Free Salesforce Certified Platform Administrator II Sample Questions

Correct answer: B — Change Donation__c's sharing setting away from Controlled by Parent so it has its own independently configurable org-wide default and sharing rules

• A. While Donation__c remains set to Controlled by Parent, its sharing is not independently configurable, and no sharing rule can be created directly on the child object; the setting must first change for standalone rules to become available.
• B. Switching Donation__c away from Controlled by Parent gives it its own org-wide default and enables independent sharing rules or manual sharing to be defined on the object, which is the required change to grant access to donations without corresponding campaign access.
• C. View All on Campaign__c would grant broad visibility into campaigns generally, and because Donation__c still mirrors the parent, this does not achieve the stated goal of donation access without campaign access; it also grants far more than intended.
• D. Folders relate to reports and dashboards, not to records of a custom object like Campaign__c, so sharing a folder has no effect on record-level access to donation records.

Correct answer: C — Delegated administrators cannot create or restructure the role hierarchy under any configuration; only users with full administrative access can do that.

• A. There is no time-based restriction that lifts after 90 days; the inability to edit role hierarchy is a permanent limitation of delegated administration, not a temporary one.
• B. View Setup and Configuration only grants read-only visibility into Setup and does not unlock the ability to create or modify roles, so adding it would not solve the underlying restriction.
• C. Delegated administration scopes only user management, assignable profiles/permission sets, and specified custom objects; editing the role hierarchy itself is reserved for users with full administrative privileges regardless of how the delegated group is configured.
• D. Assignable profiles govern which profiles a delegate can hand out to users, not whether the delegate can restructure roles; being listed there does not unlock role hierarchy editing.

Correct answers: B, C, D — Login hours restricting when users can access the org; Default record type and page layout assignment for an object; Login IP ranges that restrict access by network location

• A. Field-level security can be set on a profile, but it can equally be granted through a permission set, so this is not unique to profiles.
• B. Login hours are only defined on a profile; there is no field in a permission set for restricting the time window during which a user can log in.
• C. Default record type and page layout assignments are made per profile per record type and have no counterpart in permission sets, which do not control which layout or default record type a user sees.
• D. Login IP ranges are also profile-only, controlling which networks a user can log in from, with no equivalent restriction available in a permission set.
• E. Connected app access can be granted through both a profile's connected app access policies and through a permission set, so this capability exists in both places.

Correct answer: A — Configure Salesforce as an identity provider and define the external app as a connected service provider

• A. For Salesforce to be the source of authentication for another application, it must be set up as an identity provider, with the external application registered as a service provider that trusts Salesforce's assertions.
• B. Registering an external SAML identity provider makes Salesforce rely on that outside system for authentication, which is the reverse of what is needed when Salesforce itself must authenticate users for another app.
• C. Delegated authentication has Salesforce call an external service to validate credentials for logins into Salesforce; it does not make Salesforce the authentication source for a separate application.
• D. An authentication provider record lets Salesforce users log in using an external OAuth provider like Google or Facebook; it does not make Salesforce the identity source for another application.

Correct answer: A — It is a signed XML document from the identity provider that tells Salesforce who the user is and that authentication succeeded

• A. The SAML assertion is a signed XML statement issued by the identity provider that identifies the authenticated user and lets Salesforce trust that authentication occurred without seeing the user's actual credentials.
• B. SAML SSO does not use a long-lived cached client secret in this way; that description more closely resembles OAuth client credential handling, not a per-login SAML assertion.
• C. Salesforce never receives or forwards the user's raw password in a SAML flow; the identity provider validates credentials independently and only sends the resulting assertion.
• D. A session cookie is a browser-side artifact Salesforce manages after login completes and is unrelated to the SAML assertion exchanged between identity provider and Salesforce during authentication.

Correct answer: A — Keep the standard Business Account and Contact model, and use custom fields or a related object to enrich individual donor profiles

• A. Correct. Since Person Accounts is a permanent, org-wide change, an org wary of irreversible changes should first pursue enrichment options within the existing Business Account and Contact model, such as custom fields or a supporting object.
• B. Person Accounts is one option for individual-level data but is not the only way to enrich donor profiles, and recommending it immediately ignores the stated concern about irreversibility.
• C. Testing in a sandbox does not change the fact that enabling Person Accounts in production remains a permanent, one-way decision; a change set cannot make the feature reversible.
• D. Converting Contacts to Person Accounts still requires enabling the irreversible org-wide feature first, and doing so without further evaluation ignores the leadership's stated risk concern.

Correct answer: A — Enable Person Accounts, which can coexist with the org's existing Business Accounts, and use it only for the new individual buyers

• A. Correct. Person Accounts and Business Accounts can coexist in the same org once the feature is enabled, so the distributor can adopt Person Accounts specifically for individual buyers while leaving existing Business Accounts untouched.
• B. Grouping unrelated individuals under one generic Account misrepresents the data model and does not give each hobbyist their own individually addressable, reportable customer record the way Person Accounts would.
• C. Converting every existing Business Account into a Person Account is unnecessary and disruptive; Person Accounts are meant to supplement, not replace, an org's existing business-customer data.
• D. Account-less Contacts combined with a custom object adds ongoing maintenance overhead and forgoes the built-in reporting, layout, and field support that Person Accounts provide out of the box for individual customers.

Correct answer: A — Set a component visibility rule on the Shipping Details field section so it displays only when Order Type equals Physical Goods.

• A. Dynamic Forms field and section components support visibility rules based on other field values, so setting a rule tied to Order Type equaling Physical Goods will show or hide the Shipping Details section as needed.
• B. A validation rule only enforces data entry requirements at save time; it does not control whether a section is visually shown or hidden on the page.
• C. Building a separate page layout for the record type reverts to static, layout-based display rather than using Dynamic Forms' built-in visibility rules, and is unnecessary extra maintenance.
• D. A formula field that concatenates values into text would still display on every order type unless paired with its own visibility rule, and it does not hide the actual editable shipping fields.

Correct answers: A, B, C — Assigning the app to specific profiles on the User Profiles step of the app configuration.; Assigning the app to specific permission sets so users with those permission sets gain access regardless of profile.; Restricting the app using a permission set license tied to the feature the app depends on.

• A. The User Profiles step of App Manager lets an admin assign the app to specific profiles, controlling which users see it in their App Launcher.
• B. Apps can also be assigned to permission sets, granting access to users who hold that permission set independent of their profile assignment.
• C. When an app depends on a licensed feature, a permission set license can gate access, ensuring only users with that license can use the app.
• D. Organization-wide default sharing controls baseline record-level access for an object across the org; it does not determine which users see an app in the App Launcher.
• E. Field-level security is configured per field in Object Manager, not inside the App Manager wizard, and does not control app-level visibility.
• F. Sharing rules grant record access based on ownership, role, or criteria; they are not a mechanism for controlling which app a user can open.

Correct answer: A — Create two record types on Order, each with its own value set and layout

• A. Record types are purpose-built to let a single object present different picklist values and different page layouts per business scenario, matching the domestic-versus-international requirement precisely.
• B. Creating separate apps changes only the navigation experience and does not vary required fields, picklist values, or page layouts on the underlying Order object.
• C. Validation rules can enforce required fields conditionally, but they cannot alter which picklist values are available or which page layout a user sees, so this alone would not meet all stated needs.
• D. A related list displays child records and does not affect required fields, picklist values, or page layout differences on the Order object itself.

Correct answers: A, B, C — Whether child records must always have a parent assigned at creation; Whether the org needs roll-up summary aggregation from child to parent; Whether child records should be deleted automatically when the parent is

• A. Master-detail relationships require a parent on every child record from creation, while lookup relationships allow an optional parent, so this requirement directly influences which relationship type fits.
• B. Since native roll-up summary fields only work on master-detail relationships, a requirement for that aggregation is a key driver toward choosing master-detail over lookup.
• C. Cascade delete behavior is exclusive to master-detail relationships, so a need for automatic child deletion when the parent is removed points toward that relationship type.
• D. Auto-numbering is a Name field format option available on any custom object regardless of its relationship type, so it does not influence the choice between master-detail and lookup.
• E. Both master-detail and lookup relationships function identically in Lightning Experience and Salesforce Classic, so the interface being used does not factor into this relationship-type decision.

Correct answer: A — The log passed the 20 MB per-log cap, so its earliest lines were overwritten as new output arrived; narrowing the trace flag's categories and levels would preserve more of the early detail

• A. Each debug log is capped at 20 MB, and once a transaction's output exceeds that size, the oldest lines are overwritten by newer ones; scoping the trace flag to fewer categories or lower levels reduces output volume so the relevant early events survive.
• B. Trace flags do capture batch Apex execution like any other Apex transaction, so debug logs are still generated without needing manually inserted debug statements, though the size cap can still truncate them.
• C. Exceeding the org-wide debug log storage allocation blocks new trace flags and logs from being created rather than deleting an already-generated log after a fixed 24-hour period.
• D. A trace flag is always tied to a specific user, class, or Automated Process, not to a queue, and it does not selectively downgrade a transaction to summary-only detail.

Correct answers: A, B, C — Apply a random-value masking method to the SSN field so masked values bear no resemblance to the originals; Apply a pattern-based masking method to the Phone field so masked values retain a realistic phone number format; Apply a fixed-value masking method to the Email field so masked addresses use a consistent placeholder domain

• A. Random-value masking replaces the SSN with an unrelated substitute value each time, which satisfies the requirement that the masked value have no traceable relationship to the original.
• B. Pattern-based masking preserves the structural format of a value, so applying it to the Phone field keeps a realistic, correctly formatted number that still supports UI testing.
• C. Fixed-value masking replaces every instance of a field with the same predefined value, which matches the requirement for a consistent placeholder email domain across the sandbox.
• D. Data Mask fully supports masking a Social Security Number field; there is no restriction requiring encryption instead, and encryption alone would not obscure the value from users with field access in the sandbox.
• E. Phone numbers are commonly treated as personal data subject to masking policy for training environments, and leaving the field unmasked would expose real customer contact information to new hires.
• F. Deleting the Email value entirely would remove data the training environment likely still needs for realistic record display, whereas the stated requirement is a consistent placeholder value, not an empty field.

Correct answers: A, B, C — Click Activate to turn the update on for testing before the auto-activation date; Click Deactivate to turn a previously activated update back off during the opt-in period; Click Review to see a summary and list of potentially affected customizations

• A. Correct: Activate lets an admin turn on a pending update early to test its impact before the auto-activation date arrives.
• B. Correct: During the opt-in period, an admin can deactivate an update they previously activated, and this can be repeated as needed to evaluate impact.
• C. Correct: Review opens details about the update, including a summary and a list of customizations in the org that it might affect.
• D. Incorrect: There is no Snooze control on the Critical Updates page; the auto-activation date is fixed and the update activates permanently once it arrives.
• E. Incorrect: There is no Archive control that removes an update from the schedule; updates remain listed until they activate.

Correct answers: A, B, C — Every standard Order must be associated with an Account, which serves as the ship-to and bill-to context.; Order Products are created by selecting entries from the Price Book that is linked to the parent Order.; Once an Order is Activated, its Order Products become read-only, and a Reduction Order records any decrease.

• A. Correct. Every Order requires a related Account, which establishes the customer context the order is fulfilled and billed against.
• B. Correct. An Order must reference a Price Book, and Order Products are added by choosing from the Price Book Entries available on that linked price book, mirroring how quote and opportunity line items work.
• C. Correct. Activating an Order locks the Order and its Order Products from further edits; any subsequent decrease in quantity or value is handled by creating a related Reduction Order.
• D. Incorrect. An Order cannot have Order Products without first being linked to a Price Book, since each Order Product is created from an entry on that price book.
• E. Incorrect. Order Products do not auto-create Product records; a matching Price Book Entry must already exist for a product before it can be added as an Order Product.
• F. Incorrect. There is no such quantity restriction on Orders; an Order can contain multiple Order Products with any quantities as defined by the business need.

Correct answer: A — Assign a forecast manager to the territory, creating a node in the forecast hierarchy for review and submission

• A. Assigning a forecast manager to a territory creates a manager node in the territory-based forecast hierarchy, giving that person the ability to review, adjust, and submit the rolled-up total for the territory and its sub-territories.
• B. Sharing rules control record visibility and edit access on Opportunities, but they do not create a rollup node in the forecast hierarchy or grant the ability to adjust and submit a territory's forecast total.
• C. Forecast Category is a field on individual opportunity records used to classify deal confidence; it does not create a management layer that can review or adjust an aggregated territory total.
• D. Territory forecasts roll up through the territory hierarchy, not the role hierarchy, so building a parallel role hierarchy branch has no effect on how a territory's forecast rollup is managed.

Correct answer: A — Create separate record types for Installation Guide and Troubleshooting on the Knowledge object, each with its own page layout and assigned reviewer profiles.

• A. Record types on the standard Knowledge object are how Lightning Knowledge distinguishes article styles, each carrying its own page layout with the relevant fields and its own layout assignment by profile. This lets Firmware Version appear only on the installation record type while routing authoring and review differently per type.
• B. Lightning Knowledge articles all live on the single standard Knowledge object; creating separate custom objects is not how article variation is modeled and would break standard Knowledge features like publishing and data categories. Record types, not new objects, are the supported mechanism.
• C. Data categories classify and filter published content for search and visibility purposes; they do not control which fields or page layout appear during article authoring. Field and layout differences are driven by record types, not category selection.
• D. Salesforce Knowledge feature licenses control which users can author or manage articles, not which fields or layouts a given article style displays. Buying more licenses would not create the field or reviewer separation the scenario needs.

Correct answer: A — Add two time triggers on the milestone: one offset before the target time to fire the manager email, and another at the target time to fire the escalation Flow.

• A. Milestone time triggers can be offset earlier or later than the milestone target time, letting one trigger fire the manager email ahead of the deadline and a second trigger fire the escalation Flow once the target time is reached without the milestone being completed.
• B. A single trigger fixed to the exact target time cannot also produce a notification 30 minutes earlier, since each time trigger fires at its own offset rather than at multiple points at once.
• C. Referencing the case's Created Date instead of the milestone's target time decouples the notifications from the actual SLA deadline being tracked, so the timing would no longer reflect the milestone itself.
• D. Salesforce does not merge the pre-deadline warning and the deadline-missed escalation into a single event; each timing need is handled by its own time trigger offset from the milestone target.

Correct answers: A, B — Use a capacity-based routing model and assign a higher capacity weight to the Chat work type than to the Case work type; Set the agent's total capacity so the combined weighted units from all assigned work cannot exceed the configured maximum

• A. A capacity-based routing model lets the admin assign each work type its own weight, so setting a higher weight for chat than for case makes a single chat consume more of an agent's capacity than a single case, matching the requirement.
• B. Capacity-based routing caps an agent's total capacity across all assigned work types combined, so the agent stops receiving new items once the weighted sum of active chats and cases reaches that shared maximum.
• C. Least-active queue-based routing distributes work based on how many open items an agent currently has, but it treats each item as one unit and has no concept of weighting chat higher than case or of a combined capacity ceiling.
• D. Presence statuses control which channels an agent is generally available to receive work from, not a weighted, combined workload limit across chat and case, so manual per-channel toggling does not produce the described behavior.
• E. An Omni-Channel Flow can run logic during routing, such as flagging VIP cases, but dynamically shrinking capacity for VIP assignment is a different mechanism than defining fixed relative weights per work type, so it does not by itself establish that chats weigh more than cases.

Correct answer: D — Omni-Channel, using routing configurations, queues, and skill-based assignment across channels

• A. Sequential sign-off steps direct a record to approvers for a decision; this does not distribute live incoming work across channels to available agents by skill or capacity.
• B. A topic hierarchy that scopes article visibility organizes knowledge content; it has no mechanism for assigning live cases, chats, or social messages to agents.
• C. Milestones with warning and violation actions measure whether service commitments are met over time; they do not perform the real-time, cross-channel assignment of work to agents.
• D. Routing configurations paired with skill-based queues distribute cases, chats, and social posts to the best available agent in real time based on defined skills and remaining capacity, which is exactly this multi-channel requirement.

Correct answer: A — A reporting snapshot that runs a source report on a schedule and loads the results into fields on a custom object

• A. A reporting snapshot schedules a source report to run and maps its results into fields on a custom object, so each run preserves a point-in-time record that can be charted over time. This is the only mechanism built for retaining historical trend data beyond what a live report shows.
• B. Joined reports combine multiple report blocks into one view at run time, but every block still reflects current live data rather than stored historical values, so it cannot reconstruct what totals looked like at prior month-ends.
• C. A bucket field only recategorizes existing current-record values into ranges for the current run of a report; it does not store or preserve any historical state from previous points in time.
• D. A dashboard filter changes which current records are displayed based on a selected value, but it has no ability to retrieve or reconstruct data as it existed in past months since nothing was archived.

Correct answer: A — The duplicate rule's associated matching rule surfaces potential matches in a Potential Duplicates component as the user enters data on the form

• A. When a duplicate rule is active, its underlying matching rule can drive the Potential Duplicates component on the Lightning record creation form, comparing entered values against existing records and surfacing likely matches to the user in real time before the record is saved, which matches the requirement for an inline, pre-save check.
• B. A nightly batch job only evaluates records after they have already been created and saved, and it notifies administrators rather than the rep in the moment, so it does not provide the real-time, pre-save experience the scenario asks for.
• C. A reporting snapshot runs on its own schedule and simply copies report results into a custom object for later review, which is a retrospective process and cannot intercept or warn a user while they are actively entering data.
• D. Time-based workflow rules on legacy Workflow Rules act after a time delay following record creation, not while the user is still on the entry form, so this would not deliver the described instant, pre-save matching behavior.

Correct answer: C — It defines the fields used to query and retrieve records from a big object efficiently, and it cannot be changed once deployed.

• A. Field-level security and object permissions are governed by profiles and permission sets, not by the Big Object Index, which is strictly a query-performance mechanism.
• B. Indexes are metadata used for querying the big object itself; they do not create or maintain a duplicate copy of the data inside a separate standard custom object.
• C. A Big Object Index specifies the combination of fields used to look up and retrieve big object records efficiently at scale, and because it is fundamental to how the data is physically organized, it becomes immutable once the big object is deployed.
• D. Big Object Indexes have no automatic time-based compression behavior; they define lookup fields for queries and do not act on record age or storage consumption.

Correct answers: A, B — It runs from the command line using configuration and mapping files that a scheduler can trigger; It can target the Case object directly, unlike the wizard's fixed list of supported standard objects

• A. The Data Loader command-line interface accepts a configuration file, field-mapping file, and process-conf.xml, letting an OS-level scheduler like cron or Windows Task Scheduler trigger the load with no person present.
• B. Data Loader can target Cases directly, while the Data Import Wizard's object picker is limited to a fixed set of standard objects that does not include Case, making Data Loader the only fit for this object.
• C. Both the Data Import Wizard and Data Loader can read CSV files, so file format support is not what distinguishes the two tools for this scenario.
• D. The browser-based wizard experience described here belongs to the Data Import Wizard, and it requires a logged-in user clicking through screens rather than running unattended.
• E. Data Loader works with both standard and custom objects, including Cases, so object-type restriction is not a reason it fits an unattended nightly job.
• F. Data Loader does not send automatic email summaries as a built-in feature; any nightly notification would need to be built separately outside the tool.

Correct answer: A — The Stage column in the CSV was not mapped to the correct Salesforce field

• A. If the CSV's Stage column was left unmapped or mapped to the wrong Salesforce field during the mapping step, those values would never reach the Stage field, leaving it blank on the inserted records.
• B. Stage is a standard, writable field on Opportunity, and Data Loader can populate it like any other editable field, so there is no such restriction.
• C. Data Loader does not strip picklist values on insert; picklist fields are loaded the same way as other field types as long as the value is a valid selection.
• D. Modern supported API versions fully support inserting picklist field values, so an API version limitation would not explain blank Stage values.

Correct answers: A, B, F — Once installed, the components can be freely edited or removed by the org; There is no supported upgrade path once the source components change later; The transfer still works even though the two orgs have no relationship

• A. Unmanaged package components become ordinary, fully editable components in the target org once installed, which the administrator should factor in since the receiving org could alter or delete them.
• B. Unlike managed packages, unmanaged packages have no built-in upgrade mechanism, so any later fix to the source components would require a separate manual transfer.
• C. Unmanaged packages are not eligible for AppExchange listing, so this option would not be a route to marketplace distribution even if that became a future goal.
• D. Package installation does not rely on the deployment connection framework that change sets use, so no such configuration step is required in the receiving org.
• E. Unmanaged packages leave all code visible and editable rather than hiding it, so code protection is not a benefit an administrator gets from choosing this approach.
• F. Because installation uses a package link rather than a deployment connection, it works between orgs that have no established relationship, which is a genuine consideration favoring this approach for the partner scenario.

Correct answer: A — This deployment must meet the same 75% overall code coverage requirement as any other production deployment

• A. Deploying Apex to a production org always triggers Salesforce's standard requirement of at least 75% overall code coverage with no failing tests, and this rule applies uniformly whether the deployment happens via change set, Metadata API, or another tool.
• B. There is no exemption for change sets; the code coverage threshold is enforced by the platform at the point Apex is deployed to production, regardless of which deployment mechanism triggered it.
• C. Code coverage validation is tied to the presence of Apex classes or triggers being deployed, not to whether a Visualforce page happens to be included in the same change set.
• D. There is no threshold based on the number of existing classes; every production deployment containing Apex is checked against the same 75% coverage bar regardless of how much code already exists in the org.

Correct answer: A — That metadata type is not supported for transport through change sets and must be migrated using another tool, such as the Metadata API

• A. Not every metadata type Salesforce supports is available for change sets; some component types can only be migrated using the Metadata API, Salesforce CLI, or another deployment tool, which explains why it never appears in the picker.
• B. The change set component list reflects what currently exists in the source org, not a synced snapshot tied to sandbox refresh timing, so a newly created component in the same org would still normally appear.
• C. Access to the change set component picker is controlled by permissions related to creating and managing change sets, not by "Modify All Data," which governs record-level data access rather than metadata visibility.
• D. Change sets do not filter the component list by a modification-date window; all eligible components of a supported type are listed regardless of when they were last changed.

Correct answer: A — An approval process with sequential steps and the default submission-lock action

• A. An approval process is built for exactly this pattern: it defines ordered approval steps with named approvers, and its initial submission action locks the record by default so only approvers and admins can edit it while it is pending.
• B. A flow can send sequential emails, but it has no native approval/reject decision mechanism or record-locking behavior, so it cannot enforce sign-off routing on its own.
• C. Workflow rules are a legacy tool with no new creation allowed, and even when active they only fire time-dependent actions on a schedule, not sequential human approval routing with record locking.
• D. A validation rule only blocks a save based on field conditions at the moment of edit; it cannot route a record to specific people in sequence or lock it for a pending review period.

Correct answer: A — An Apex trigger with explicit transaction and bulk-safe DML handling

• A. Apex gives full control over transaction boundaries and rollback behavior, and a well-written bulk-safe trigger can process large batches of records within governor limits far more reliably than declarative loops.
• B. A flow loop updating records one at a time across three related objects for 5,000 records is likely to exceed flow interview limits and does not offer the same fine-grained transaction rollback control as code.
• C. A manually launched screen flow depends on a person running it for every shipment, which does not scale to an automated batch of 5,000 records and introduces the risk of missed runs.
• D. Workflow rules are a retired tool with no new creation, and chaining field updates across three objects would still not provide multi-object transactional rollback.

Correct answer: D — A record-triggered flow configured to run before the record is saved, updating the field directly without a separate database operation

• A. A nightly batch introduces an unnecessary delay for a value that needs to reflect the current save, and it does not take advantage of the record already being in memory during that transaction.
• B. Requiring a manual quick action adds a dependency on user behavior for a calculation that should happen automatically every time the source field changes, and it does not fit a bulk data-load scenario.
• C. Updating the same record after it has already saved forces a second, separate database update, which is slower and consumes more resources during bulk operations than making the change before the record commits.
• D. A before-save update can change a field on the triggering record directly as part of the original save operation, avoiding an extra database transaction and making it the most efficient option for bulk performance.

Correct answers: A, B, D — Looping through a collection of related records to perform calculations before the parent record is submitted for approval; Making an HTTP callout to an external system as part of the automation logic; Displaying an interactive multi-screen interface that collects additional information from a user before submission

• A. Approval processes have no loop or collection-processing element; iterating over related records to perform calculations is a capability unique to Flow Builder among these declarative tools.
• B. An approval process cannot make an outbound HTTP callout as part of its approval logic; only a flow (or Apex) can perform that kind of external integration.
• C. Requiring one named person to approve a record before it is final is a core, built-in capability of approval processes themselves, so it is not something only Flow Builder can provide.
• D. Presenting an interactive, multi-screen interface to collect information from a user is a screen flow capability; an approval process has no equivalent interactive screen mechanism of its own.
• E. Approval processes natively support dynamically determined approvers, including routing to a queue or to a value looked up from a field on the record, so this is not exclusive to Flow Builder.

Correct answer: A — Roll-up summary recalculation can run asynchronously relative to the triggering flow's transaction, causing a short display delay

• A. Correct. Roll-up summary recalculation triggered by changes made through automation such as flows can occur asynchronously relative to the initiating transaction in some cases, which is a documented reason a dependent roll-up value may not reflect an update instantly.
• B. Incorrect. There is no "Recalculate Roll-Up Summary" flow element; roll-up summary fields are a declarative field type on the master object that recalculate automatically based on platform logic, not through a manual flow element.
• C. Incorrect. Roll-up summary fields recalculate in response to relevant child record changes throughout the day, not on a single daily maintenance schedule, so this does not match documented behavior.
• D. Incorrect. Roll-up summary fields remain a fully supported native field type and have not been deprecated or replaced by a Flow-based equivalent.

Correct answer: A — Open Flow Trigger Explorer for the Account object and set an explicit run order for the three flows

• A. Flow Trigger Explorer is built for exactly this situation, letting an admin assign an explicit run order to multiple record-triggered flows on the same object and event so results become predictable.
• B. Salesforce allows multiple active record-triggered flows of the same type on one object; there is no one-flow-per-object limitation forcing deactivation of the others.
• C. Converting to scheduled flows would decouple the logic from the actual record save event, which changes the automation's behavior rather than fixing the ordering conflict on save.
• D. A formula field recalculates automatically based on other field values, but it does not control or resolve the underlying conflict in run order between the three record-triggered flows.

Correct answer: A — After the transaction commits, since async processing is deferred until the save finalizes

• A. Outbound messages and future Apex calls are queued as asynchronous work and only actually execute once the triggering transaction has successfully committed, ensuring they act on final, saved data.
• B. Asynchronous actions are deliberately deferred rather than executed inside the same uncommitted transaction as triggers and workflow rules, since they must wait for a successful commit.
• C. The after-trigger phase runs synchronously as part of the same transaction, whereas outbound messages and future methods are dispatched separately after that transaction finishes.
• D. There is no manual Setup action required to process queued asynchronous work; Salesforce automatically executes it once the triggering transaction commits.