MCE-Con-201 Salesforce Certified Marketing Cloud Engagement Consultant Questions and Answers

Because the Marketing Cloud Connect app-level connection itself reports as successful, but Sales Cloud objects still fail to appear or synchronize within Contact Builder, the most direct and efficient troubleshooting step is to verify that the specific users involved in the integration — both on the Marketing Cloud side (the integration/API user) and the Sales Cloud side — hold the correct permission sets and role-based access required to expose and synchronize Salesforce objects into Contact Builder, since a successful app connection does not automatically guarantee sufficient object-level and field-level permissions for synchronization to function. This is a targeted, low-effort first diagnostic step consistent with standard troubleshooting methodology: verify permissions before escalating to more invasive fixes. Option A, sending a test email to validate tracking, addresses a different layer of the integration (send-tracking sync) rather than the object-visibility issue described in Contact Builder specifically. Option B, creating new users and reinstalling entirely, is a drastic, high-effort action that should only be considered after simpler causes like permissions have been ruled out. Option C, validating Sales Cloud users have BU access, is narrower and less complete than validating permissions across both platforms, which is the more thorough and correct first diagnostic. Reference: Marketing Cloud Connect Troubleshooting, Contact Builder Salesforce Data.

Tracking Extract, run as an Automation Studio Data Extract Activity, is the documented mechanism for pulling detailed, per-send delivery status data — including records marked as Not Sent — for journey-driven email sends. Unlike the standard Sent Data View, which primarily reflects records that were successfully processed for sending, Tracking Extract exposes the fuller set of delivery outcome codes, including exclusions and failures, giving the granularity needed to specifically isolate and report on the ' Not Sent ' population from journey emails. Option A, querying the Sent Data View, is not reliable for this purpose because that view is generally populated with records that were successfully sent, not the ones that were excluded or failed to send — it is the wrong data source for isolating non-sends. Option B references a ' Subscribers Not Sent to ' standard report, which is not a real, out-of-the-box Marketing Cloud report name. Option D, Analytics Builder Intelligence Reports, refers to a different reporting layer focused on broader engagement and journey performance visualizations rather than granular per-record delivery-status extraction, and is not the documented tool for this specific ' Not Sent ' reporting need. Reference: Tracking Extract, Automation Studio Data Extract Activity.

Identifying subscribers who did not receive their intended email over a rolling 24-hour window requires filtering logic against send-outcome data (such as the Sent, Bounce, or related data views) based on a specific date range and a negative-outcome condition — precisely the kind of conditional, criteria-based filtering that a SQL Query Activity is built to perform. This activity is the essential step that actually identifies and isolates the correct population of non-received subscribers before anything can be exported; without it, there is no mechanism within the automation to determine which specific subscribers meet the ' did not receive ' criteria. While a Data Extract Activity (C) would ultimately be needed to produce the exportable file and place it on the SFTP, the scenario is asking specifically which activity is required to correctly identify the target population in the first place — and that identification step is the SQL Query Activity, making it the most essential, non-optional component of this automation given the filtering complexity described. A Report Definition Activity (A) is not the standard mechanism for this kind of granular, criteria-based subscriber-level filtering. A ' SendLog Extract ' (B) is not a distinctly named standard Automation Studio activity separate from the general Data Extract Activity. Reference: Automation Studio SQL Query Activity, Data-Driven Segmentation for Exports.

Send Logs, once enabled through Send Logging, provide a unified, cross-channel record of send activity that spans email, SMS, and push notifications, giving the consultant a single, queryable structure to build reporting that covers all three channels together rather than needing to separately reconcile disparate channel-specific data sources. This cross-channel breadth is precisely what the requirement calls for, since the scenario explicitly asks about send data spanning SMS, Push, and email together, not just one channel in isolation. The Sent Data View (B) is scoped specifically to email send activity and does not natively capture SMS or Push send events in the same structure, making it insufficient on its own for a unified, cross-channel query. Tracking Extracts (C) similarly focus primarily on email delivery and engagement tracking data and are not the standard mechanism for consolidated multi-channel (SMS/Push/email) send reporting. Contact History (D) is not a defined, standard Marketing Cloud reporting object for this purpose in this context. When a scenario explicitly spans multiple messaging channels and asks for a single queryable send-data source, Send Logs is the correct, unifying answer among the choices presented. Reference: Send Logging, Cross-Channel Reporting.

The View As Web Page link relies on the underlying tracking/job data remaining associated with a stable subscriber record across time, which is why the daily-refreshed audience data extension must be maintained using an upsert (update-if-exists, insert-if-new) write pattern rather than a destructive overwrite. An upsert preserves existing subscriber records and their historical tracking associations while still adding new qualifying contacts and removing/updating those no longer relevant, keeping the View As Web Page link ' s underlying references intact for previously sent messages. A full overwrite (option B), by contrast, effectively deletes and recreates every row in the data extension daily, which can sever the referential continuity that the View As Web Page link and its tracking depend on, causing it to break for recipients across refresh cycles. Option A, using AMPscript against a send logging data extension for content display logic, addresses dynamic content selection, not the structural data-refresh mechanism that protects the View As Web Page link. Option D, exclusion lists, manages suppression of already-sent recipients on subsequent sends — a send-list-hygiene concern unrelated to preserving the web-page-link ' s integrity through daily audience changes. Reference: SQL Query Upsert Behavior, View As Web Page / View Email as Webpage.

Marketing Cloud Connect ' s integration user requires both the API User checkbox enabled — which designates the account as a dedicated, non-interactive system/integration user rather than a standard human login, exempting it from normal password-expiration and login-count restrictions that would otherwise break the integration — and sufficiently elevated role permissions to perform the full range of actions Marketing Cloud Connect needs, which is why Salesforce ' s documented setup specifies both the Administrator and Marketing Cloud Administrator roles be applied to this user. Together, these two settings ensure the integration user can reliably authenticate on an ongoing, unattended basis while having the necessary permissions across both the Marketing Cloud and Salesforce sides of the connection to support synchronization, triggered sends, and Connect app functionality. Option B leaves the user without the roles needed to actually perform Connect ' s required operations, which would cause integration failures. Options C and D both leave the API User checkbox deselected, meaning the account behaves as a standard interactive user subject to expiration and session policies unsuited to an always-on system integration, and option D ' s Marketing Cloud Viewer role is read-only and insufficient for an integration user that must write and synchronize data. Reference: Marketing Cloud Connect API User Setup.

Tracking Extract, configured as a scheduled Automation Studio Data Extract Activity, is the standard, documented mechanism for pulling detailed, recurring email engagement metrics — send, open, click, bounce, and unsubscribe events — on a defined cadence such as hourly, producing a file the analytics team can ingest into their own reporting or next-best-action systems on a predictable schedule. This is precisely designed for the kind of recurring, scheduled, granular metrics pull the scenario describes, giving the analytics team consistent hourly visibility to drive timely downstream actions for the sales team. Data views (A) are queryable in real time via SQL, but the scenario specifically calls for a repeatable, scheduled hourly delivery mechanism, which is what Tracking Extract, run on an hourly automation schedule, is purpose-built to provide as an exportable, consumable file. Scheduled Reports (B) in Marketing Cloud are generally oriented toward summary-level dashboards and standard reporting views rather than granular, extract-level engagement data suited to feeding an external next-best-action process. Send Logs (D) capture per-send detail (address, content used) rather than the aggregate engagement metrics (send/open/click counts) the analytics team specifically needs for trend-based next-best-action decisions. Reference: Tracking Extract, Scheduled Automation for Reporting.

Because the customer already possesses a fixed, curated list of known competitor email addresses — a static, externally-sourced list rather than a set dynamically derived from an internal query against existing behavioral or attribute data — the correct approach is to load these known addresses directly into an auto-suppression list, which then automatically applies to all qualifying future sends and prevents any of the listed addresses from receiving marketing communications going forward without requiring the list to be manually re-attached to each individual send. Populating it directly with the known email addresses (rather than subscriber keys, as in option B) is appropriate here specifically because the source data the customer has in hand is a list of raw email addresses, not internal subscriber key values, and the suppression mechanism can match directly against the email address field. Option A describes building a list via query to ' exclude the subscriber key, ' which misapplies a query-based approach to what is fundamentally a static, externally-provided list rather than data that needs to be dynamically derived through SQL logic. Option D ' s ' exclusion list ' is not the correct, standard Marketing Cloud terminology or mechanism for this scenario — auto-suppression list is the actual, documented feature that automatically applies across future sends. Reference: Auto-Suppression Lists.

SSH key-based authentication is the correct, security-conscious solution because it replaces the password-based login entirely with a cryptographic key pair — the external data warehouse system authenticates using its private key rather than a password subject to the 90-day expiration policy, eliminating the recurring maintenance burden of updating the file transfer script every time a password rotates, while actually strengthening security compared to password authentication rather than compromising it. This is precisely why Salesforce documents SSH Key Management as the recommended approach for unattended, script-driven, recurring file transfer integrations that need to avoid password-expiration disruption. Option A, disabling password expiration entirely for FTP users, directly conflicts with the explicit requirement that the customer does ' not want to compromise security ' — permanently exempting an account from password rotation policy is a security downgrade, not a secure workaround. Option B, enabling SAML Single Sign-On, is designed for interactive human user authentication into the Marketing Cloud application interface, not for unattended, script-based SFTP file transfer authentication. Option D is a non-solution — the scenario already specifies an Enhanced FTP account, which uses SFTP protocol; simply restating ' use SFTP ' does not address the underlying password-expiration/automation problem at all. Reference: SSH Key Management, Enhanced FTP Authentication.

The web analytics file arrives unpredictably, so a File Drop Automation — not a Scheduled Automation — is the correct trigger, immediately ruling out options C and D. After the Import File Activity loads the raw web analytics data into a staging DE, a Verification Activity must run next: it evaluates whether the target DE contains records and, if empty, stops the automation from proceeding, which is exactly the ' do not want the automation to finish running ' requirement. Only after that gate passes should the two Query Activities run: the first joins the web analytics staging DE against the Customer Reference DE to append Subscriber Key/Email Address (resolving identity for the first file), and the second performs the equivalent join for the five-day non-purchase file, since it also lacks identity fields and will be needed later for the journey ' s decision split. Data Extension Entry — the activity that makes the population available for journey injection — logically comes last, once both supporting data sets have been fully resolved and are ready for consumption. Placing Data Extension Entry before the queries (as in options B, C, and D) would inject contacts before the decision-split data is prepared. Reference: File Drop Automations, Verification Activity, Query Activity Sequencing for Journey Data Prep.

Send Logging, once enabled, writes granular, per-send detail — including the exact email address and content used for each individual message — into a dedicated Send Log Data Extension that marketers can directly query for troubleshooting purposes such as verifying which address received which content on a specific send. Because the scenario explicitly states the Send Log is already being populated and is growing in size, the natural and correct recommendation is to query that existing Send Log Data Extension directly via SQL rather than standing up an entirely separate reporting mechanism. Option A, a Filter Activity on the Sendable Data Extension, only reflects current-state subscriber data, not the historical, point-in-time send detail needed to troubleshoot a specific past send. Option B, Tracking Extracts, capture engagement and delivery-status events rather than the detailed content/address-level send record that Send Logging specifically provides — it answers a different question (did it send/open/bounce) rather than ' which exact address and content combination went out. ' Option D, Sent and Subscriber Data Views, reflect general send and current subscriber status but similarly lack the granular per-send content-and-address audit trail that Send Logging is uniquely designed to preserve. Reference: Send Logging, SendLog Data Extension.

The Business Unit-level Email setting ' Subscribers will be unsubscribed from this business unit only ' is the direct, native, low-effort configuration that satisfies this exact requirement: it scopes an unsubscribe action taken within the children ' s fashion business unit to that business unit alone, leaving the subscriber ' s opt-in status intact across all other business units, including the parent Enterprise business unit — with no custom development, attribute engineering, or filter-building required, which aligns with the customer ' s minimal technical resources. This is a single checkbox-level setting change, making it by far the lowest-effort, most maintainable solution among the options. ' Use separate Subscription Center for this business unit ' (A) affects which subscription-preference page a subscriber sees when managing their preferences, not whether an unsubscribe event is scoped to one business unit versus propagating account-wide — it solves a different problem. Option B describes manually engineering a custom filter against the All Subscribers list to track unsubscribes, and option D proposes custom attribute-based tracking — both are unnecessarily complex, developer-dependent workarounds for a requirement natively solved by a single setting. Reference: Business Unit Email Settings, Unsubscribe Scope Configuration.

Field-Level Encryption protects sensitive PII (such as SSNs, payment details, or other regulated data) by encrypting it at rest within the data extension field, but this security benefit comes with a documented functional trade-off: because the data is stored in encrypted form, it cannot be used as a condition in SQL Query Activities, Data Filters, or Decision Splits — encrypted fields cannot be segmented, filtered, or queried against. This is the critical architectural limitation a consultant must surface before recommending encryption for any field the business also intends to use for personalization logic or audience segmentation, since doing both on the same field is mutually exclusive. Option B is inaccurate; Field-Level Encryption is not restricted to list-based sending and applies at the data-extension field level regardless of the sending method used downstream. Option C incorrectly implies encryption keys are managed per business unit as a hard requirement — key management is an account-level consideration, not something mandated per BU by the feature itself. Option D is also incorrect; encrypted fields remain encrypted and masked in standard reporting and data views precisely because the encryption is meant to protect the data from unauthorized visibility, not selectively expose it in reports. Reference: Marketing Cloud Field-Level Encryption.

Because the new dataset is refreshed independently, on its own separate feed cadence, from the main contact data, it should live in its own dedicated data extension rather than being merged directly into the core contact record. In Data Designer, this new data extension is then linked (related) to the existing contact-based data extensions through a defined relationship on a common key, allowing the data to be joined for segmentation, personalization, and decision-split logic while preserving each source ' s independent update lifecycle. This relational approach is precisely what Data Designer is built for — modeling one-to-one or one-to-many relationships between separately-maintained data extensions without collapsing them into a single monolithic table. Option A ' s ' new population ' terminology does not correspond to a real Data Designer configuration concept — Data Designer works in terms of related data extensions, not ' populations. ' Option B unnecessarily specifies a ' Salesforce data extension ' (implying a synchronized object from a connected org), which is not indicated anywhere in the scenario and adds an unsupported assumption. Option D, adding attribute fields directly to the contact record, would break the independent-feed requirement, since the contact data extension and the new dataset are updated on different schedules and by different sources. Reference: Data Designer, Data Extension Relationships.

The most reliable, documented pattern for cross-journey handoff based on engagement is to capture the engagement outcome as a persisted, queryable data point rather than attempting to read it indirectly from system-level journey data views, which do not reliably expose per-contact Engagement Split path results in a directly queryable format. In this pattern, Journey A includes an Engagement Split after the email send, and the branch a contact falls into (engaged vs. not) is written out via an Update Contact Activity into a Boolean field on a dedicated Engagement data extension — creating a durable, explicit record of that contact ' s engagement outcome. A separate Automation Studio process then queries this Engagement data extension directly to build the entry-source population for journey B, giving a clean, reliable, and fully controllable handoff mechanism between the two journeys. Options A, B, and D all attempt to derive the engagement outcome by querying system Journey/Journey Activity data views for Engagement Split results, which is not a standard, reliably queryable data structure for this purpose — Engagement Split path outcomes are not natively exposed as a clean Boolean field in those system data views the way this scenario ' s options imply. Reference: Journey Builder Engagement Split, Update Contact Activity, Cross-Journey Data Handoff Patterns.

Pulling live, external RSS feed content into an email at the moment of send requires the platform to reach out to and retrieve data from an outside URL during send processing, which is precisely what External Content (B) is designed for: it is a Content Builder feature that fetches and injects HTML/text from an external source, including RSS-driven blog feeds, at send time. AMPscript (D) achieves the same underlying capability through code — using HTTP-retrieval functions such as HTTPGet() combined with parsing logic to pull the RSS feed ' s XML and render the latest blog entries dynamically within the email body, giving developers a more customizable path to the same real-time-pull outcome. Personalization Strings (A) only substitute predefined subscriber, sender, or system-level attribute values already known to Marketing Cloud at send time — they have no capability to reach out to an external URL or retrieve live third-party content. Dynamic Content (C) selects between pre-authored content variations based on subscriber attributes and audience rules; it does not fetch or pull in external, live-updating data from an RSS feed — its content options must already exist within Marketing Cloud rather than being retrieved externally. Reference: External Content, AMPscript HTTP Functions.

The Contact Data View is the authoritative system data view that reflects the full Contact model in Marketing Cloud, including all contacts synchronized from connected Salesforce orgs and those created natively. A scheduled Automation Studio sequence — SQL Query Activity against the Contact Data View, followed by a Data Extract Activity configured to email the resulting file or a notification — satisfies a recurring, hands-off weekly report to the CMO. Option C is close in mechanics but targets the Subscriber Data View, which reflects the legacy All Subscribers list rather than the full Contact model; in a Contact Model-enabled account, this undercounts or misrepresents the true contact base. Option A references a ' Contact Extract Activity, ' which is not a standard Automation Studio activity name — Data Extract Activity is the correct term, and it operates on Data Extensions, not directly on a live count without a preceding query. Option B ' s ' Contacts Counts Standard Report ' is not an out-of-the-box Marketing Cloud report. Consultants should default to querying the correct data view for the object being counted, then automating the export/notification with Automation Studio rather than relying on manufactured report names. Reference: Marketing Cloud Data Views, Automation Studio Query and Extract Activities.

Because the same Contact/Subscriber record is being reused with a changing Email Address value at send time, standard data views and reports that key off the current subscriber record will not preserve the specific address that was actually used for a given historical send — they reflect present-state data, not point-in-time send detail. Send Logging is the feature purpose-built for this exact requirement: when enabled, it captures a persistent, granular record for every individual send, storing the exact email address (and other attributes) as they existed at the moment of send, independent of any later changes to the subscriber record. That data lands in a queryable Send Log Data Extension the consultant can report from. Data Views (A) reflect current, not historical, attribute values and would show whichever email address is on the record now, not what was used per send. Recent Email Send Report (C) is a high-level send-count summary, not per-subscriber address-level detail. Tracking Extract (D) captures engagement events (opens, clicks) rather than the address used at send time and is not designed to solve the ' which address received which specific send ' problem. Reference: Send Logging, Marketing Cloud Reporting.

This is a direct restatement of Q16 ' s underlying mechanism: the Event Notification Service is Marketing Cloud ' s real-time, webhook-based method for pushing granular send-status events — including Sent and NotSent outcomes — to an external, subscribed endpoint as they occur, rather than requiring the analytics team to poll or wait for a batch process. Given the explicit legal-obligation framing, where the team needs assured, immediate visibility into whether a transactional message actually reached the recipient, a push-based, real-time notification mechanism is the only option among those listed that genuinely delivers status ' in real time ' rather than on a delay. Platform Events API (A) is a Salesforce Sales/Service Cloud pub/sub messaging framework for custom Salesforce-side event architectures — it is not the native Marketing Cloud mechanism for transactional send-status notifications. ' Data Retrieves on the Send Object ' (B) is not a defined, real, real-time Marketing Cloud feature or standard terminology. Data View Export ' with every send ' (D) describes a batch or scheduled extraction pattern that, even if run frequently, introduces polling latency and operational overhead incompatible with genuine real-time delivery, unlike a push-based webhook subscription. Reference: Event Notification Service, Transactional Messaging.

SMS/MMS sends in Journey Builder require a fully normalized phone number — country code plus the local number with no additional formatting characters — present directly in the entry-source data extension at the point the SMS activity executes; Marketing Cloud does not automatically infer or append a missing country code on the platform ' s own initiative. Because the root cause described is that the Salesforce-sourced mobile number field lacks the country code altogether, the fix must occur at the data source: formatting the phone field value in Salesforce to explicitly include the country code prefix (+1 for United States numbers) before or as it syncs into the journey ' s entry data extension, ensuring the normalized format Journey Builder requires is present from the start. There is no native ' Normalize Phone Number ' toggle in Journey Builder settings (option A) that automatically fixes missing country codes — this is a fabricated distractor. Configuring a Salesforce field ' s locale setting (option B) affects display/input formatting conventions within Salesforce itself, not the actual stored digit string synced to Marketing Cloud. There is likewise no ' sending country ' selector within the SMS activity configuration (option C) that retroactively corrects malformed source data. Reference: SMS/MMS Activity Phone Number Requirements, Journey Builder Salesforce Entry Sources.

Given the explicit statement that immediate ingestion is not required — the data only needs to be available in time for a daily scheduled send — and that submissions arrive at an extremely high frequency (thousands per minute), a daily bulk export from the external system to SFTP, followed by a scheduled Automation Studio import, is by far the most efficient and appropriately-scaled architecture: it avoids the overhead, cost, and API rate-limit risk of processing potentially hundreds of thousands of individual real-time API calls per day when a single, efficient nightly batch transfer accomplishes the same business outcome with far less operational complexity and infrastructure load. Option B, pushing each individual record via a real-time API call on every form submit, is a significant over-engineering of the requirement given real-time delivery was explicitly stated as unnecessary, and would introduce meaningful API volume and rate-limiting concerns at the stated submission frequency. Option C, recreating the external form natively inside Marketing Cloud, is an unnecessary architectural change that abandons existing, presumably functional external infrastructure without clear justification. Option D misapplies a Journey Data Source, which is designed for individual, real-time journey-entry-triggering events, not for bulk batch contact acquisition feeding a scheduled send. Reference: Automation Studio Batch Import, SFTP-Based Data Ingestion.

The Transactional Messaging API is Salesforce ' s purpose-built, API-driven mechanism for sending exactly this class of message — a one-off, highly time-sensitive, security-relevant notification (new-IP login alert) that must be dispatched with minimal latency directly from an external system ' s real-time event trigger, without waiting on any batch, scheduled, or journey-evaluation cycle. Because the request is initiated synchronously via a direct API call at the exact moment the security event (new IP login) occurs, and the response includes send-status confirmation, it delivers the fastest possible deployment path among the available options while also supporting the dynamic, per-message content (the specific login IP) that needs to be included. Injecting the subscriber into a Journey (A) introduces journey entry-evaluation timing and processing overhead that, while often fast, is not as directly immediate as a synchronous transactional API call purpose-built for exactly this kind of urgent, individual, security-triggered message. ' User Initiated Emails Interaction ' (B) and ' Triggered Emails Interaction ' (D) are not standard, precisely-named Marketing Cloud features matching this real-time, API-driven transactional use case as directly as the Transactional Messaging API. Reference: Transactional Messaging API, Security/Fraud Alert Messaging Patterns.

Salesforce ' s documented sandbox-refresh process for a Marketing Cloud Connect integration requires disconnecting the existing Marketing Cloud Connect connection first, before the underlying Salesforce sandbox refresh actually occurs — attempting to refresh the sandbox while still connected risks corrupting or orphaning the existing integration state, since a sandbox refresh fundamentally replaces the sandbox ' s underlying data and configuration with a new copy from production. After disconnecting, the sandbox refresh itself is performed on the Salesforce side. Once the refreshed sandbox is ready, the Marketing Cloud Connect connection must be re-established by reconnecting, followed by re-authenticating the integration credentials against the newly-refreshed sandbox instance, and finally synchronizing data to rebuild the Contact Builder object mappings and synchronized data extensions against the refreshed org ' s current state — synchronization must come last because it depends on a fully authenticated, active connection already being in place. Options A and C both incorrectly place Refresh before Disconnect, risking connection corruption during the refresh process itself. Option D places Synchronize before Authenticate, which is technically out of order since synchronization cannot meaningfully occur without an already-authenticated connection. Reference: Marketing Cloud Connect Sandbox Refresh Procedure.

Introducing Sales Cloud alongside an existing Marketing Cloud instance that has been keyed on Email Address for a year creates a structural identity mismatch, because Sales Cloud-driven integrations (Marketing Cloud Connect, Synchronized Data Extensions) key subscriber/contact records on the Salesforce record ' s own unique identifier, not raw email address — so without remediation, the same real-world person can end up represented as two disconnected records once the Salesforce integration goes live. Uploading a CSV that maps and updates existing Marketing Cloud records to use properly migrated Subscriber Keys aligned to their future Salesforce identifiers (B) directly resolves this by re-keying the existing base before Salesforce data starts flowing in, preventing a second, duplicate record from being created for the same person. Alternatively, engaging a dedicated services/migration effort to have existing records comprehensively updated with new keys sourced from the Sales Cloud instance (D) accomplishes the same de-duplication goal through a structured, professionally-managed remediation project. Option A is incorrect — Marketing Cloud does not natively ' re-key ' or deduplicate existing Email-Address-keyed records simply because Sales Cloud is later introduced; without remediation, duplication risk remains. Option C, purging historical records, is a destructive approach that loses valuable existing subscriber history rather than preserving and correctly re-keying it. Reference: Subscriber Key Migration, Sales Cloud Integration De-Duplication.

The requirement explicitly calls for an attribute-to-attribute comparison between the customer number field in the journey ' s entry source data extension and the customer number field in a separate transaction data extension. A Decision Split configured for ' both Contact and Journey Data ' allows the split ' s filter criteria to reference attributes from Attribute Groups configured in Contact Builder (relational, contact-scoped data such as the transaction/redemption DE) alongside the journey ' s own entry-source attributes in a single condition — precisely the cross-DE, field-to-field comparison described. A split using only Journey Data (A) is confined to the fields present in the entry source itself and cannot reach into the separate transactional data extension. A split using only Contact Data (B) can query related, contact-linked data but loses the ability to reference journey-specific attributes in the same condition, which the scenario also implies is needed for the comparison. Einstein Split (C) is a machine-learning-driven random/predictive path allocator, not a rules-based attribute comparison tool, and is unrelated to this use case. Consultants should recognize ' compare a field in the entry source against a field in a related data extension ' as the signature use case for the combined Contact-and-Journey Data Decision Split. Reference: Journey Builder Decision Split, Data Designer Relationships, Attribute Groups.

Journey Builder performs best when the heavy lifting of segmentation happens before contacts are evaluated for entry, not inside the journey itself. Performing large-scale segmentation in Automation Studio (B) offloads CPU-intensive filtering to SQL Query Activities, which are optimized for bulk processing across millions of rows, rather than relying on Journey Builder ' s per-contact entry evaluation engine. Creating a prefiltered, dedicated sendable copy of the source data extension for each journey (D) avoids lock contention and re-evaluation overhead that occurs when multiple journeys share a single large entry source, and it ensures each journey only processes the exact population it needs. Option A is a valid best practice for decision-split performance generally, but Attribute Groups govern contact-level data joins for Contact Data-based splits and does not address entry-source scale at 10 million records specifically — it ' s a secondary optimization, not one of the two primary levers for this scenario. Option C describes filtering directly on the entry source Data Extension, which still forces Journey Builder to evaluate the full underlying population against the filter at entry time, rather than pre-segmenting it, and is less efficient than doing that work in Automation Studio beforehand. Reference: Journey Builder Performance Best Practices, Automation Studio Segmentation.