Welcome to your trial of Zendesk Copilot! We can’t wait for you to explore the powerful features that will enhance your customer support service. With Copilot, you will streamline your operations, improve agent productivity, and elevate customer satisfaction. We know adopting AI can be overwhelming, so this guide will help you take it one step at a time and discover the tangible impact that AI can have on your business.
- If you’re new to our approach to Copilot, see About Zendesk Copilot.
- If you’re a power user, dive deeper into Resources for Zendesk Copilot.
This article provides a map to maximizing your Copilot trial. Topics include:
- (Optional) Week 0: Optimize your knowledge base for AI
- Week 1: Get started with intelligent triage
- Week 2: Get started with agent copilot
- Week 3: Launch your AI agents
- Week 4: AI dashboards and crafting your pitch
- Week 5: Wrap up
(Optional) Week 0: Optimize your knowledge base for AI
A well-organized knowledge base is crucial for helping users find answers and troubleshoot issues. Creating one is just the beginning: Your knowledge base must be tailored to meet user needs and continuously updated to remain relevant.
To optimize
- Know your goals and users. Start by defining what you want to achieve with your knowledge base and who will be using it. Understanding your audience helps you create content that meets their needs.
- Identify common issues. Look at frequent support tickets and popular macros to spot recurring problems. This insight will guide you in creating relevant content. Organize your information clearly, and write in a way that’s easy for users to understand.
- Make content easy to find. Use tags and labels to boost the visibility of your articles. Implementing semantic search can also help users quickly locate the information they need.
- Keep improving. Track how your knowledge base is performing by monitoring article views, feedback, and comments. This will help you spot areas for improvement and ensure your content stays useful over time.
Additional resources
Week 1: Get started with intelligent triage
Intelligent triage uses AI to automatically detect why a customer is reaching out (their intent), in which language the ticket is written, and whether the customer's message is positive or negative (its sentiment). You can also use entity detection to detect specific products and other key words (entities) and automate workflows based on that information.
Why this step? Setting up intelligent triage will allow you to understand your CS operations and act on insights faster, automate top recurring requests, eliminate manual triage and boost operational efficiency, and streamline your operations by identifying and acting fast on urgent issues.
Week 1, Day 1: Set yourself up for success and configure intelligent triage
-
Set goals for a successful trial. Decide which metrics are important to you and your leadership team, and configure your trial in a way that will meet your goals. Zendesk recommends orienting around the following KPIs:
- Customer sentiment
- Agent efficiency. First Response Time and Agent Handle Time
- Customer experience analysis. Customer interactions and retention
- Block your calendar. Set aside time halfway through the trial (end of week 2) and near the end of the trial (end of week 4) to track performance and make sure you are aligning to your stakeholders’ goals.
-
Notify your agents about the trial, especially the agent copilot features. Use our template to encourage your employees to try these features as soon as possible.
Note: We dedicate Week 2 to using agent features.
-
Get set up. Configure intelligent triage.
- Verify that intelligent triage is turned on in your account and read how to configure intelligent triage settings.
- Watch a demo for a visual walkthrough.
Week 1, Day 2: Understand your operations and automate top recurring requests
- Understand your operations. Discover top incoming request topics (intents), and shift teams’ priorities in real time to identify knowledge gaps and resolve incoming requests faster. For example, you could group together similar intents for a specific team to streamline operational efficiency.
- Automate top recurring requests. Allow customers to self-serve across common request types (intents) where documentation is readily available, like return policy, profile update, or subscription cancellation. For example, ensure you have the right knowledge base articles to deflect tickets.
Week 1, Day 3: Eliminate manual triage
Display incoming requests to specific agent or agent groups based on the intent, custom entity (entity detection), sentiment, and language so the tickets are being tackled by the right agents. Improve the efficiency of workflows by routing the ticket to the correct approval flow based on the intent. You can assign incoming requests to the right agent or agent group automatically.
- Use custom entity (entity detection) to detect products mentioned in tickets and show them in a view for agents.
- Use the ticket’s automatically detected language to assign the ticket to agents who speak that language.
- Improve workflow efficiency by routing tickets to the correct approval flow based on intent.
Week 1, Day 4: Streamline operations
- Identify escalations easily and customize SLA rules based on intent and sentiment to prioritize requests.
- Automatically request missing information for a request type where the mandatory field is missing and reduce multiple-touch conversations.
- Collect critical customer information proactively and update ticket information based on intent.
Week 1, Day 5: Track your progress
Use the Zendesk intelligent triage dashboard to track your progress. The Explore intelligent triage dashboard provides valuable insights into your tickets, which can be used to improve your workflow and determine how effective your support is. For example, you can see why your users are reaching out (with intents) and how they feel (with sentiment).
Week 2: Get started with agent copilot
Why this step? Zendesk's agent copilot empowers customer support representatives by automatically generating personalized, contextually relevant draft responses to reduce response times, improve consistency, and allow agents to focus on more complex problem-solving and customer relationship building. These features provide immediate value to agents by reducing agent handle time and aiding agents in composing friendly and professional responses in seconds.
Week 2, Day 1: Familiarize yourself with agent copilot
- Watch the demo .
-
Read our agent copilot resource to see copilot’s potential.
Note: Copilot offers many capabilities, but for this trial, we recommend setting up two features: enhance writing and ticket summaries.
Week 2, Day 2: Notify your team about agent copilot features
Notify your team on the copilot features they can now access if you haven’t already. View our best practices guide here to help ease them into the transition of using AI in their role. We know it can be overwhelming, and our best practices will help assuage their emotions and keep your trial on track
Week 2, Days 3 and 4: Discover more agent copilot features
- Learn about Enhance writing.
- Learn about Ticket summaries, including Talk ticket summaries.
Week 2, Day 5: Nearly halfway
- Track metrics with our Generative AI Agent Tools dashboard.
- Check in with your goals, team, and dashboards.
-
Learn. Take agent copilot to the next level by enabling auto assist to ensure adherence to business procedures and even complete actions on behalf of your agents.
- Read about Configuring auto assist.
- Watch a quick demo of auto assist and procedures.
- Read how to create and manage procedures.
Week 3: Launch your AI agents
Why this step? When people think of AI, they often think of bots, which we call AI agents, first. Since you’ve already optimized your knowledge base (see Week 0), these AI agents will pull in the most accurate data from that knowledge base when responding to customers on your behalf. AI agents are included in Zendesk Suite.
Week 3, Day 1: Launch your AI agents
Launch your AI agents in just a few steps, without any training required.
- With generative replies, AI agents use knowledge from your existing help center to quickly craft accurate, conversational responses to your customers’ questions. This helps automate repetitive requests, enhancing your CX and freeing up your teams to focus on more sensitive or complex tasks. To get started:
- Name your AI agent and choose its tone of voice.
- Connect your brand and knowledge base.
- Create and test your AI agent. Then you’re ready to publish it and start automating.
Week 3, Day 2: Set up AI agent personas
Set up AI agent personas to ensure your generative replies are consistently representing your brand’s voice and tone.
To set up personas
-
Make sure your AI agents are a direct reflection of your brand and are delivered in a consistent tone with AI agent personas. There are two key components of personas: tone of voice and business profile.
- Tone of voice. You already chose a tone of voice when setting up your AI agent. You can change this setting at any time.
- Business profile. Guide your AI agent responses by populating the business profile to ensure it understands your business and its purpose.
- (Optional) Adjust reply length to specify if you’d like your AI agent to keep responses snappy, or if you prefer a longer response.
Week 3, Day 3: Fill in knowledge gaps
- Fill in knowledge gaps. With time, new questions will pop up from your customers. Discover which common requests are not successfully resolved by your AI agent, and enhance your knowledge base or AI agent responses to fill in any gaps.
Week 3, Days 4 and 5: Track progress and learn more
- Track progress. Analyze your automated resolutions data, and check in with your team to see how AI agents are improving their workday as a front line defense.
- Learn more. The AI agent you've launched in trial is designed to help you instantly start automating customer requests using cutting-edge generative AI technology. If you're interested in additional customization and more advanced use cases, like enabling your AI agent to fully automate complex tasks, check out our AI agents – Advanced offering.
Week 4: AI dashboards and crafting your pitch
Why this step? Zendesk's AI-powered dashboards and reporting tools provide you with actionable insights to measure the effectiveness of this trial now – and in the future, to continue automatically analyzing customer interaction data, identifying performance trends, predicting potential bottlenecks, and offering predictive analytics that enable more strategic decision-making and continuous improvement of customer support operations.
Week 4, Day 1: Evaluate your trial
Now is the time to measure your trial’s success and communicate that success with key stakeholders. Compare your KPIs and stakeholder goals to your quantitative trial data and your team’s qualitative experience. You can leverage these dashboards:
- Review the intelligent triage dashboard.
-
Review the Generative AI Agent Tools dashboard.
- Watch Gen AI Agent tools.
- Review your automated resolutions data to track deflections with AI agents.
Week 4, Days 2 to 5: Prepare and present your pitch
Prepare and present your pitch to key stakeholders and cross-functional partners.
-
Utilize our best practices to build your pitch to key stakeholders:
- Highlight metrics from the trial in comparison to your stakeholder goals.
- Introduce anecdotal feedback from your agents, highlighting specific quotes.
- Show metrics and feedback that your stakeholders didn’t initially consider.
- Leave them with your list of goals for Copilot if you were to purchase.
- Ask them for any questions to consider when deciding to convert.
- Customer stories. See how others are using Copilot to provide better customer service with AI. Check out our customer stories and filter to find one that matches your company’s use case as further validation for your pitch.
- Future proof. Remind stakeholders that this trial was meant to give you a taste of the power of Copilot, but there is so much more you can do. Check out other features, like setting up shortcuts, additional copilot features, and our AI agents – Advanced add-on for example.
Week 5: Wrap up
Take one last look at your data in comparison to your goals. Export any final metrics you may need in making your pitch to stakeholders.
Your trial ends after 30 days. We hope you had a great trial experience and that AI has made your work day easier, whether it's optimizing workflows, boosting agent efficiency, or deflecting customer requests. To purchase Copilot, see Buying the Copilot add-on.
0 comments