Here’s how customer service teams are actually using AI
Generative AI tools for customer service driving higher productivity, happier agents and satisfied customers
Last updated February 23, 2024
Let’s take a moment to recognise how hard customer service teams work to keep your customers happy. That’s a big mandate even during the best of times. Today, it feels like businesses are dealing with higher ticket volumes and more customer issues than ever before. Add uncertainties to the mix – like supply chain issues, staffing shortages and hiring freezes – and agents have a lot cut out for them. Your customers are probably seeing more friction in their experience than they need to, as well.
When we talk about AI in customer service, a lot of people immediately default to thinking about chatbots that act like humans. But that’s not the case. One of the most powerful ways businesses are using AI is by making their service teams more agile, especially when dealing with high volumes of customer requests.
Service teams have a tonne on their plates
Between high ticket volumes and repetitive tasks, support agents have their hands full. As customer issues come in, service teams can spend a lot of time and resources categorising each one manually to ensure it gets to the right place. Inside the actual conversation, they might lack crucial insights that could result in a quicker resolution. All the time spent inside conversations doing repetitive tasks eventually amounts to hours that could be spent on higher-value tasks.
Of course, automation isn’t meant to replace humans. It should make their jobs easier. On top of that, your customers still want a human experience. Rather than replace your whole service team with sentient bots, AI can be used to automate repetitive and time-consuming tasks to enhance your customers’ experience while creating a more efficient workflow for your agents.
Automation isn’t meant to replace humans. It should make their jobs easier.
Here are ways businesses can use AI from Zendesk to make their service teams more efficient. First, there are advanced bots that come with pre-trained customer intents specific to CX teams, which means more personalised and accurate responses, more productive agents and faster setup. Second, you have intelligent triage, which automatically classifies and categorises incoming customer conversations based on customer intent, sentiment and language. It helps service teams automatically prioritise and route the conversations to the right agent. Third, there’s intelligence in the context panel, an agent-facing application that provides insights about customers to help agents respond more effectively. And lastly, macro suggestions for admins, AI-powered suggestions for new shared macros that admins should create to make their agents more effective.
Advanced bots with Generative AI
Bots are a support agent's best friend. They’re extremely effective at taking repetitive questions and tasks – like password resets and refund requests – off a service team’s plate, enabling agents to focus on higher-value conversations. Bots powered by Zendesk are easy to set up and deploy across all messaging channels, and they're also easy to customise and change on the fly.
Zendesk's bots are not only easy to set up across all messaging channels, but also simple to customise and adjust as needed. With the addition of Zendesk AI, these bots become even more potent and user-friendly.
Zendesk's bot builder, a tool that requires no coding, is the secret to this adaptability. Admins can tailor the bot's responses to individual customer issues using pre-set messages, help centre articles and more, all without needing a developer. This allows you to shape how the bot interacts with your customers in a way that best suits your business.
Our bots also employ Generative AI, using knowledge from your help centre to swiftly create accurate, conversational responses for your customers. You can even construct your bot's persona, ensuring it maintains a consistent level of service across channels, whether it's friendly, formal or playful.
Our advanced bots are intelligent tools for messaging and email, leveraging a vast database of customer intents specific to CX teams in your industry. This results in more personalised and accurate responses, increased agent productivity and quicker setup. These bots come equipped with tools that provide admins with recommendations on what needs to be automated.
When a conversation appears in an agent’s workspace, they need to figure out how to prioritise it against other conversations, who can best solve the issue and how the customer is feeling.
An experienced agent might be able to determine these things at a glance. Still, when the tickets are piling up and customers are pressed for a resolution, the agent’s time would probably be better spent actually resolving the issue.
That’s where AI is a powerful friend – it can automatically analyse requests and determine customer intent, sentiment and language in order to route and prioritise issues. That way it lands in the hands of the right agent who can solve the issue most effectively.
For an AI tool like Intelligent Triage to automate tasks such as sentiment analysis and routing, it requires a large set of data at its foundation. The majority of companies don’t have time to build out this data set from their customer conversations, as doing so is time consuming and costly.
To work around that, Zendesk has already done a lot of the heavy lifting, using our own data to train AI to understand repetitive tasks and common queries. Companies that roll out Intelligent Triage for their service teams can see value quickly. The end result is service agents who are more agile and efficient.
Intelligence in the context panel
While intelligent triage works behind the scenes to route conversations, agents can also use AI to gain insights and help them respond in a more understanding, effective way. Once conversations are routed and prioritised, agents can use intelligence in the context panel to glean insights about the customer and determine the best course of action in the conversation.
Alongside the conversation, agents might see a suggestion to automate a response to the customer without having to do manual work. For example, a retail agent might see a prompt to suggest a refund or an update on an order status or a macro that sends a relevant knowledge base article. With insights that show intent, language and sentiment, it puts agents on the quickest and most efficient path to solving a customer issue.
Like a co-pilot, intelligence in the context panel helps agents do their jobs more efficiently, removing the guesswork and repetitive manual tasks. It makes interacting with customers easier through intuitive tools and accessible suggestions, thereby reducing pressure on agents who are experiencing higher volumes of requests and customers who are facing more complex challenges.
Macro suggestions for admins
Zendesk AI can surface relevant help centre articles in your customer conversations and it can fill in the gaps where articles don’t yet exist. Content Cues are friendly, AI-powered nudges for improving your knowledge base. Zendesk AI scans every inbound support request and collects those trending topics for knowledge management teams. If an admin sees a topic pop up, they can work with their knowledge management team to create a help centre article, so customers can get the answers that they need on their own. Generative AI for your knowledge base effortlessly crafts content to boosts self-service, accelerating and streamlining content creation. The content's tone can be shifted to be more friendly or formal, and creativity can be inspired.
Zendesk AI can also support admins behind the scenes to improve your operations and streamline resolutions with macro suggestions for admins, which are new, AI-suggested shared responses that admins can create to help agents respond faster and more consistently. Our AI identifies gaps in existing macros and creates relevant responses to reduce time spent on analysing tickets and conversations.
AI adds up to savings
On average, intelligent triage and intelligence in the context panel can save 45 seconds per customer issue. For an average enterprise retailer, that can amount to 120 hours a month in time saved by AI. It enables service teams to tackle high volumes of conversations by automating repetitive tasks and requests using an enormous trove of data.
The modern customer journey is complex and constantly evolving. And many businesses feel pressure to predict the where, when and how of their customer interactions. Meanwhile, the conversations are piling up.
On average, intelligent triage and intelligence in the context panel can save 45 seconds per customer issue… that can amount to 120 hours a month saved by AI.
Thankfully, AI simplifies a great deal of the customer experience. It helps resolve issues faster and more efficiently. Better yet, it’s easier than ever for businesses to roll out – in an instant, not weeks or months. Zendesk has made it beautifully simple.
With Zendesk, our easy-to-use AI helps service teams scale and simplify support. Deploy quickly to start solving issues right away, surface relevant insights to aid your team, reduce repetitive work and start building better bots.
The best part is that it’s already here.