Recently, we addressed questions from our community about Zendesk's AI tools for Guide and Self-Service Best Practices. These tools and methodologies are designed to boost efficiency and introduce automation into your support strategy. The event featured a detailed Q&A session with Zendesk's Product and Documentation teams, who offered insights into their practical applications and shared best practices.
If you missed it or joined us but want to review, you can find a link to the full recording in the links below. We also encourage you to read the event summary article, where we’ve provided answers to all of the questions we were not able to answer live during the event. Keep an eye out for further updates on our AI tools and join us for upcoming events.
For more in-depth discussions and the latest in CX innovation, join us at Zendesk Relate from April 16 - 18 in sunny Las Vegas for our premier user conference.
Video Recording
Top Questions from the Q&A
Q1: Will the new Zendesk AI Tools in development be an extra cost option?
Some AI features are included with standalone product and Suite plans, with specific features varying by plan level, while other features are available with the Advanced AI add-on for Suite and Support Professional and above plans. See Zendesk AI and Advanced AI at a glance.
Resources
Zendesk AI and Advanced AI at a glance
Q2: How can we personalize the self-service experience to cater to individual customer preferences and history?
You can connect bots with your front- and back-end systems like a CRM or external order system. These types of integrations can give agents historical information on customers’ past purchases, return history, loyalty information, and more to tailor the customer experience and ensure they keep coming back. See Creating webhooks to interact with third-party systems.
Resources
Creating webhooks to interact with third-party systems
Q3: How can we ensure knowledge-base / help centre chat bots don’t hallucinate?
Our product team is constantly working on making sure that the our AI is able to provide the most accurate support. With our in house intent models, and prompt engineering for generative AI that significantly reduces hallucinations. Also being able to connect your bots to the help center rather than allowing your AI to gather information from non specific sources limits the exposure of AI to work within the bounds of the information that its provided.
Q4: Are there any guarantees that the data is ethically sourced and nothing we feed the AI is then usable by other subscribers of Zendesk?
Building responsible AI is a part of Zendesk’s commitment to providing our customers with trusted products and solutions. Zendesk leverages a set of design principles with respect to our AI products, which include privacy, security and compliance by design.
Resources
Zendesk AI Data Use Information
Q5: What are the most important success metrics for self-service AI and how can they be used for optimizing content or user flow/experience?
The article Guide reporting tools for measuring self-service acts as a good starting point for measuring the effectiveness of your self-service offering. It walks through helpful reporting information about calculating your self-service score, reporting on tickets solved by auto replies with articles, and more. Additionally, the Zendesk Answer Bot dashboard in Explore gives you insight into automated resolutions from your bot.
Resources
Guide reporting tools for measuring self-service
Analyzing your bot builder activity
Q6: How can AI systematically make improvements to the self service experience?
A great tool for this is Intent Suggestions. Intent Suggestions takes the guesswork out of what to automate for your bots. The intents tab within the conversation bot, highlights top asked intents with no answers attached, that allows admins to know exactly what needs automation. And you can select up to 10 intents to an answer. Content Cues is a great tool that identifies where you may want to make improvements to articles in the help center. It will show you trending topics and also relevant tickets and tags to those topics
Q7: Can Chatbot pull from non-public articles and ask the user to login and see more info?
Our generative replies feature (part of the Generative AI EAP) and article recommendations feature can use content from articles that require authentication, according to the permissions you’ve set up in Guide admin for the corresponding brand help center. For details on how this works with generative replies, see Using generative AI to generate replies in a conversation bot. For details on how this works with article recommendations we have extra documentation to help.
Resources
Generative AI EAP
Designing a conversation bot using answers
Q8: If AI is the one interacting during chat/messaging, is there a way for our Agents/team leads to monitor a live interaction?
We want to allow customers to self-serve to save agents time, which is why agents cannot editor monitor bot responses generated by GenAI.
Q9: How can customer feedback be used to continuously improve and expand self-service capabilities?
We would recommend enable comments on your help center articles and assign specific people or teams to monitor and respond to comments. Often, comments on articles can indicate a need for updates to the article content itself. See Setting up to collect feedback about your knowledge base content. If your customers are actually internal agents, they can flag needed article updates in Knowledge.
Resources
Flagging articles with the Knowledge Capture app
Getting started with self-service
Q10: How can we leverage AI to categorize tickets, as well as reduce volume/manage high volume instances?
Macro suggestions for admins are refreshed at the beginning of each month. But if you referring to suggested macros for the agent feature, we update the models about once a month. Once you're eligible, you may have to check your account settings to enable the feature when you are eligible.
Resources
Creating macros from macro suggestions for admins
Enabling and disabling suggested macros
Presentation Material:
Future Events & Opportunities