apps, Dataverse and Power Automate, but they should not be described as the current booking blueprint. No single new exam should be called its automatic replacement without issuer evidence.
Some learners should begin with Microsoft PL-900 to learn the platform’s scope, connectors, environments and business value. Others with real implementation experience should compare current role-based credentials against the tasks of their job, not against a simple exam-code progression.
Gather requirements before selecting a component
Interview a business user about the decision they need to make, the data they trust, the approval they cannot bypass and what a failure looks like. Translate that discussion into acceptance criteria and a traceable process. A consultant needs to challenge a requested feature when a simpler workflow or clearer ownership model would solve the underlying problem.
Separate the user interface from the record model and the process. One screen can create a case, but who owns it, who can reassign it and how exceptions are reported are separate design questions.
Model Dataverse and secure access
Design tables from business entities and relationships, not from the order in which screens are built. State which fields are required, how reference data is validated and how updates are audited. Avoid adding a complex workflow that only exists to compensate for a poorly understood schema.
Agents and connectors must honor the same data boundaries as users. When evaluating AI-enabled business processes, Dataverse grounding and permissions demonstrates why a convenient language interface must not grant new record access. Test a negative permission case before demonstrating the happy path.
Automate only processes with defined exceptions
Power Automate can coordinate approvals, notifications and integrations. For each flow record the trigger, retry behavior, required identity, idempotency conditions and the person who handles a failed action. If a flow updates a case twice after a retry, a successful run history may still conceal corrupted business state.
Use a scenario with an invalid record and a temporary connector failure. Test who receives an alert and how the data is corrected. The consulting skill is choosing a reliable process, not merely attaching the maximum number of cloud services.
Work through a Power Platform requirement from intake to signoff
An HR team requests an application to collect onboarding equipment approvals. Clarify whether a record belongs to the employee, hiring manager or IT queue, which fields are confidential and who may edit the decision after approval. Model the tables, relationships and business rules before choosing a canvas or model-driven app. Show what happens if a manager delegates authority or a new employee leaves before equipment is issued.
A cloud flow should handle failed notifications and duplicate approvals without leaving a misleading ‘complete’ status. Define test users with different roles and verify field and row permissions at both the app and Dataverse layers. Store the solution in a managed deployment pipeline, record dependencies and obtain user acceptance evidence. These design skills are transferable even though the old PL-200 exam is retired.
Use ALM and user validation to finish the work
Separate development and production environments, document solution dependencies and test a rollback or recovery. Use controlled deployment, representative accounts and explicit acceptance criteria. A configuration can be technically deployable yet fail because a salesperson cannot find the required field or a manager cannot approve an exception.
A durable portfolio contains the original requirements, model diagram, role permissions, one automated process, two negative test cases and evidence of a user-requested improvement. Those artifacts matter even after an examination retires.