Salesforce Service Cloud Consultant cheat sheet

Service Cloud editions in ascending capability: Essentials, Pro Suite, Enterprise, Unlimited; console apps, Omni-Channel, entitlements, and Knowledge are licensed at Enterprise and above.

Field Service is a separate license and product; Service Cloud does not include work orders, service resources, or service appointments.

Core service metrics: CSAT = positive survey responses / total responses; FCR = cases closed on first contact / cases closed; AHT = total handle time / contacts handled; SLA attainment = milestones met / milestones due.

Channels to inventory during discovery: phone/CTI, email, Messaging for In-App and Web, SMS, social, Web-to-Case forms, self-service portal, and legacy chat.

Customer-facing self-service runs on Experience Cloud sites with Customer Community or Customer Community Plus licenses, not full Salesforce licenses.

Service Cloud Voice is delivered with a telephony partner (Amazon Connect, Five9, Genesys, NICE) and adds voice call records, transcription, and real-time coaching.

Knowledge, Digital Engagement, and Einstein/Agentforce capabilities are frequently edition-gated or credit-metered; confirm entitlements before solution design.

Discovery must quantify current state: monthly case volume, channel mix, AHT, FCR, backlog, and peak concurrency drive routing, capacity, and staffing design.

Salesforce Well-Architected pillars: Trusted (secure and compliant), Easy (intuitive and fast to adopt), Adaptable (scalable and low maintenance).

Prefer standard Case with record types and support processes over custom objects; each custom object adds maintenance, reporting, and integration cost.

Sandbox types: Developer and Developer Pro (metadata plus small data), Partial Copy (sampling via template), Full Copy (all data, 29-day refresh) - use Full Copy for UAT and load testing.

Load tooling by volume: Data Import Wizard for under 50,000 records and standard objects; Data Loader or Bulk API for larger or automated loads.

A requirements traceability matrix maps every requirement to a configuration component and a test case; sign-off gates each phase transition.

Phased rollout pattern: pilot one channel or business unit, measure against the baseline, then expand; full cutover on day one is the highest-risk option.

AppExchange managed packages (CTI, e-signature, surveys, telephony) are usually cheaper than custom builds when functional fit is roughly 80% or better.

Support processes control which Case Status picklist values are selectable for a record type; each record type is tied to exactly one support process.

Entitlements match a case to an Entitlement Process, which defines milestones whose start and completion criteria reference entitlement, case, and milestone fields.

Milestone actions fire on milestone completion or violation: email alert, field update, outbound message, quick action, or flow.

On case creation, assignment rules and auto-response rules run after save; Omni-Channel routes the case only when it lands in an Omni-Channel-enabled queue for that object.

The default case owner setting (a user or a queue) applies whenever no assignment rule or flow sets an owner.

Sharing design: set Case organization-wide default to Private, then open access through role hierarchy, criteria-based sharing rules, queues, and case teams; use Apex managed sharing for dynamic rules.

Case teams grant case access to users outside the owner's role hierarchy through predefined case team roles and members.

Validation rules enforce closure discipline (required closure reason and closed date); field history tracking supplies the audit trail most service programs require.

Lightning Knowledge article types map to record types; each article type carries its own page layout, validation rules, and available data categories.

Publishing creates a new KnowledgeArticleVersion; Draft, Published, and Archived states belong to the version, not to the article as a whole.

Data categories control article visibility per channel and per profile and populate the filter sidebar in Knowledge search.

Search relevance is driven by promoted search terms, title and body matches, data category filters, and synonyms; stop words are ignored.

In the console, suggested articles surface from case field matches; attaching an article creates a CaseArticle record that supports reuse reporting.

Approval processes can gate the Draft-to-Published transition, and article validation rules are evaluated at publish time.

A public knowledge base on an Experience Cloud site requires published articles plus channel and data category visibility; drafts are never shown to customers.

Article voting and feedback supply quality signals; report on KnowledgeArticleVersion and CaseArticle to measure reuse and deflection.

Messaging sessions are asynchronous and persist across sessions and devices; legacy Chat is synchronous and ends when the window closes.

Omni-Channel routing models: Least Active (fewest open work items), Most Available (largest remaining capacity), Most Productive (most items completed).

Presence statuses and capacity are assigned per user or per profile, and capacity can be weighted differently per work item type.

Web-to-Case and Web-to-Lead default to 5,000 requests per day per org; request an increase before high-traffic campaigns.

Email-to-Case requires a customer-hosted email agent; On-Demand Email-to-Case uses Salesforce servers and is simpler but has attachment size limits.

Service Cloud Voice requires a partner telephony integration; transcripts and conversation insights can populate case fields and quality scores.

Omni-Channel supervisors monitor wait times, agent status, and backlog in real time and can reassign or flag work items.

Case statuses come from the support process; closed cases can be protected from edits with a validation rule or closed-case behavior settings.

Case assignment rules evaluate in order and stop at the first matching entry; each entry assigns to a user or a queue.

Escalation rules evaluate case age against business hours and can reassign, notify, or update a field; they are independent of entitlement milestones.

Case merge supports up to three cases: the master retains its own field values, inherits related records, and merged cases become read-only.

Parent-child cases model incident hierarchies; roll-up summary fields are not native to Case, so use Apex or a custom rollup for child counts.

Case Origin, Priority, and Type are standard picklists; fix these reporting dimensions before go-live because later changes break historical reporting.

Case comments are internal by default and become customer-visible only when published to a portal or Experience Cloud site.

A case with no matching entitlement produces no milestones and is excluded from SLA attainment reporting.

Service Console apps use split view (list plus detail), subtabs, a highlights panel, and a utility bar; record pages are configured in Lightning App Builder.

Utility bar items persist across the app: Omni-Channel, History, Notes, Macros, CTI softphone, and Flow.

Macros bundle quick actions, field updates, emails, and flows into a single click and can run immediately or after agent review.

The Highlights panel displays key case fields and actions, and field visibility can be set conditionally per record page.

The docked composer keeps email and messaging composition visible while the agent navigates between records.

Keyboard shortcuts, pinned list views, and the recent-items tab reduce navigation time and are common adoption levers.

The Omni-Channel supervisor component shows real-time agent and queue metrics and can be placed in the utility bar or on a dedicated app page.

Standard case reports show current field values only; use Case History, Case Snapshot, or Case Milestone reporting for time-based analysis.

Reporting snapshots copy case data into a custom object on a schedule, enabling trend analysis of status, owner, and backlog by day or week.

Row duplication appears when a case joins to a child object with multiple rows; use a custom report type with 'with or without' and count distinct rows where possible.

AHT = total handle time / contacts handled; FCR = cases closed on first contact / cases closed; both should be sliced by channel and case type.

SLA attainment = milestones met / milestones due, reported from Case Milestone filtered on completion and violation dates.

Dynamic dashboards run with the viewing user's security settings; a fixed running user is required for org-wide executive service dashboards.

CRM Analytics service apps add prebuilt datasets for case, channel, and knowledge analytics, plus predicted case escalation.

Omni-Channel analytics uses AgentWork and presence data to measure wait time, transfer rate, and capacity utilization.

Einstein for Service includes case classification, case wrap-up field prediction, reply recommendations, article recommendations, and service replies.

Agentforce agents are assembled from topics (scope plus instructions), actions (flows, Apex, prompt templates), and a reasoning engine that selects the topic at runtime.

The Einstein Trust Layer enforces zero data retention by the model provider, data masking, toxicity detection, and an audit trail of prompts and responses.

Grounding matters: prompt templates retrieve Knowledge articles and CRM records as context; ungrounded prompts raise hallucination risk.

Agentforce Service Agent handles customer-facing self-service on messaging and web channels, while the agent-facing assistant supports reps inside the console.

Generative AI features consume Flex Credits; enable Einstein and assign the required permission sets before building anything.

Test agents in the Agentforce Testing Center with representative utterances, and scope each topic narrowly to prevent misrouting.

Einstein Case Routing can predict the best agent or queue from historical case outcomes, complementing skills-based Omni-Channel routing.

Case relates to Account, Contact, Entitlement, Asset, Product, and ServiceContract, and through junction records to Knowledge (CaseArticle) and case teams.

Key standard objects: Case, CaseComment, CaseHistory, CaseTeamMember, CaseMilestone, Entitlement, ServiceContract, Asset, WorkOrder, KnowledgeArticleVersion.

Ownership skew - one user owning a disproportionate share of records - degrades sharing recalculation and query performance; spread ownership across queues and roles.

Load order: Account, Contact, Entitlement/ServiceContract, Case, then related records; use external ID fields with upsert to make loads repeatable.

Duplicate and matching rules block inserts; deactivate them or add bypass conditions for integration users during bulk loads.

Case organization-wide default of Private is the standard starting point; open access with role hierarchy, sharing rules, queues, and case teams.

At large data volumes, archive closed cases to a big object or external store and report on aggregates rather than raw rows.

Field history tracking is capped (20 fields per object by default), must be enabled before go-live, and is never retroactive.