The PL-900 blueprint is split across 5 domains. Weights are the official share of the exam; see the official exam guide for the authoritative breakdown.
What you must be able to do. Given a business problem stated in plain terms, identify which Power Platform service solves it, how the platform connects to Microsoft 365, Dynamics 365, Azure, and third-party apps, and which governance control applies.
In one sentenceThe why-and-which layer: matching a business need to the right Power Platform service and knowing how the suite is governed and connected.
Recall check: answer these from memory first
- A team re-keys data between two systems daily and keeps forgetting to notify the customer. Name the service built to fix this and say why it is not Power Apps.
- A council wants residents to self-serve against the same permit records staff already manage in Dataverse. Which service builds that external site, and why is no second copy of the data created?
- Inside Microsoft Teams, which built-in component lets an app, a flow, and an agent all read and write one shared set of records?
What it tests. Whether you can read a business problem and name the service that solves it without writing code. It covers the business value of each service (Power Automate for workflow automation, Power Apps for custom applications, Power Pages for external sites, Dataverse as the shared store, and Dataverse for Teams as the in-Teams low-code platform); how Power Platform connects to Microsoft 365, Dynamics 365, Azure, and third-party applications through connectors; and platform administration and security, including environments, data policies, and tenant isolation that govern what makers can build and where data can flow.
How to study it. Learn each service by the one-line job it does and the verb that signals it: automate a process means Power Automate, build a tailored app means Power Apps, publish an external site on existing records means Power Pages, store the shared data means Dataverse. Drill the scenario form by covering the options and naming the service from the need alone. For governance, fix three facts: data policies and tenant isolation apply to Teams and Dataverse for Teams the same as other environments, environments are how administration and data residency are scoped, and Dataverse for Teams is the single shared platform that Power Apps, Power Automate, and Copilot Studio all build against inside Teams.
Easy to confuse
- Power Automate versus Power Apps. Power Automate runs an automated workflow that moves data and sends notifications between apps and services; Power Apps builds a custom interface for people to enter and view data. If the verb is automate a process, it is Power Automate; if the verb is build an app to use, it is Power Apps.
- Power Pages versus Power Apps. Power Pages creates secure external-facing websites for customers and partners, reusing the same Dataverse records; Power Apps builds internal business applications for staff. The deciding signal is who the users are - external public visitors point to Power Pages, internal staff point to Power Apps.
- Dataverse versus Dataverse for Teams. Dataverse is the full low-code data platform used across the whole tenant; Dataverse for Teams is the built-in, lighter version scoped inside a Teams environment that Power Apps, Power Automate, and Copilot Studio build against. Choose Dataverse for Teams when the scenario keeps staff working inside Teams.
Worked example from the PL-900 bank
lock_openFree sampleDescribe the Business Value of Microsoft Power Platformeasy
A logistics company manually re-keys delivery confirmations between two systems every day, and staff often forget to send the customer notification. Which Power Platform service is designed to optimise this business process by running an automated workflow between the apps and services involved?
- APower Automate, which creates automated workflows between apps and services to synchronise files, collect data, and send notifications.check_circle Correct
- BPower Apps, which builds custom business applications that let users enter and view organisational data through a tailored interface.
- CPower Pages, which creates and hosts secure external-facing websites that customers and partners use across browsers and devices.
- DDataverse, which securely stores and manages the data used by business applications across the Power Platform.
Power Automate is the service for automating repetitive processes and workflows that move data and send notifications between apps and services. Power Automate optimises business processes and automates repetitive tasks by creating automated workflows between apps and services to synchronise files, get notifications, and collect data, which is exactly the re-keying and notification work described.
Why A is correct: Correct. Power Automate optimises business processes and automates repetitive tasks by creating automated workflows between apps and services to synchronise files, get notifications, and collect data, which is exactly the re-keying and notification work described.
Why B is wrong: Tempting because an app could capture the confirmations, but the grounding scopes Power Apps to building applications, not to running automated workflows between systems on a schedule.
Why C is wrong: Power Pages is for external-facing business websites per the grounding, not for automating an internal re-keying and notification process between back-office systems.
Why D is wrong: Dataverse is the data store per the grounding; it holds the data but does not itself run the workflow that moves and notifies between the two systems.
What you must be able to do. Given a data, integration, or governance need, identify the core Power Platform component that fits, from Dataverse tables to standard, premium, and custom connectors to AI Builder, environment roles, and the platform's security model.
In one sentenceThe building-blocks layer: Dataverse structure, the three kinds of connector, AI and Copilot capabilities, environment roles, and how the platform secures and locates data.
Recall check: answer these from memory first
- In Dataverse, what is the structure made of rows and columns where each column stores one type of data, and how does it differ from a solution and a business rule?
- Which environment role lets a citizen developer create apps, connections, custom connectors, and flows but grants no administrative control over the environment?
- Which mechanism encrypts Power Platform data in transit and which encrypts it at rest, and why is it wrong to say TLS provides encryption at rest?
What it tests. The core components that everything else is built on. It covers Dataverse as the underlying data platform (tables made of rows and columns, each column holding one data type, plus business rules and solutions); connectors that integrate data sources, split into standard, premium, and custom; AI Builder and Copilot capabilities for adding intelligence without code; environment roles such as Environment Maker and Environment Admin that scope who can build versus govern; and the security and residency model, where environments are bound to a geographic region and data is protected by TLS in transit and Transparent Data Encryption at rest.
How to study it. Pin down the vocabulary, because the traps here are precise. A Dataverse table is rows and columns where each column stores one type; a solution packages components, a business rule validates form data, a dataflow transforms incoming data - do not confuse them. Learn the connector ladder: standard connectors are included, premium connectors need a premium licence, custom connectors reach a source with no prebuilt connector. Memorise that Environment Maker builds apps, connections, custom connectors, and flows but cannot administer the environment, while Environment Admin governs it. For security, separate the two encryption mechanisms cleanly: TLS 1.2 or higher encrypts data in transit between devices and datacentres; Transparent Data Encryption encrypts data at rest on disk. Attributing at-rest encryption to TLS is a classic false statement.
Easy to confuse
- Environment Maker versus Environment Admin. Environment Maker is the build role - it creates apps, connections, custom connectors, and flows inside one environment; Environment Admin is the govern role - it provisions the Dataverse database and manages users and roles. If the scenario wants to build resources without tenant control, it is Environment Maker.
- TLS versus Transparent Data Encryption. TLS 1.2 or higher secures the connection between user devices and Microsoft datacentres, encrypting data in transit; Transparent Data Encryption performs real-time encryption of data as it is written to disk, providing encryption at rest. They are two distinct mechanisms protecting two different states of the data.
- Standard versus premium versus custom connector. Standard connectors come included for common services, premium connectors reach licensed or enterprise sources and need a premium plan, and a custom connector lets a service reach an external source that has no prebuilt connector at all. The deciding factor is whether a connector exists and what licence it needs.
Worked example from the PL-900 bank
lock_openFree sampleDescribe the Core Components of Microsoft Power Platformeasy
A retail company asks a citizen developer to build a stock-takedown canvas app, its supporting flows, and the connections those flows rely on, all inside one environment. Which built-in environment role grants exactly the access needed to create those resources without granting tenant-level administrative control?
- AEnvironment Maker, which lets the assignee create apps, connections, custom connectors, and Power Automate flows within that environmentcheck_circle Correct
- BEnvironment Admin, which lets the assignee provision the Dataverse database and add or remove other users from environment roles
- CSystem Administrator, which is the Dataverse security role used to manage the environment after a database is provisioned
- DPower Platform administrator, the Microsoft Entra ID role that opens the security scorecard for the whole tenant
Environment Maker is the build role: it permits creating apps, connections, custom connectors, and flows, but not environment administration. The Environment Maker role exists so makers can create resources within an environment, specifically apps, connections, custom connectors, and flows using Power Automate, without holding administrative control over the environment or tenant.
Why A is correct: Correct. The Environment Maker role exists so makers can create resources within an environment, specifically apps, connections, custom connectors, and flows using Power Automate, without holding administrative control over the environment or tenant.
Why B is wrong: Tempting because admins can also create resources, but the Environment Admin role grants far more than building apps. It performs all administrative actions, which exceeds the access the developer needs.
Why C is wrong: Plausible because it appears in the same grounding, but System Administrator is a Dataverse security role used in place of Environment Admin after the database exists, not the role for citizen-developer resource creation.
Why D is wrong: Plausible because it is a real administrative role, but it is a tenant-wide Entra ID role for centralised security management, not the per-environment role that authorises building apps and flows.
What you must be able to do. Given an app requirement, choose between a canvas app and a model-driven app based on the data sources, the kind of work, and the user experience needed, and recognise how Power Apps integrates with portals, mobile, and Teams.
In one sentenceThe biggest domain: telling canvas apps from model-driven apps by their data source and the work they suit, and knowing how Power Apps fits the wider suite.
Recall check: answer these from memory first
- A maker must combine records from SharePoint, an Excel workbook, and Salesforce on one screen. Which Power Apps type binds to that range of sources, and why not a model-driven app?
- A team manages a data-dense sales process where users move constantly between related opportunity, contact, and account records. Which app type does Microsoft recommend, and why?
- After you share a model-driven app with users, what does the maker still have to do for those users to reach it?
What it tests. How Power Apps lets people build custom applications without traditional code. It covers canvas apps, which start from a blank canvas and connect to many data sources through connectors for a hand-designed experience; model-driven apps, which are built on Dataverse with structured, component-driven layouts that suit data-dense, process-driven work; and the broader capabilities and use cases, including business rules for no-code form logic, sharing behaviour, mobile apps, portals, and Teams integration. This is the heaviest domain by weight, so the canvas-versus-model-driven distinction is worth owning cold.
How to study it. Make the canvas-versus-model-driven call automatic, because it is the most tested distinction on the exam. Canvas apps bind to many data sources - SharePoint, Excel, Salesforce, Dataverse - through connectors and let the maker hand-place every control, so they win when the scenario combines varied sources or wants a tailored screen. Model-driven apps read from Dataverse tables and generate structured layouts, so they win when the scenario is data-dense and process-driven, with users moving between related records like opportunities, contacts, and accounts. Learn the precise facts the yes-or-no items lean on: business rules apply form logic without JavaScript or plug-ins, and sharing a model-driven app assigns access but sends no email, so the maker must distribute the link separately.
Easy to confuse
- Canvas app versus model-driven app. A canvas app connects to many data sources through connectors and is hand-designed control by control, ideal for tailored experiences over varied sources; a model-driven app is built on Dataverse and generates structured layouts, ideal for data-dense, process-driven work across related records. The data source and the kind of work decide it.
- Sharing a canvas app versus sharing a model-driven app. Sharing a canvas app sends users an email with a link; sharing a model-driven app assigns access but sends no email, so the maker must obtain the direct web link and send it separately. Do not assume both notify users the same way.
- Business rule versus JavaScript or plug-in. A business rule applies model-driven form logic - showing, hiding, setting, locking, or validating columns - through a no-code interface of conditions and actions; JavaScript and plug-ins are the pro-code routes for logic beyond what business rules cover. If the scenario says without writing code, it is a business rule.
Worked example from the PL-900 bank
lock_openFree sampleDescribe Microsoft Power Appseasy
A team needs an app to manage a data-dense sales process where users constantly move between related opportunity, contact and account records. Which Power Apps app type does Microsoft describe as especially well suited to this kind of process-driven scenario?
- AA canvas app, because the maker hand-places every screen control
- BA model-driven app, because it suits data-dense process workcheck_circle Correct
- CA Copilot-generated app, because natural language builds it
- DA SharePoint-list app, because the records live in a list
Model-driven apps are the recommended choice for data-dense, process-driven work spanning related records. Microsoft states model-driven apps are especially well suited to process-driven apps that are data dense and make it easy for users to move between related records, giving managing a sales process as an example.
Why A is wrong: A canvas app is a real app type and tempting for tailored UI, but Microsoft frames model-driven apps, not canvas, as especially suited to data-dense process-driven scenarios.
Why B is correct: Correct. Microsoft states model-driven apps are especially well suited to process-driven apps that are data dense and make it easy for users to move between related records, giving managing a sales process as an example.
Why C is wrong: Copilot can generate an app and its data model, but it is an authoring method rather than the app type Microsoft names as suited to data-dense process scenarios.
Why D is wrong: An app built on a SharePoint list is a valid canvas data source, but Microsoft names model-driven apps, not list-backed apps, for data-dense process-driven work.
What you must be able to do. Given a process that needs automating, identify the right flow type by what starts it - automated, instant, or scheduled - and tell cloud flows apart from desktop flows and robotic process automation.
In one sentenceThe automation layer: choosing the flow type from what triggers the run, and knowing when cloud flows give way to desktop flows and RPA.
Recall check: answer these from memory first
- A cloud flow must begin on its own the moment a new record appears in a connected service, with nobody pressing a button and no fixed clock time. Which trigger type fits?
- After a trigger starts a cloud flow, what is the sign-off step the flow performs called, and how does it differ from the trigger?
- From which Microsoft surfaces can an approver respond to a Power Automate approval request without leaving their normal working tools?
What it tests. How Power Automate automates repetitive work. It covers cloud flows and their three trigger types - automated flows that start by themselves in response to an event, instant flows a person starts manually, and scheduled flows that run on a recurrence - plus the building blocks of triggers, actions, templates, and connectors; and desktop flows with robotic process automation for automating legacy and user-interface-based tasks that have no connector. It also covers where approvals surface, such as a Teams adaptive card, an Outlook email, or the Power Automate action centre.
How to study it. Anchor everything on what starts the flow, because that is how the exam frames the trigger questions. An automated trigger fires by itself when an event happens in a connected service, an instant trigger waits for a person to press a button, and a scheduled trigger runs on a fixed recurrence. Separate the two core vocabulary terms cleanly: the trigger is the configured event that begins each run, and actions are the steps the flow performs afterwards, such as requesting an approval. Know that templates are prebuilt flows you search by scenario and then edit, not packaging or connectors. For RPA, remember desktop flows automate legacy and UI-based tasks that have no connector, which is the signal to leave cloud flows behind.
Easy to confuse
- Trigger versus action. A trigger is the configured event that begins each run of a cloud flow, such as a new item being added to a list; an action is a step the flow performs after it starts, such as sending an approval request. One starts the flow, the others are everything it does next.
- Automated versus instant versus scheduled trigger. An automated trigger starts the flow by itself in response to an event, an instant trigger waits for a person to start it manually, and a scheduled trigger runs on a fixed recurrence. Read the scenario for what begins the run - an event, a button press, or a clock.
- Cloud flow versus desktop flow (RPA). A cloud flow connects apps and services through connectors and runs in the cloud; a desktop flow uses robotic process automation to drive legacy and user-interface-based tasks that have no connector. If the scenario describes automating a system through its screen with no API, it is a desktop flow.
Worked example from the PL-900 bank
lock_openFree sampleDescribe Microsoft Power Automatemedium
A retail operations team builds a cloud flow that must begin on its own the moment a new record appears in a connected sales service, with nobody pressing a button and no fixed clock time. Which cloud-flow component fits how this flow needs to begin?
- AAn instant trigger, so a team member starts the flow manually on demand whenever a sale is recorded
- BA scheduled trigger, so the flow runs on a fixed recurrence such as every hour to check for new sales records
- CA Start and wait for an approval action, so the flow begins once an approver reviews each new sales record
- DAn automatic trigger, so the flow begins by itself in response to the new record event in the connected servicecheck_circle Correct
An automatic trigger starts a cloud flow by itself in response to an event, unlike instant (manual) or scheduled (recurring) triggers. Microsoft Learn explains that a cloud flow can be triggered automatically, instantly, or via a schedule, and that actions are the events a flow performs after the trigger starts it. An automatic trigger is the type that starts a flow by itself in response to an event in a connected service, which is what a flow reacting to a new record requires.
Why A is wrong: Instant is a real cloud-flow trigger type, which makes it tempting, but it starts the flow only when a person chooses to run it. The team wants the flow to begin by itself when data changes, so a manual instant start does not fit.
Why B is wrong: Scheduled is a genuine cloud-flow trigger type and is tempting because it removes manual effort, but it fires on a fixed clock recurrence rather than the instant a record appears. The scenario rules out a fixed time, so this lags behind the event.
Why C is wrong: Start and wait for an approval is a real grounded component, but it is an action the flow performs after the trigger starts it, not the mechanism that begins the flow. It pauses for a response rather than initiating the run on a data event.
Why D is correct: Correct. Microsoft Learn explains that a cloud flow can be triggered automatically, instantly, or via a schedule, and that actions are the events a flow performs after the trigger starts it.
What you must be able to do. Given a need to publish an external-facing site, recognise Power Pages as the answer, know how the design studio builds and structures pages, and identify how visitors and staff sign in.
In one sentenceThe external-site layer: knowing when Power Pages is the right service and how its design studio builds, nests, and secures pages.
Recall check: answer these from memory first
- A maker cannot find a template matching the customer portal they want and prefers to design every page from the ground up. Which start option gives them a site with nothing pre-built?
- In the Pages workspace editor, what is the gesture for adding a new horizontal band to the canvas before dropping in text and image components?
- Which built-in identity provider lets staff sign in to a Power Pages site with their existing organisation account, and is it OpenID Connect or OAuth 2.0?
What it tests. How Power Pages builds secure, external-facing business websites on top of Dataverse. It covers when Power Pages is the right choice for a customer or partner portal; how the design studio works, including starting from a business-need template or a blank page, adding sections through the WYSIWYG editor, and nesting subpages that inherit their parent's page template; and how authentication works, with built-in identity providers split between OpenID Connect providers such as Microsoft Entra ID and OAuth 2.0 providers such as Google, LinkedIn, and Facebook for consumer sign-in.
How to study it. First, lock the trigger for the service itself: external-facing website for customers or partners, built on existing Dataverse records, means Power Pages. Then learn the design studio basics the exam asks about. Starting a site, choose a business-need template, a starter layout, or a blank page when nothing fits and you want to build from scratch. Adding content, hover over an editable section area and select the plus sign to add a section, then drop components in. Structuring, a new subpage inherits its parent page's page template. For authentication, separate the two provider families: Microsoft Entra ID is built in as an OpenID Connect provider for staff signing in with their organisation account, while Google, LinkedIn, Facebook, and similar consumer logins are built-in OAuth 2.0 providers.
Easy to confuse
- Blank page versus starter layout versus business-need template. A business-need template provisions a site preconfigured for a scenario, a starter layout applies a cross-industry base, and a blank page starts with nothing so the maker customises the whole site from scratch. Choose blank page only when no template fits and full hand-design is wanted.
- Microsoft Entra ID versus Google, LinkedIn, and Facebook sign-in. Microsoft Entra ID is built into Power Pages as an OpenID Connect provider for organisation accounts; Google, LinkedIn, and Facebook are built-in OAuth 2.0 providers for consumer sign-in. The exam tests which protocol family each provider belongs to, not just that they exist.
- Adding a section versus styling a page. Adding a section means hovering over an editable section area and selecting the plus sign to insert a layout band you can fill with components; styling means opening the Styling workspace to apply themes and visual settings. One adds structure, the other changes appearance.
Worked example from the PL-900 bank
lock_openFree sampleDescribe Microsoft Power Pagesmedium
A retail maker on the Power Pages home page cannot find a business-need template that matches the customer portal they want, and they prefer to design every page themselves from the ground up. Which option lets them start the site with nothing pre-built?
- ABlank page, which lets the maker customise the whole website from scratchcheck_circle Correct
- BPreview template, which opens device views of a selected starter layout
- CStarter layout, which applies a cross-industry solution to the new site
- DChoose this template, which provisions the highlighted business template
Blank page lets a maker build a Power Pages site from scratch when no template matches their needs. When none of the business-need templates fit, the design studio lets makers choose a Starter layout or choose Blank page to customise the website from scratch, so Blank page is the start-with-nothing option.
Why A is correct: Correct. When none of the business-need templates fit, the design studio lets makers choose a Starter layout or choose Blank page to customise the website from scratch, so Blank page is the start-with-nothing option.
Why B is wrong: Preview template only shows how a chosen template looks across devices; it previews an existing design rather than starting an empty site.
Why C is wrong: A Starter layout supplies a ready cross-industry solution, so the site is not blank; it is the choice for a generic head-start, not for building from scratch.
Why D is wrong: Choose this template confirms a specific business-need template and provisions it, which is the opposite of starting with nothing pre-built.