50 Tips to Get More Out of Dynamics 365 Customer Service
50 practical tips for Dynamics 365 Customer Service, covering workspace productivity, omnichannel routing, Copilot, AI agents, and contact center administration.
At Community Summit 2026 in Nashville, Abraham Saldana and I presented 50 Tips in 50 Minutes: Dynamics 365 Customer Service.
Our goal was simple: show organizations how to get more out of the technology they already own while exploring the rapidly expanding capabilities of Microsoft’s service platform.
Dynamics 365 Customer Service has evolved considerably. What started as a traditional case management platform now includes intelligent routing, configurable agent workspaces, generative AI, autonomous agents, and increasingly sophisticated contact center operations.
Yet many organizations are still using a fraction of what’s available.
Here are 50 practical tips to help change that.
1. Make the Workspace Work for Your Team
The best productivity improvements aren’t necessarily complicated customizations. Often, they’re configuration changes that eliminate unnecessary steps.
Tips 1–10: Workspace and Productivity
1. Use experience profiles instead of building custom apps.
Experience profiles let you configure different workspaces for different teams without maintaining separate applications. Each team can have its own session templates, channels, and productivity tools.
2. Build session templates for each channel.
A representative handling WhatsApp shouldn’t necessarily see the same workspace as someone handling a phone call. Configure anchor tabs and application tabs to provide relevant context automatically.
3. Enable the inbox for high-volume work.
The inbox brings cases, conversations, and activities together in a unified view. It’s particularly useful for asynchronous channels such as SMS, WhatsApp, and persistent chat.
4. Automate repetitive actions with macros.
Macros can open forms, populate fields, send messages, and execute multiple actions in sequence. Start with common workflows that require repetitive clicks.
5. Refresh session context in your macros.
Session context is cached. When a macro updates a record and needs to reuse those updated values, include the Refresh Session Context action.
6. Guide representatives with agent scripts.
Scripts provide standardized processes while allowing macros to execute individual steps. You can configure default scripts through session templates.
7. Replace retired smart assist suggestions with Copilot.
Microsoft retired the older smart assist case and knowledge suggestions in June 2025. Copilot provides knowledge-grounded answers and suggested responses through the productivity pane.
8. Improve notification templates.
Give representatives useful information before they accept an interaction: customer name, queue, wait time, or other contextual details.
9. Bring external tools into application tabs.
Application tab templates can display dashboards, records, and supported third-party pages within an active session.
10. Use Microsoft’s workspace training resources.
The free Copilot Service workspace training covers experience profiles, macros, scripts, templates, and multisession functionality.
My Take
Before investing in custom development, spend time understanding experience profiles, session templates, and macros.
I’ve seen organizations build functionality that already exists in the platform. A small amount of thoughtful configuration can eliminate unnecessary complexity and reduce long-term maintenance.
2. Route Customers Intelligently
Routing isn’t just about getting a customer to the next available representative.
It’s about connecting the customer with the person best equipped to resolve their issue.
Tips 11–20: Omnichannel and Routing
11. Start with guided channel setup.
The Copilot Service admin center provides guided setup for channels, permissions, and initial routing configurations. It’s a useful starting point for new deployments.
12. Manage skills through the Skills Hub.
Skills-based routing helps match customer requirements with representative capabilities. Keep your skills manageable rather than creating hundreds of attributes nobody maintains.
13. Route conversations based on customer intent.
Intent-driven routing uses the reason a customer is contacting support to direct the conversation to an appropriate representative or team.
14. Transfer conversations using intent.
Representatives can use intent to identify the right workgroup, queue, or colleague instead of guessing which queue handles a particular issue.
15. Use consult-to-queue for warm transfers.
Rather than searching for a specific colleague, representatives can consult an entire queue and connect with someone available.
16. Pin frequently used transfer destinations.
Pinned contacts and queues reduce navigation during active conversations.
17. Tell customers how long they’ll wait.
Configure queue position and estimated wait-time updates for supported channels. Customers should know what to expect rather than waiting indefinitely.
18. Offer callbacks instead of keeping customers on hold.
Enhanced customer-first callbacks allow callers to leave the queue without remaining on the phone.
19. Explore proactive outbound engagement.
Outbound voice campaigns, preview dialing, and proactive SMS allow service teams to reach customers before an issue escalates.
20. Review Microsoft’s deprecation announcements.
Channel capabilities and forecasting tools continue to change. For example, Apple Messages for Business configuration was removed in September 2026, and legacy conversation forecasting is scheduled to end in October 2026.
My Take
Organizations spend considerable time optimizing how representatives handle interactions but often overlook how those interactions reach them.
Better routing can improve service without requiring additional headcount. Getting a conversation to the right person the first time matters.
3. Make Copilot Useful, Not Just Available
Turning on Copilot isn’t the same as implementing it successfully.
The real value comes from integrating AI into workflows, providing reliable knowledge, and measuring whether representatives actually benefit.
Tips 21–30: Copilot for Service Representatives
21. Use AI summaries at every handoff.
Copilot can summarize cases and conversations so customers don’t need to repeat information after a transfer.
22. Customize the information included in summaries.
Configure summaries around the fields and related records representatives actually need. Custom summaries can also support other record types.
23. Ground Copilot in trusted knowledge.
Review your knowledge articles and approved domains before enabling AI-powered answers. Outdated documentation produces unreliable guidance.
24. Draft customer emails with Copilot.
Allow Copilot to generate contextual responses while keeping representatives responsible for reviewing accuracy, tone, and content before sending.
25. Use proactive Copilot prompts.
Suggested prompts help representatives discover useful AI capabilities without needing to learn specific commands.
26. Explore Service Agent inside the workspace.
Service Agent lets representatives use natural language to retrieve information, summarize cases, draft communications, update records, and create notes.
27. Extend Service Agent into Microsoft 365.
Service Agent can work with Dataverse and Microsoft Graph, allowing authorized users to access service information from applications such as Teams and Outlook.
28. Surface customer sentiment on case records.
Sentiment indicators can help representatives identify potentially frustrated customers and adjust their approach.
29. Use real-time translation for multilingual support.
Supported chat channels can translate conversations, helping representatives assist customers across language barriers.
30. Track Copilot usage and feedback.
Capture representative feedback and AI interaction data. Analyze adoption, usefulness, and knowledge gaps rather than relying on anecdotal impressions.
My Take
AI adoption should be measured through outcomes, not licenses assigned.
Are representatives resolving cases faster? Is after-call work decreasing? Are customers being transferred less frequently?
If an AI feature doesn’t improve something measurable, organizations should question whether they’ve implemented it correctly or whether it’s the right capability for the problem.
4. Move from AI Assistance to AI Agents
One of the most significant changes in Dynamics 365 Customer Service is the shift from AI that suggests actions to AI that can execute them.
Microsoft’s service agents introduce new opportunities for case management, knowledge creation, routing, and resolution.
They also introduce new operational responsibilities.
Tips 31–40: AI Agents
31. Enable AI agents at the environment level.
Before configuring individual service agents, administrators must enable AI agents for the environment in the Power Platform admin center.
32. Let Case Management Agent populate case records.
The agent can create and update cases using information from conversations across supported channels, including voice, chat, and email.
33. Configure conditional case update rules.
Use conditions to control which additional fields the agent predicts. For example, a product defect might require a product, priority, and serial number.
34. Begin with semi-autonomous operation.
Start with AI-generated recommendations or actions that representatives approve. Expand autonomy when accuracy and reliability have been demonstrated.
35. Test agents using shadow mode.
Shadow mode allows organizations to evaluate agent predictions without allowing those predictions to change production records.
36. Automate incoming email classification.
Case Management Agent can classify emails into predefined categories and subcategories to support routing and downstream automation.
37. Automate follow-up and case closure.
Agents can send follow-up communications, create closure summaries, and close resolved cases according to configured rules.
38. Turn resolved cases into knowledge.
Customer Knowledge Management Agent can identify information from closed cases and draft or update knowledge articles.
39. Discover customer intents automatically.
Customer Intent Agent analyzes conversations to identify common reasons customers contact support and recommend intents.
40. Start with Microsoft’s service agent templates.
Templates such as After Hours Service, Case Categorization and Enrichment, and Automated Case Follow-ups provide starting points for agent implementations.
My Take
The most important question isn’t whether an AI agent can complete a task.
It’s whether the organization can understand, govern, and verify what the agent is doing.
I strongly recommend beginning with supervised execution and measurable acceptance criteria. Autonomous operation should be something an agent earns through testing rather than a default deployment choice.
5. Give Supervisors Better Visibility and Control
AI doesn’t eliminate the need for quality management or operational oversight.
If anything, increasing automation makes governance, measurement, and visibility more important.
Tips 41–50: Supervisors, Quality, and Administration
41. Evaluate more than a sample of customer interactions.
Quality Evaluation Agent can assess conversations and cases against defined evaluation criteria, expanding quality analysis beyond limited manual sampling.
42. Provide real-time coaching with Quality Assurance Agent.
The agent can analyze live conversations and provide guidance to representatives while alerting supervisors to potential compliance concerns.
43. Apply governance rules to outbound communications.
Quality Management capabilities can inspect outbound emails and attachments against policies, allowing organizations to log or block violations.
44. Use screen recordings to understand workflows.
Conversation transcripts explain what was said. Screen recordings provide visibility into how representatives navigated systems and completed their work.
45. Improve audio quality and recording compliance.
Deep noise suppression reduces background noise, while consent-based recording helps organizations respect customer recording preferences.
46. Use Service Operations Agent for administration.
Service Operations Agent assists administrators with configuration, onboarding, and operational setup tasks.
47. Manage users through Microsoft Entra security groups.
Security-group-based management simplifies onboarding, role assignment, and employee transitions.
48. Analyze conversation metrics by segment.
When interactions move between queues, segment-level metrics provide insight into each stage instead of attributing the entire interaction to the final queue.
49. Mask sensitive customer information.
Data masking helps protect sensitive information within conversations and transcripts, with security roles controlling access to redacted content.
50. Follow the AI at Work roadmap.
Microsoft moved away from traditional release plans in September 2026. The AI at Work roadmap now provides ongoing visibility into upcoming Dynamics 365, Power Platform, and Dataverse capabilities.
You can find it at aka.ms/AIatWorkRoadmap.