7 Ways Companies Use AI for Customer Service

At Help Scout, we believe great customer support is more than providing correct information. It’s about creating connections, demonstrating empathy, making proactive suggestions to move customers forward, and more.

While AI is here to stay in customer service, it will never entirely replace human agents. AI tools help customer service teams do better work — it doesn’t replace them.

In this article, we’ve collected seven successful examples of companies that have embraced AI in customer service — not to take the place of human agents, but to complement their abilities and make them more effective than ever.

What is AI in customer service?

At the highest level, AI in customer service means using artificial intelligence and machine learning technologies to automate customer support processes and interactions.

AI in customer service looks different depending on the business. For example, one company might offer a conversational chatbot that pulls from their help center to assist customers in a human-like way. Another might use AI features to improve an agent’s email response so it sounds more on brand or to summarize a lengthy email thread in seconds.

How are companies using AI for customer service?

As more companies are incorporating AI into their toolboxes, these are some of the most common ways the technology is being used in customer service:

To engage in conversations and respond to customer queries: Chatbots use generative AI to understand and respond to customers’ questions, no matter the time of day. By integrating this technology with an existing knowledge base, chatbots can be incredibly effective at solving common customer issues without escalating them to a human agent. However, like any other technology, it’s not 100% reliable. There will be cases where the information provided might not correspond to the question or is simply wrong. You’ll need to map out those scenarios and make it easy for customers to contact a human agent if their problem isn’t solved.
To understand satisfaction levels: AI is being used in analytics tools to analyze customer feedback at scale. From emails to reviews to social media posts, this AI-powered analysis is helping customer service leaders better gauge customer sentiment and emotions. Technology is faster and more reliable at finding common themes or issues before they become a concern so businesses can address them proactively.
To offer multilingual customer support: Customers want and expect customer support in their own language. However, having support agents available in all languages can be costly. Using machine learning algorithms to produce accurate translations quickly, businesses can communicate with customers in almost any language. For example, companies are using AI in their ticketing software to quickly change the language of an agent’s response to match the customer’s needs.
To categorize and prioritize incoming support tickets: AI in customer service can handle tedious, time-consuming tasks previously handled by agents, such as triaging customer support tickets. AI-powered ticket routing can automatically categorize incoming customer conversations based on their content and urgency so that they reach the appropriate team more efficiently, reducing response time and freeing human agents to carry out more valuable tasks.
To improve human agents’ answers: A common use of AI is to help agents do their best work. AI technology summarizes a customer’s question so the agent saves time figuring out what the customer wants. AI writing assistance tools can shorten or lengthen an agent’s answer and even edit a response to match a business tone.

This list contains some of the most common scenarios of companies using AI in customer service, but it’s constantly evolving. As AI improves and is integrated into more tools and processes, businesses continue to find new ways to use AI to improve and optimize customer support experiences, increasing satisfaction and retention.

7 examples of companies successfully using AI in customer service

We’ve collected seven real world examples from companies using AI to improve their customer experience.

1. Airbnb – Chatbot and self-service

The world of customer service chatbots can be divided into two groups: chatbots that are powered by AI and chatbots that aren’t.

You’ll usually recognize chatbots that use AI, because they are surprisingly good at understanding and responding to customer questions. Chatbots that don’t leverage AI are much simpler and more structured — they follow pre-defined rules to interact and lead users through a decision tree. Unfortunately, even AI chatbots often fail to provide relevant information unless customers use the “right” prompt, which can cause some odd agreements.

Airbnb demonstrates what an experience with conversational AI in customer service should be: clear, smooth, and informative. The best part is that it doesn’t purposely make connecting customers with a human agent hard. If you don’t get the help you need through self-service, reaching someone is easy.

This type of AI chatbot is really good at letting customers self-serve while businesses can reduce support costs, improve efficiency, and enhance satisfaction.

2. Wanderlog app – Conversational AI

If you’ve ever had to organize a long holiday trip, you know how much time goes into it. It’s almost a full-time job: reading travel blogs and guides, finding the best places to eat or stay, and figuring out how to get from A to B.

If you’re lucky, you might know someone who can give you some personal recommendations.

Wanderlog is a handy little app that helps you plan and organize your trip. It comes with an interesting AI feature that uses ChatGPT’s vast data to act as a travel agent, providing recommendations and tips on activities and even organizing your daily itineraries.

By using a generative AI bot, Wanderlog makes it fun and easy for travelers to get recommendations, explore activities, and find travel assistance with fast, detailed answers without having to reach out to a customer service agent.

3. TransferGo – Multilingual chat automation

While English is often considered the language of international business, only one in five people can actually speak it. Offering multi-language support is important for businesses, as 75% of customers are more likely to return if support is in their language.

TransferGo, an international money transfer company that lets people send money quickly, is now using AI to offer multilingual chat automation to their customers. The company uses a virtual agent fluent in 11 different languages.

Using AI, their multilingual virtual agent reliably and quickly handles self-serve intents, such as changing names, addresses, or phone numbers in seven languages.

4. Uber – Sentiment analysis

AI-powered sentiment analysis tools can analyze customer feedback to gauge customers’ sentiments and emotions. This helps businesses understand and address concerns in real time before they become problems.

Uber uses AI technology to measure the effectiveness of their strategies. For example, after rolling out a new rider app, Uber used AI to identify unhappy users and used feedback to address issues in the app before they became a major problem. This approach gets upstream of your customer service team, enabling you to fix issues before your support inbox gets slammed with angry customers.

5. Help Scout – AI writing assistant

At Help Scout, we’re proud to use our AI features to improve our customers’ experiences while helping our team work more efficiently and effectively. However, some of our AI writing assistance features, AI drafts in particular, have even benefitted people beyond our support team.

Every team member, from our engineers to our marketing team to our CEO, spends time responding to real customers in the Help Scout queue. And while we’re all familiar with our product, it can still be a little scary for those who may not be used to working directly with customers.

AI drafts is a tool available in Help Scout that can generate responses to incoming conversations based on data within your account, such as conversation histories and knowledge base content. It’s made a big difference when it comes to increasing comfort in the queue.

Kristi Ernst Thompson, one of our senior technical support specialists, says, “AI drafts have made Whole Company Support feel a lot less daunting. Our support team always encourages everyone to fact check and revise drafts so they’re written in their own voice, but we’ve found this gives them a great jumping-off point.”

There can be a certain amount of anxiety that comes with staring at a blank page, and AI drafts makes it just a little bit easier to get everyone in the inbox and helping customers.

6. IKEA — Virtual room planning

Making a “quick visit to IKEA” isn’t a thing — and that’s especially true if you need interior design help. Now, with IKEA’s AI room planning tool, users can speed up the design process.

IKEA customers can use the AI tool to create their designs when it’s convenient for them, rather than spending lengthy sessions on the phone or in person with a customer service associate.

While this may not seem like a standard use case of AI in customer service, it’s a great example of zooming out to think about the broader customer experience. Prior to using this AI technology, IKEA customers had to spend a lot of time and effort to design their spaces. Using AI technology makes that process far easier and more convenient, and it also decreases the likelihood of future returns and dissatisfied customers.

7. Delta Airlines — Knowledge management

While low-cost airlines have traditionally focused on price, Delta Airlines has been trying to compete by offering a great customer experience. At a conference last year, Delta’s CEO, Ed Bastian, explained how AI is helping their support teams be more efficient by making information easier to find.

“We’re working with our reservations team to try to help our reservations agents parse the historical policies and questions and things that you may call into a real agent,” he explained.

An example of this in practice is when a passenger calls with a question about traveling with a pet. Using AI, the customer service rep can quickly access an answer from a procedural manual. This kind of information has historically been available in help centers and FAQs, but Delta’s working to make it more accessible to call center agents. “People go on and are on hold for five minutes waiting for an answer; they should only be on hold for five seconds,” Bastian said. “That’s what AI can do, and that’s one of the first applications that we’re deploying.”

Best practices for using AI in customer support

AI is far from perfect. It can misunderstand requests, give out unclear or incorrect information, and suggest unrelated help center articles.

However, while not flawless, these companies (and many more) have already benefited from using AI to improve their customer service. Below are some best practices for implementing AI in your own customer service journey:

Set clear objectives. Having clear objectives will help you find the AI solution that fits your needs. Perhaps you want to reduce response time in languages other than your team’s primary language or you want to start monitoring sentiment to offer more proactive support. Whatever your objective is, it’s the basis for figuring out which tool best fits your needs.
Choose use cases that fit the team’s needs. Find the specific use cases where AI can have the most significant impact on customer support. Prioritize repetitive, time-consuming tasks, such as answering common queries, routing tickets, or providing self-service options.
Start by leveraging your existing tools. Many tools, like Help Scout, have now released AI-powered features and functionality. Before investing in additional tools that make big promises, optimize your existing tech stack to take advantage of these new features.
Test, test, and test again. Unfortunately, many businesses still offer a sub-par AI-powered customer support experience. It’s not the AI’s fault; it’s because the humans didn’t take the time to fine-tune it. AI is a powerful tool, but it needs to be monitored to ensure the accuracy of AI-driven decisions.
Have human agents overseeing the AI models. When using AI to assist human agents, be sure to have a process in place for agents to validate the AI’s work. This is to avoid any inaccuracies coming from the AI model that can be quickly corrected by a human.
Monitor performance by setting up KPIs. Start by deciding which metrics to measure and how to measure them. Without KPIs, you won’t know whether the AI is meeting your expectations. You can improve and update the AI models periodically as your needs evolve and you collect customer and employee feedback. Deploying AI in customer support is not a static option; instead, it’ll evolve, especially as the technology keeps on developing.

AI will radically impact how customer service teams are organized, but AI is no substitute for the skill and empathy that support professionals possess.

Optimize your customer service with AI

As artificial intelligence rapidly evolves, it’s obvious that it has many use cases in customer service. It’s especially relevant when applied strategically to automate repetitive or tedious tasks, freeing up customer support agents to do more impactful work.

In the examples we’ve seen, companies have successfully implemented AI to complement the work of human agents and to improve the broader customer experience.

One thing is crystal clear: The benefits of AI are highly dependent on how thoughtfully you integrate AI into your customer service tools and processes.