In our recent webinar, Unlocking AI Success with Zendesk Knowledge Base, we explored innovative strategies to elevate customer experience using AI. This session featured valuable insights from our expert panel, including members from the Zendesk CX, Product, and Community Teams. In this article you will find our event recording, Q&A Summary, resources from our event and our full event presentation deck.
Event Recording
Q&A Summary
With your enterprise plan, you can enhance your knowledge base using several AI-powered tools:
- Intent dashboard: Identify common customer issues to populate your knowledge base.
- Generative AI tools: Use features like ""Expand"" and ""Tone Shift"" to streamline article creation.
- Generative search: Improve information retrieval for agents and customers in your help center.
- AI agents: Leverage AI agents to assist with customer inquiries.
Resources: 5 strategies for building up your help center content for AI
Currently, Zendesk's generative AI responses utilize only text-based content. This includes information from articles, community posts, and federated sources when using our federated search.
For AI agents, responses are also based solely on text from articles and federated search sources.
We’re excited to share that support for embedded graphics, images, and text files (including PDFs) is on our roadmap and will be available later this year!
Yes, generative search results will eventually be available for anonymous users, meaning you won't need to sign in to access them. Initially, during the early access program, this feature will be limited to logged-in users. Once we officially release it to the public, all users will be able to enjoy generative search, including those who are not logged in.
While the organization of your knowledge base articles doesn’t directly affect how AI agents operate, it does play an important indirect role. A well-organized knowledge base allows you and your team to maintain accurate and up-to-date information, which ultimately benefits your AI agents.
When articles are organized effectively, it’s easier to identify gaps and ensure that your knowledge base remains healthy. Good organization is key to unlocking the full potential of AI in your support processes!
Resources: Best practices for creating a successful knowledge base
You can leverage your ticket data to strengthen your knowledge base and improve your AI capabilities. Here are some effective strategies:
- Identify common questions: Review frequent tickets to find recurring questions. If customers often ask the same thing, consider creating a helpful article to address it.
- Analyze categorized tickets: Look at categorized tickets to understand the main topics or issues your customers face. This will help ensure your knowledge base covers essential information.
- Utilize popular macros: If your team frequently uses the same responses, these can be turned into articles or refined for AI use.
- Transform agent responses: Take effective responses from your agents, copy them into an article, and use the generative AI expand feature to create a polished, comprehensive article.
Resources: 5 strategies for building up your help center content for AI
Currently, there is no direct replacement for the soon-to-be sunsetted content cues. While we don’t plan to create a separate feature for this, we are excited to introduce data insights and benchmarks within the product in a more contextual way.
Although we can’t specify what will replace content cues, we assure you that this year, new data points will be integrated closer to your workflow as an admin. This will help you take meaningful actions to enhance your knowledge base.
Resources: Announcing the removal of Guide Content Cues
When your knowledge base has gaps, it can significantly affect the performance of your AI. If important information is missing, AI-generated answers and AI agents may struggle to provide the assistance customers need.
This not only limits the effectiveness of your AI but also negatively impacts the overall customer experience. A strong knowledge base is essential for supporting both your agents and your customers, so it's important to keep it comprehensive and up to date.
Resources: Best practices for creating a successful knowledge base
AI agents locate information based on the content of your articles. Key factors include:
- Titles and initial sentences: The title and the first 75 words are essential. Use clear titles and language that matches how users search. If the article is long, include a summary at the top.
- Labels: Labels help organize content and can guide AI on what to include or exclude from search results.
By structuring your content effectively, you can enhance the AI's ability to suggest relevant articles and improve the customer experience.
Yes, AI agents consider both the title of the support article and the first 75 words when determining whether to pull information from it.
As a best practice, avoid duplicating the exact wording in the title and the initial sentences. Instead, use this opportunity to rephrase or incorporate synonyms that resonate with your audience. By doing so, you can ensure that you capture a wide range of customer-friendly terms in both the title and the opening of your articles.
Resources: Optimizing your help center content for AI agents
To enhance AI answers in your knowledge base, you can utilize three main sources: articles, posts, and external content accessed through the federated search API or crawlers.
It's important to note that no single source is prioritized over another, so you can't influence their usage rates. However, using search filters can impact the generated answers by providing different results for the AI to draw from. This means that applying the right filters can help tailor the information your AI provides to better meet user needs.
When deciding whether to combine information into a single article or break it into separate articles, consider the topic you're addressing:
- Single topics or FAQs: If you’re answering a specific question that frequently arises, it’s beneficial to create a standalone article. This makes it easier for users to find the information they need and can improve SEO.
- Complex processes: For topics that involve multiple steps or are more complex, it’s best to keep all related information together in one article. This helps prevent confusion, as AI agents rely solely on the context provided in the article.
- Use separate articles for clear FAQs and combine information for more complex topics to enhance the effectiveness of your AI responses and improve user experience.
Here are two effective tips for content writers to save time while ensuring accuracy with AI:
- Generative AI tools: Utilize generative AI tools to expand and enhance your article content. At Zendesk, agents can contribute to content creation by copying answers from ticket responses, pasting them into the article editor, and then using AI tools to optimize and structure the content for better presentation.
- Knowledge context panel and templates: Take advantage of the knowledge context panel, especially the templates feature. This helps streamline your published content and ensures consistency across articles.
By integrating these tools into your writing process, you can save time and improve the quality of your knowledge base!
Resources: 5 strategies for building up your help center content for AI
When creating a knowledge base for AI, attention to detail in content structure is crucial. Here are some important tips:
- Clear titles and subtitles: Be precise with your titles and subtitles. For example, if you have an article about FAQs, specify what the FAQs are about. This clarity helps AI better understand the content and provide more accurate answers.
- Use of lists: Incorporate bullet points and numbered lists effectively. Use numbered lists for sequences that customers need to follow (e.g., steps in a process), and bullet points for non-sequential information. This organization makes it easier for both users and AI to navigate the content.
To measure and improve your knowledge base, focus on these key performance indicators (KPIs):
- Self-service ratio: Track how often users find answers in your knowledge base compared to submitting support tickets. A higher ratio indicates better utilization.
- Content audits: Regularly review content related to high-ticket topics to identify gaps and optimize findability. This helps improve user experience and reduce ticket volume.
- Traffic and engagement: Monitor changes in traffic and engagement for updated content over time to assess the impact of your improvements.
- Customer feedback: Gather insights through surveys or interviews to identify areas for improvement and measure user sentiment.
Combining these quantitative and qualitative measures will help you enhance the effectiveness of your knowledge base.
Resources: Best practices: Measuring and optimizing your self-service channels
At Zendesk, we view our knowledge base as the foundation for enhancing all customer experiences, powering AI searches, AI agents, and the Copilot feature. Here are some key areas of focus for our roadmap:
- Improved searchability: We are working on enhancing the searchability and accuracy of knowledge retrieval for AI agents and users, ensuring they can easily access relevant information.
- Integration of additional sources: We aim to incorporate knowledge from other platforms, such as Slack and Confluence, alongside the articles in your knowledge base.
- Enhanced copilot support: We are developing features within the knowledge base to better guide Copilot in handling tickets, making it more robust and effective.
- Scalable management features: We are introducing features like article multi-placement to simplify managing large volumes of content. A new article editor will also be available soon, allowing for advanced components like drag-and-drop tabulation and accordions.
- Self-service dashboards: We will launch dashboards that track anonymous user activity, which has been highly requested.
- Attachment improvements: We continue to enhance the experience of adding attachments to articles and are working on features like search functionality within the media library.
These initiatives aim to create a more efficient and user-friendly knowledge base, ultimately improving the overall support experience.
Event Resources
Event Presentation
Knowledge Manager User Group
Event Recording
0 comments