Overview
Intake and Interaction Channels is 13% of the exam and has the most objectives in the guide, five. The first three are about channel design: recommending intake approaches such as Email-to-Case, Web-to-Case, Experience Cloud forms and phone; discussing use cases for interaction channels such as Messaging for In-App and Web, enhanced messaging on WhatsApp, SMS and Facebook Messenger, and Service Cloud Voice; and applying best practices when configuring them, especially Omni-Channel routing, capacity and presence.
The last two objectives are agentic. You recommend where AI agents belong in the service process, for example order status, returns or password resets, and where people must stay involved. You also explain Agentforce Service Agent capabilities and setup: the agent user, topics and actions, grounding in Knowledge, deployment to messaging channels, testing, and escalation to a rep through an Omni-Channel flow.
How it's tested
Questions describe customer contact patterns, team skills and business goals. You pick the intake channel, interaction channel, routing model or AI agent design that fits. Watch for volume, real-time needs, required skills and the point at which a person must take over from automation.
-
Pick the channel
Customers contact the business in a stated way, and you choose Email-to-Case, Web-to-Case, Messaging, Voice or a site form.
-
Route the work
Work must reach the right rep, and you choose queue-based, skills-based or attribute-based routing with suitable capacity.
-
Place the AI agent
An AI agent is proposed, and you decide which requests it handles, which actions it needs and how it escalates.
A bank wants an AI agent on its mobile app for card questions. Balance checks and card freezes are routine and well defined, so Agentforce Service Agent can handle them with actions grounded in Salesforce data. Fraud disputes involve judgment and regulation, so the agent should escalate them through an Omni-Channel flow to a skilled rep with the conversation transcript attached.
Objectives 5
Official objectives from the Salesforce outline, in outline order.
5.1 Given business process requirements, recommend the appropriate approach to intake channels and their design.Choosing how cases enter Salesforce.
- What it means
- Intake channels create cases or work items: Email-to-Case, Web-to-Case, Experience Cloud case forms, phone through Voice or CTI, and messaging sessions that may become cases.
- What to understand
- On-Demand Email-to-Case processes email inside Salesforce with threading by Lightning threading tokens or message headers, routing addresses and auto-response rules. Web-to-Case suits simple anonymous forms with a daily limit of 5,000 cases. Experience Cloud forms suit authenticated users and allow richer data. Design for required fields, deduplication of repeated emails, attachments, spam handling and case assignment rules applied at intake.
- How it shows up
- A client describes how customers reach them, and you recommend the intake channel and its key settings.
- Common confusion
- Recommending Web-to-Case for authenticated portal users who should log in and track cases.
5.2 Given a set of business goals, discuss the use cases and functionality for proposed interaction channels.Matching real-time and asynchronous channels to goals.
- What it means
- Interaction channels are where reps or AI agents converse with customers: Messaging for In-App and Web, enhanced messaging on WhatsApp, SMS, Facebook Messenger and Apple Messages for Business, Service Cloud Voice and email.
- What to understand
- Messaging is persistent and asynchronous, so customers can leave and return, and reps can handle several sessions at once. Voice suits urgent or complex issues and integrates telephony with transcription through Amazon Connect or partner providers. Email suits detailed, non-urgent requests. Match channels to customer demographics, urgency, concurrency and cost, and avoid launching channels the team cannot staff.
- How it shows up
- Business goals are given, and you pick the interaction channel that delivers them.
- Common confusion
- Treating messaging like phone, ignoring that sessions persist and reps can work several at once.
5.3 Assess the design considerations and best practices when configuring an interaction channel solution.Routing, capacity, presence and supervision.
- What it means
- Most channels depend on Omni-Channel, which pushes work to available reps based on routing configuration, capacity and presence status.
- What to understand
- Know queue-based routing, skills-based routing and Omni-Channel Flow for attribute-based or complex logic. Configure service channels, routing configurations with priority and units of capacity, presence configurations and statuses, and status-based or tab-based capacity. Plan overflow and secondary routing, business hours for channels, and Omni Supervisor for real-time monitoring. Consider pre-chat forms, consent, transcripts and conversation context passed between bots and reps.
- How it shows up
- A routing or capacity issue is described, and you choose the configuration change that fixes it.
- Common confusion
- Using queue-based routing when the scenario requires matching specific rep skills or languages.
5.4 Given a set of business requirements, recommend an appropriate use for AI agents in the customer service process.Deciding what an AI agent should and should not do.
- What it means
- AI agents such as Agentforce Service Agent hold natural conversations, reason over requests and take actions. Placing them well means matching their strengths to suitable requests and keeping humans involved where needed.
- What to understand
- Good candidates are high-volume, well-defined requests with clear data and actions: order status, appointment changes, returns, password resets and common questions grounded in Knowledge. Keep people involved for complex, emotional, high-value or regulated cases, and design escalation with context. Consider guardrails, data access through the agent user, success measures such as containment and resolution, and agent-assist features like Service Replies for reps.
- How it shows up
- A use case is described, and you decide whether an AI agent, rep assistance or a human-only path fits.
- Common confusion
- Assuming the goal is maximum containment rather than resolved issues and satisfied customers.
5.5 Explain agentic service capabilities, use cases, and configuration processes.Building and deploying Agentforce Service Agent.
- What it means
- Agentforce Service Agent is configured in Agentforce Builder from a service agent template, then connected to channels such as Messaging for In-App and Web.
- What to understand
- Configuration includes the agent user and its permission sets, topics with classification descriptions, scope and instructions, actions (standard actions such as answering with Knowledge, plus flows, Apex or prompt templates), and data access. Connect the agent to a messaging channel and an Omni-Channel flow, define escalation through an outbound route to a queue, test in Agentforce Testing Center, then activate. Know adjacent capabilities such as Service Replies, Work Summaries and Einstein Bots as the legacy rules-based option.
- How it shows up
- A setup step is missing or wrong, and you identify which configuration element fixes the agent's behavior.
- Common confusion
- Forgetting the agent user's permissions, so actions fail even though topics are configured correctly.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
On-Demand Email-to-Case
Email processed into cases inside Salesforce.
- Remember
- Uses routing addresses and threading.
- Tested as
- Email intake scenarios.
- Trap
- Choosing the legacy agent-based version.
-
Web-to-Case
HTML form that creates cases.
- Remember
- Daily limit of 5,000 cases.
- Tested as
- Simple anonymous intake.
- Trap
- Using it for authenticated customers.
-
Messaging for In-App and Web
Persistent messaging for websites and apps.
- Remember
- Successor to legacy Chat.
- Tested as
- Web conversation scenarios.
- Trap
- Recommending legacy Chat for new projects.
-
Enhanced messaging channels
WhatsApp, SMS, Messenger, Apple Messages.
- Remember
- Asynchronous, sessions persist.
- Tested as
- Mobile-first customers.
- Trap
- Ignoring consent and opt-in rules.
-
Service Cloud Voice
Telephony built into the console.
- Remember
- Amazon Connect or partner telephony, with transcripts.
- Tested as
- Voice scenarios.
- Trap
- Confusing it with Open CTI only.
-
Skills-based routing
Routes to reps with required skills.
- Remember
- Supports language and product skills.
- Tested as
- Matching specific expertise.
- Trap
- Using queues when skills matter.
-
Omni-Channel Flow
Flow that routes work with custom logic.
- Remember
- Attribute-based routing and AI agent handoff.
- Tested as
- Complex routing.
- Trap
- Writing Apex routing instead.
-
Capacity model
Units of work a rep can hold.
- Remember
- Tab-based or status-based.
- Tested as
- Concurrency scenarios.
- Trap
- Giving voice and messaging equal weight.
-
Omni Supervisor
Real-time view of reps and queues.
- Remember
- Monitor, reassign and coach live.
- Tested as
- Supervisor scenarios.
- Trap
- Using static reports for live monitoring.
-
Agentforce Service Agent
Autonomous AI agent for customers.
- Remember
- Topics, actions, agent user, channel connection.
- Tested as
- Agentic scenarios.
- Trap
- Missing the escalation route.
-
Agent user
User record the AI agent runs as.
- Remember
- Its permissions limit data and actions.
- Tested as
- Failing action scenarios.
- Trap
- Granting broad admin access.
-
Escalation to rep
Handoff from AI agent to a person.
- Remember
- Outbound Omni-Channel flow route to a queue.
- Tested as
- Handoff design.
- Trap
- Losing conversation context.
-
Containment rate
Share of conversations finished without a rep.
- Remember
- Pair with resolution and CSAT.
- Tested as
- AI agent success measures.
- Trap
- Optimizing it alone.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
Routing models
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Queue-based | Work routed from queues to members | Simple, team based | Ignores individual skills |
| Skills-based | Work matched to rep skills | Language or product expertise | Over-detailed skill lists |
| Omni-Channel Flow | Flow decides routing | Custom logic and AI handoff | Building logic flows cannot need |
Channel characteristics
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Voice | Real-time phone | One at a time, urgent issues | Using it for simple status checks |
| Messaging | Persistent asynchronous chat | Several sessions per rep | Treating sessions as one-off |
| Asynchronous written requests | Detailed, non-urgent | Promising instant replies |
AI in service
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Agentforce Service Agent | AI agent talking to customers | Acts autonomously | No human fallback |
| Service Replies | Drafted responses for reps | Rep reviews and sends | Expecting it to answer customers alone |
| Einstein Bots | Rules-based dialog bot | Scripted paths, legacy option | Choosing it for open-ended requests |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
Maximum containment
- What candidates pick
- Configure the AI agent to avoid escalations.
- Why it looks right
- Containment looks like savings.
- Why it's wrong
- Customers stuck in automation abandon or call back; success means resolved issues and satisfied customers.
Clue to watch for Complaints about not reaching a person.
-
Legacy Chat
- What candidates pick
- Set up legacy Chat for a new website.
- Why it looks right
- It is familiar.
- Why it's wrong
- Messaging for In-App and Web is the current, persistent channel and the one Agentforce connects to.
Clue to watch for A new web conversation channel is requested.
-
Queues for skills
- What candidates pick
- Create a queue per language.
- Why it looks right
- Queues are simple to build.
- Why it's wrong
- Queue sprawl grows quickly; skills-based routing matches language and product skills more flexibly.
Clue to watch for Several languages and products combine.
-
All channels at once
- What candidates pick
- Launch voice, messaging and social together.
- Why it looks right
- Customers want choice.
- Why it's wrong
- Channels without staffing and capacity planning raise wait times and abandonment.
Clue to watch for The team size is fixed.
-
Topic fixes permissions
- What candidates pick
- Rewrite topic instructions when an action fails.
- Why it looks right
- Instructions guide the agent.
- Why it's wrong
- If the agent user lacks access, the action fails regardless of instructions.
Clue to watch for Errors mention access or missing records.
-
Bot for everything
- What candidates pick
- Use a scripted bot for open-ended questions.
- Why it looks right
- Bots are cheaper.
- Why it's wrong
- Scripted dialogs handle fixed paths; open-ended requests need a reasoning AI agent or a rep.
Clue to watch for Customer questions vary widely.
Answer framework
Use these steps for intake, channel, routing or AI agent questions.
- Identify how customers reach the business today, and how urgent, complex and frequent their requests are.
- Choose the intake channel and its settings, such as Email-to-Case routing addresses, threading and auto-responses, or authenticated forms on a site.
- Choose interaction channels the team can staff, then route with queues, skills or an Omni-Channel flow and set capacity.
- Decide which requests an AI agent should handle, based on volume, clarity, available actions and regulatory risk.
- Configure the AI agent's user, permissions, topics, actions and Knowledge grounding, test it in Testing Center, and design escalation to a skilled rep that keeps the full conversation context.
Practice question
Want more like this? The Service-Con-201 practice test has questions for every domain.
Checklist
0 of 13
Tick each one you can do without notes. Your ticks stay in this browser.
Study strategy
Split this section into channels and agents. For channels, use a Developer org to enable Omni-Channel, create a service channel for cases, a queue, a routing configuration and presence statuses, and route an Email-to-Case message to yourself. Add skills-based routing for one skill and watch the difference. Set up Messaging for In-App and Web if your org supports it, and read the Service Cloud Voice overview for telephony options.
For agents, complete the Trailhead modules on Agentforce Service Agent and Agentforce Builder. In an org with Agentforce enabled, create a service agent from the template, review its topics and actions, check the agent user's permission sets, connect it to messaging and test an escalation. Write a list of ten service requests and mark each AI agent, rep with AI assistance, or human only.
High-priority ideas are routing models, capacity and escalation design. Underestimated are objectives 5.4 and 5.5, which older Service Cloud material does not cover.
FAQ
How much is Intake and Interaction Channels worth on the exam?
Intake and Interaction Channels is 13% of the exam per third-party copies of the current guide, around eight scored questions as a planning estimate, across five objectives.
Why does this section have five objectives at only 13%?
Two objectives cover AI agents, added to the guide in the Agentforce era. Each objective therefore contributes only one or two questions.
Do I need to configure Service Cloud Voice?
Know its use cases, telephony options and how it fits the console and Omni-Channel. Detailed telephony setup is beyond the exam.
Is Einstein Bots still tested?
It may appear as the scripted, rules-based alternative. For open-ended customer conversations, Agentforce Service Agent is the current answer.
What is the most common AI agent mistake in scenarios?
Designing without escalation. The expected answer usually includes a route to a skilled rep that keeps the conversation history.
Does the Agentforce Specialist exam cover the same agent content?
It overlaps on topics and actions, but this exam focuses on service use cases, channel connection and handoff, not deep agent building.
Other CRM & Business Applications certifications
Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.
- HubSpot Inbound Marketing Certification Study guide
- Salesforce Certified Platform Developer I Study guide
- Google Ads Search Certification Study guide
- ServiceNow Certified System Administrator Study guide
- Salesforce Certified Platform Administrator Study guide
- ServiceNow Certified Application Developer Study guide
Test your Domain 5 knowledge
You have covered intake channels, interaction channels, Omni-Channel routing and Agentforce Service Agent. Check whether you can design routing and AI handoff for a scenario.
Keep going
- Training Training Built on your Weak areas
- Practice test Questions by domain, with explanations
- Mock exam Full length, timed like the real exam
- Readiness test Find your weakest domain in minutes
- Cheat sheet Key terms and formulas on one page
- Ask the Service-Con-201 AI tutor Get any concept explained your way