AI-based Self-Service with Answer Bot
Customer service has undergone significant changes over time. In the past, customers typically contacted the support team through traditional methods such as emailing or calling a toll-free number. However, this often involved waiting for extended periods of time and being stuck listening to hold music, before being able to speak with an agent. In contrast, modern customer service has shifted towards more flexible and empowering self-service options.
Tools and resources that enable customers to independently resolve issues are gaining popularity. This is because research has consistently shown that approximately 80% of consumers prefer to tackle problems on their own before seeking assistance from a support agent. This preference for self-reliance and DIY problem-solving is a testament to the value of autonomy and independence.
There are a range of self-service tools on the market, such as online forums and knowledge hubs, as well as automated tools like chatbots, voice assistants, and live support for customers to access vital information at their convenience. Meanwhile, many companies are adopting AI-powered customer service, which enables them to develop AI-centric strategies for self-service solutions.
Availability
Permission Required Users with Zia administrative permissions can configure the Answer Bot, however, for support admins, this permission will be enabled by default.
Check Feature Availability and Limits
Importance of Bots in AI self-service
A bot is a self-service tool that allows customers to interact with the application, mimicking human-like conversation and granting a personalized experience. Bots can automate tasks that are simple, repetitive, informational, or transactional. If setup correctly, bots can play a vital role in increasing customer satisfaction and loyalty.
Customers who are struggling to find the right solution by navigating through the knowledge base articles, FAQs, blogs, and so on can easily find the answer by asking a bot. An answer bot is typically trained to comprehend queries and fetch a quick response. Each time a question is raised, answer bot leverages the information available in the form of FAQs, knowledge base articles, and other support material to provide quick and accurate responses. This is a simple example of how bots are vital in customer retention, driving improved satisfaction, and reducing load on the support teams by enabling ticket deflection.
Answer Bot for Zoho Desk Knowledge Base
The knowledge base is a centralized repository of information where product or service-related articles, educational materials, use cases, white papers, troubleshooting tips, FAQs and so on are maintained in an organized and structured manner that is available online for customers to access at their convenience, making it one of the most reliant self-service tool.
While the knowledge base holds extensive information, it can often be challenging to find relevant answers without scouring multiple topics and pages. Users may sometimes need instant answers for general queries like "What is your refund policy?", "How can I take a printout of the ticket?", and "What riding skill do I need for the Trans Euro Trail?"
This is where informational bots are convenient. Users can ask their questions to the bot, which gives simple responses and also includes the FAQs or articles that have the exact answer. They are vital in saving time.
Note: Zoho Desk Answer Bot and SalesIQ Answer Bot are distinct features available on the Desk and SalesIQ platforms, respectively. While they are different bots, they offer similar functionality.
Important benefits of a self-service bot in any business include:
Zoho Desk's Answer Bot is powered by Zia, Zoho's built-in AI tool. It is a self-learning bot that trains itself over the articles present in the Knowledge Base and generates responses to customer queries and ticket replies based on its ability to analyze, train, and interpret.
The Answer Bot is useful for customers and agents alike.
A customer who is confused on how to carry out an action like "how to take a printout of a ticket" can ask the Answer Bot for a solution instead of reaching out to the customer support agents. The Answer Bot can share the article that has this information and help the customer with a quick answer.
A support agent handling multiple customer interactions simultaneously can utilize the Answer Bot to get quick access to relevant Knowledge Base articles and key steps to the query raised without leaving the ticket detail view.
Best practices for training the Answer Bot
The Answer Bot's responses will be able to provide answers based on the information that you have facilitated in your knowledge base. The Answer Bot requires 30+ articles in a department to train itself on the furnished information. To get optimal results from the Answer Bot, here are some of the best practices we recommend while writing an article.
1. Opt for an easy-to-understand language
The Answer Bot can support 13 languages. It is recommended to use straightforward and easy-to-understand language while writing the articles to ensure clarity. If technical terms are necessary, provide clear explanations alongside them to ensure thorough understanding.
For example, while writing about a software enhancement, if technical terms like 'encryption' are unavoidable, provide an easy-to-understand explanation to ensure that its benefits and implications are clear.
The languages supported by the Answer Bot, along with English are:
| Spanish | Dutch |
| German | Danish |
| Russian | Swedish |
| French | Hindi |
| Portuguese | Arabic |
| Italian | Hebrew |
2. Structure the content
Ensure the content is organized using titles, headings, and subheadings that are clear, concise, and appropriate. This helps Answer Bot differentiate and understand the sections better. The inclusion of FAQ format also helps Answer Bot train better. The presence of headings and subheadings also helps the answer bot retrieve the most relevant section as answer to the question.
For example, an article on "How to Reset Your Smart Watch" can be structured as follows:
Title: How to Reset Your Smart Watch
Heading 1: Why Reset Your Smart Watch?
Subheading: When to Reset
Subheading: Benefits of Resetting
Heading 2: Steps to Reset Your Smart Watch
Subheading: Access Settings
This structured format helps Answer Bot navigate the article with ease and identify the right section.
3. Focus on one topic at a time
Ensure that each passage focuses on a single topic and try to avoid combining multiple topics into one passage. This approach provides better clarity , while training the Answer Bot.
For example, if a user asks specifically about how to restore data from drive to device, the Answer Bot can pull just the restore data passage and provide it as an answer to the user. This makes the Answer Bot's response faster and more relevant too
4. Keep your articles up to date
Keeping knowledge base articles up to date is an important practice. Revisiting articles time and again ensures the content is relevant, accurate, has no missing details, and includes the latest updates. This will help Answer Bot train on the available information and provide customers with the most accurate information.
For example, if a project management software has an update regarding third-party tool integration, if the information is not updated in the help doc on how the users can integrate and use the enhancement, Answer Bot will not be equipped to answer and support the influx of queries that tend to raise.
Creating an Answer Bot
Admins can create an Answer Bot for every department. However, each department must have more than 30 published articles for the Answer Bot to provide optimum responses. It's also important to note that even if the KB has 30+ articles, if there is not sufficient information on the query that a customer asks, the Answer Bot will not be able to fetch a response or suggest a KB article.
Businesses can monitor the tickets for which the Answer Bot couldn't provide a response, analyze the question, and publish more articles such as FAQs, user manuals, and training documents in the KB for the Answer Bot to train.
Note
- Admins can create an Answer Bot for all departments or for each department depending on their business needs.
- Admins can create one Answer Bot per department.
Creating an answer bot includes the following steps:
Bot name and description
The bot must be given a suitable name and description for users to understand its purpose. For example, for a travel and tourism agency, the bot might be named TripGuide.
Set visibility
Admins can create an Answer Bot and enable it for both agents and customers. One Answer Bot can be created for the customers that could be functional across departments. If every department has an Answer Bot, then the agent can access or chat only the bot respective to the department.
Customer can use web and mobile
For customers, the Answer Bot can be enabled for both webpages like the website, help-center, and landing page. It is also supported on mobile (via mobile SDKs), both iOS and Android.
Enable generative AI abilities
When creating an Answer Bot, the admin can choose to enable Generative AI integration. With this option enabled, each time a query is made, the Answer Bot will send the query and the relevant knowledge base article to the preconfigured generative AI to generate a well-structured ticket response. The Answer Bot will provide answers only if the Knowledge Base contains relevant information. It will not use Generative AI to fetch answers from open-domain sources.
Set default messages
A set of default messages are set up in order to maintain consistent communication and re-direct the user in cases the bot can not fetch an answer.
The admin can set:
- Welcome Message - The welcome message sets the tone for the interaction and informs users about the type of assistance the bot can offer. For example, it could be a general introduction like, "Hello, how may I assist you?" or a specific one like "Hello! Welcome to ABC Corporation. How can I help you?"
- When the Answer Bot is unable to respond - This default message manages situations where the bot cannot provide an answer. It acknowledges the limitation of the bot, like "Sorry, I can't help with this right now. Please check our help center or contact support for further assistance."
- When the user gives thumbs down - This default message is to address feedback from users who are dissatisfied with the bot's response. An alternative like contacting a support agent can also be included to improve customer experience. For example, "Sorry to hear that. We’re constantly working to improve. Please contact support at support@example.com for further assistance."
To create an Answer Bot
- Navigate to Setup > Zia > Answer Bot.
- In the Answer Bot page, choose a Department and click Add Answer Bot.
- In the Create Answer Bot page, provide the following details in Answer Bot details.
Enter a Bot name.
Add a description.
Enable the Bot either for agents or customers or both.
If enabled for customers, select the channels where the Bot will be deployed.
Enable Generative AI integration.
Set up the Default Messages.
Click Save.
Setting up Answer Bot for customers
Desk's Help Center is multifunctional and allows customers to submit tickets, engage in community discussions, and access the Knowledge Base hub. They can refer to the KB articles to find solutions to their queries independently. However, a quick assistance from the Answer Bot can save time and effort when people are looking for a quick solution. For businesses that want an answer bot exclusively for their customers, an All department bot is the most suitable option, this is because the bot will train itself on the resources available across the department and will have knowledge on varied topics, this will help the customers find precise solutions.
Charles, is planning his first trip to the Maldives and urgently needs information on the Visa on Arrival process and required documents. He is short on time and doesn't want to spend time searching for information or connecting with an agent. An Answer Bot can provide him with accurate and concise information on the visa process and document requirements.
Charles enters his query in the chatbot, "What are the document requirements and process for a Visa on arrival at Maldives?" The Answer Bot analyzes and comprehends the question, matches it to the articles available in the knowledge base, and then retrieves and delivers the most relevant piece of information to the Charles. He can click on the suggested article and find the required information. Read More Setting up an Answer Bot for the Help Center
Answer bot is a round-the-clock support assistant that:
- Saves time
- Improves self-service capabilities of the platform
- Enhances customer satisfaction
- Improves retention rate
- Reduces the load on the support teams
- Promotes faster resolution
- Reduces the response time
Note
- An all department bot will train itself on resources from across departments, this enriches its analytical and interpretative abilities.
- If you have created only an all department bot or don't have a bot for the respective department, then upon clicking the Answer bot option from the ticket, the agent may get an error stating "No bot is created" this can happen if the bot doesn't have enough training material on the respective topic and thus couldn't find an appropriate article to suggest.
- The Answer Bot can be enabled for each department. Before that, make sure that Zia is enabled for the department.
- The Answer Bot can work with a single successfully trained article, but for the best results, a minimum of 30 articles per department is recommended.
- The Answer Bot is available for the Help Center, webpages, and landing pages.It can be enabled via ASAP by embedding the given script and for mobile via Mobile SDK.
- The Answer Bot for customers will only fetch answers from knowledge base articles under registered users and all users.
Setting up the answer bot for agents
Support agents must adhere to a rigid schedule when addressing customer inquiries and resolving issues. They must comply with the Service Level Agreement (SLA) and prioritize critical tickets according to the First Response Time (FRT) guidelines, as specified in the business's agreement with its customers.
Each time a ticket is created, they must review the ticket to understand the query and then draft a solution for the reply. Often, the agent needs to refer to the knowledge base to find solutions or add supporting materials to their response. It can be challenging to perform these activities for multiple tickets they receive throughout the day.
Answer Bot can help the agents by suggesting the right articles from the knowledge base and a suitable response for the ticket. For example, Answer Bot can recommend an article that has information about service plans and billing issues, helping the agent save time in searching for the right article. In addition, it will also generate a response for the ticket based on its analysis, interpretation of the conversation, and the information in the KB article.
Overall, this provides an edge in efficient ticket handling by saving time and effort of the support agents, which they can spend handling other critical activities.
Note
- Ticket response will be generated only when generative AI (ChatGPT) is enabled for the answer bot. Read more
Using the Answer Bot in tickets
Agents can access the Answer Bot on the ticket detail page by clicking on the iBar on Zia followed by the Answer Bot. The Answer Bot for the department the ticket is in will open, and the agent can ask questions to the bot for assistance.
While agents are handling multiple queries at the same time, they have to refer to the knowledge base every now and then to find the right solution and supporting articles. The agent can ask the answer bot to find the solution and relevant help doc for the query.
For example, a ticket was raised enquiring if a single agent can be added to multiple departments. Here, the agent asked the same question to the Answer Bot and got a brief and a list of relevant articles to solve the query. The agent can save time spent searching for the articles and drafting a response.
Note
- In the ticket detail view, the Answer Bot will always display the article permissions for the articles used to generate the response. This helps the agent understand the source of the information.
- Agents also have the option to disable the auto-suggestion to use Zia Answer Bot , that appears in the iBar after clicking on "Reply All".
- To disable auto-suggestion of use of Zia Answer Bot, navigate to Personalization > Preferences > Disable Auto Suggestion.
Training and retraining Answer Bot
After the Answer Bot is created and saved, it will automatically start training itself on the available articles. After the first round of training, the user can view the status of training, such as:
- How many articles the Answer Bot has trained on
- How many articles are yet to be trained on
- If the bot failed in training on any articles
Each time an article is created or updated in the knowledge base, an option to re-train the Answer Bot will be enabled. While the scheduler is set for training Answer Bot every thirty minutes, the user can manually enable re-training.
Note
- Once every 30 minutes, the Answer Bot will train on articles that it is yet to train on.
- A minimum of five minutes should be given between two training or re-training sessions that are initiated manually.
- Answer Bot fails to train on the articles when:
- The article does not meet the minimum requirement of 100 characters or exceeds the maximum character requirement of 100,000 characters.
- The article is not well-structured, and each paragraph discusses multiple topics, adding complexity.
- Each paragraph does not follow the ideal word limit of 1000 words per paragraph.
- The article is in a language that is not supported by the Answer Bot.
- The article is embedded with HTML/numeric values that were unable to be translated.
- The article has images added without sub-content or there is large space between the image and content.
Disable or delete an Answer Bot
When agents access a ticket in the Ticket Detail view and choose the "Reply All" option, the Answer Bot chat window automatically opens with a pre-set default message to assist the agent. If an agent prefers not to use the Answer Bot’s assistance, they can disable it.
To disable an Answer Bot
- Navigate to Setup > Zia > Answer Bot
- Select the desired department from the drop-down.
- Hover the Bot and click the Edit icon.
- Under Enable for, uncheck Agent.
- Click Save.
To delete an Answer Bot
- Navigate to Setup > Zia > Answer Bot.
- Select the department of the Answer Bot that is to be deleted.
- Click the Delete icon on the right side of the Answer Bot.
- Click Yes, Delete option to confirm.
Access your files securely from anywhere
Zoho CRM Training Programs
Learn how to use the best tools for sales force automation and better customer engagement from Zoho's implementation specialists.
Zoho DataPrep Personalized Demo
If you'd like a personalized walk-through of our data preparation tool, please request a demo and we'll be happy to show you how to get the best out of Zoho DataPrep.
- Zoho Workerly Home
- Forums
- Connect With Us:
- Zoho Recruit Home
- Forums
- Connect With Us:
New to Zoho CRM Plus?
Deliver unforgettable customer experiences
New to Zoho CRM Plus?
Deliver unforgettable customer experiences
New to Zoho Marketing Plus?
Everything you need to run your marketing
New to Zoho Marketing Plus?
Everything you need to run your marketing
You are currently viewing the help pages of Qntrl’s earlier version. Click here to view our latest version—Qntrl 3.0's help articles.
New to Zoho Survey?
Zoho Sheet Resources
Zoho Forms Resources
New to Zoho Sign?
Zoho Sign Resources
Sign, Paperless!
New to Zoho TeamInbox?
Zoho TeamInbox Resources
New to Zoho ZeptoMail?
Zoho DataPrep Resources
Design. Discuss. Deliver.
New to Zoho Workerly?
New to Zoho Recruit?
New to Zoho CRM?
New to Zoho Projects?
New to Zoho Sprints?
New to Zoho Assist?
New to Bigin?
Related Articles
Zia - An Overview
Zia is an AI-driven support assistant who makes your customer support team more effective through data mining and machine learning. She provides support agents with analyses, customer forecasting, alerts, and suggestions through a pre-configured ...FAQs: Zia powered by ChatGPT and ChatGPT for Zoho Desk extension
How is Zia's integration with ChatGPT useful for my business? The power of generative AI (ChatGPT) coupled with Zia's capabilities can bring a spectrum of benefits to the customer support processes in businesses Contextual responses to tickets: Zia ...Zia - Field Predictions
Better customer service that eventually results in long-term engagement and builds stronger relationships with the customers is an asset that every business wants to invest in. While on the surface, customer support service may seem just about ...Introduction to Zoho Desk
Help Desk is a department in an organization that is in charge of providing the support needed by customers while using the company's product or a service. Customers contact the help desk when they have a question or a problem and a product expert ...Integration of Zia with OpenAI (ChatGPT) in Zoho Desk
Availability: Zia Powered by GPT is accessible in Standard, Professional, and Enterprise editions. Currently, it is only opened for US, AU, EU, and India data centers in order to comply with GDPR regulations. Customers located in countries outside of ...
New to Zoho LandingPage?
Zoho LandingPage Resources