Bringing AI into your customer service workforce has never been easier, and implementing chatbots can help your business to save up to 30% on your customer service costs, according to Gartner.
When you choose to work with Netomi, you don’t have to worry about code and nothing changes for your customer service team. Based on the end user experience that you want to provide and the level of customization you want to have for each use case, you can choose between two options:
- AI premium service: This is our best-in-class solution that comes with a dedicated customer success team that provides white-glove support every step of the way. We’ll create a really engaging, natural chat experience. You will have complete customization on the conversational design and the ability to integrate with multiple back-end systems.
- Knowledge base service: A self-serve option that links to your knowledge base and will be up and running immediately.
Whichever option you choose, launching Netomi’s AI on Zendesk chat is quick and simple.
Zendesk Chatbot: Netomi’s Premium Service
When you select our conversational customization option, you’ll create a highly-tailored experience that is expertly designed for the chat interface. The below six steps are completed in as little as two weeks.
Step 1: Gather historical data
Historical data is the best place to start when you’re launching a new conversational AI experience for customer service. Pull historical customer service tickets from your Zendesk account. You’ll send an export to your dedicated Netomi customer success team for analysis.
Step 2: Analyze data to identify use cases
It’s critical that each ticket is managed properly. By leveraging historical tickets, you’re taking the guesswork out of which tickets to delegate to a chatbot, and which should be always routed to a human agent.
High-risk, critical or rare issues should always be routed to a human agent. Tickets that are highly-repeatable, pose low-to-medium business risk and require low-to-medium exception management are the ideal situations to automate.
During our clustering process, we’ll identify the right use cases. Essentially, our algorithms will group together historical tickets that have the same intent, making it really easy to determine which use cases account for a high volume of tickets and uncover trends in how human agents have responded in the past. During this step, you’ll select the tickets you will delegate to your AI.
For instance, a subscription company will likely see that order modifications, skipping a delivery or address updates account for a lot of tickets and therefore would make great candidates for AI. Edge cases, such as specific questions regarding a recipe or preparation, would likely account for fewer tickets and therefore would yield a lower benefit when automated.
This clustering process also helps jumpstart utterance training, giving your AI the ability to understand how people ask things in real-life, not simply the wording on your online wiki or FAQ page. This extensive utterance training helps your AI more accurately classify your customers intent from launch, ensuring a positive customer experience.
Step 3: Conversational design
Once the use cases that will be delegated to AI have been selected, the conversational design process begins where you bring your virtual agent to life. This consists of a few key things:
- Bot Personality: Within Netomi’s AI Studio, you’ll bring your bot to life through the tone of your bot responses, making sure it aligns with your brand.
- Tailoring Use Case Journeys: Create and tailor your responses for each use case within AI Studio. Don’t forget to add emojis and other conversational U/X features like buttons, suggested responses and more.
- Conversational Elements: You will also create responses for conversational copy (greetings, common questions), create menus and bridges back to core functionality.
Step 4: Connect your Netomi bot with your Zendesk Chat
It’s really easy to integrate a bot with your Zendesk chat account. You’ll simply need to create a dedicated Netomi API client for your particular Zendesk subdomain which will have the ability to access Zendesk, just like any human agent would.
To link Zendesk to a bot on AI studio, your dedicated customer success team at Netomi will create a simple social configuration on our end.
Step 5: Connect to your knowledge base
Do you want your bot to provide recommended, relevant articles to use cases that you haven’t trained? You can integrate with your Zendesk help center by linking to your knowledge base in AI Studio. This way, even if your AI isn’t providing a full resolution within the chat interface, it can offer the potential for self-service and subsequently deflect tickets from other channels.
Step 6: Launch, measure and optimize
All that’s left now is to embed the Zendesk chat Web widget on your website and we’ll activate your bot. You’ll want to watch performance and review how the AI is interacting with your customers to identify opportunities to enhance the training, add additional use cases or adjust the conversational flow and responses.
You should start seeing increases in customer satisfaction and agent productivity within days. Check for all the most critical customer service KPIs within the Netomi AI Studio.
Zendesk Chatbot: Netomi’s Knowledge Base Service
Do you want to get up and running immediately? Our quick and easy Knowledge Base integration is for you. Just enter the URL of your knowledge base into AI Studio and you’re up and running. When your customers ask your bot a question, they will be sent relevant articles to supercharge self-service.
Step 1: Train the AI by integrating with your knowledge base
Are you looking to launch immediately and have a robust knowledge base on your website? You can train our AI by simply providing a link to your Knowledge Base. Our AI will immediately get to work analyzing your information so that it can serve up the most relevant ones when users ask a question.
Step 2: Create conversational and human handoff copy
Within Netomi’s AI Studio, you will have the ability to customize how your bot responds to customers with greetings, hand-offs to human agents and other conversational elements.
Step 3: Launch
Press “Go Live” and you’re all set!
Step 4: Measure
See trending topics, customer engagement and other key customer service metrics within AI studio.