Artificial intelligence (AI) has quickly conquered the customer service realm, and that comes as no surprise.
Companies know they must understand and deliver what customers want to remain successful or stay in business. AI plays into this picture in numerous ways.
Artificial intelligence can reduce the number of steps people go through to get their problems resolved, help route calls to the agents best equipped to take care of customers’ needs and, in some cases, answer queries at any time of the day or night.
It also aids in driving sales and steering customers towards the products most likely to meet their needs. Amazon is famous for this approach, suggesting other things people should buy based on the merchandise already in their shopping carts. Then, even though individuals often spend more on retailers’ sites due to this technique, they typically end up more satisfied due to the relevance of the suggestions and the perception that the company wants to help.
Further, AI does an excellent job of predicting the factors that cause customer churn — where people stop doing business with one company and pick another. Similarly, it can flag potential warning signs through changes in consumer habits.
For example, it might detect that a customer who used to spend at least $100 per month at a store now only shops there every few months and spends less than $20 each time. Then, representatives can intervene to find out what went wrong and try to remedy the situation before it’s too late.
The following seven examples highlight specific industry examples of how companies use AI in customer service to meet their goals.
7 ways AI in customer service is used
- Improve content personalization
- Find out how customers feel about loyalty programs
- Offer faster responses to common queries
- Streamline the customer journey with chatbots and other virtual assistants
- Shorten ticket resolution times
- Improve real-time emotional intelligence
- Slash return rates
1. Improve content personalization
Companies are in exceptionally demanding positions when they serve customers who have various needs to fill, speak several languages or hail from different regions. Such situations make it challenging to offer sufficiently relevant and interesting content. However, AI can help.
For example, Swisscom AG sought to stand out via exceptional and seamless offline and online experiences. It used several Adobe products, including AI-powered Sensei, to provide highly personalized content for customers.
Sensei also makes A/B testing simpler by identifying consumer behavior patterns and automatically routing them towards the best-performing version of a site based on customer characteristics. Furthermore, the tool detects things like a customer’s geographic area and gives them the most appropriate information.
Using AI in these ways allows Swisscom to enhance personalization while reducing labor-intensive processes. AI saves the company time without sacrificing the personalized content that keeps customers coming back. As such, it disproves the possible assumption that having hyper-personalized material requires significantly more team members and longer processes.
2. Find out how customers feel about loyalty programs
Customers appreciate loyalty programs that give them the rewards they want and the benefits they’ll use. In an industry such as travel, where so many airlines and hotel chains have dedicated programs, it’s especially important to ascertain how customers feel about those offerings.
Hong Kong airline Cathay Pacific has a loyalty program called Marco Polo. After the program went through substantial changes concerning what awards it offered and how customers participated, researchers used machine learning — a subset of AI that focuses on algorithms getting smarter over time based on the data received, and sentiment analysis to gauge how customers responded to Marco Polo’s updates.
The team initially focused on Cathay Pacific’s loyalty program members but then expanded the machine learning algorithms to look at the factors that make people more likely to switch to other loyalty programs. Thanks to AI, the researchers analyzed more than 400 million words that people said about airline loyalty programs.
They discovered that when individuals simultaneously spoke negatively about Marco Polo and talked about other airlines or their loyalty programs, they were highly likely to book flights with other airlines or join their loyalty programs within 85 days.
Any company could apply a similar strategy before or after altering their loyalty programs. The popularity of social media means plenty of sources can give companies glimpses into customers’ feelings.
3. Offer faster responses to common queries
Most companies realize their customers ask certain questions more than others. Unfortunately, although they’re often simple, those queries put burdens on customer service agents who need to spend their time handling more complex situations. Instead, AI offers a better alternative.
One example is Course Hero, an online learning destination with millions of resources. The company wanted to give their customers faster responses while automating some of the parts of the customer journey that take the most time.
A company called DigitalGenius used in-house AI tools to meet both those requirements. DigitalGenuis AI products handle 33% of inbound customer queries, and more than half of those get resolved without human involvement.
This use of AI gives Course Hero’s employees more time to handle more complicated questions, and it has saved the company tens of thousands of dollars. Course Hero’s customers have ample praise for the company’s customer service practices and mention how their questions were answered thoroughly and efficiently when interacting with the technology.
4. Streamline the customer journey with chatbots and other virtual assistants
The Course Hero illustration above is one example of how a customer could get the answers they need faster when a company applies AI to their processes. But, it’s far from the only option. Some businesses use AI to guide people through each step of their buying journey by encouraging them to communicate with chatbots. Statistics show that 60% of consumers used a chatbot to talk to a brand within the last 12 months, which suggests they’re embracing the tech.
See how rapidly chatbots are taking the world by storm by downloading our roundup of the top chatbot statistics!
The brand 1-800-Flowers is primarily an online retailer that helps people send flowers to their loved ones. It recently deployed a virtual concierge called Gwyn that assists customers with getting bouquets and other gifts. It understands vague sentences like “I need a gift for my grandmother” and can also help with more specific queries like, “Show me your highest-quality roses.”
The interface asks follow-up questions throughout the conversation and uses the customer’s responses to make more accurate and tailored suggestions. Moreover, the chat tool gets smarter the more that customers interact with it. Since 88% of the company’s business operates online, it’s crucial that internet-based shoppers find what they need without hassles. Gwyn achieves that aim.
TIP: Check out how AI in retail is used in 2019, with thrilling examples like the one above! |
Chatbots also help shorten the customer journey by determining when it’s best for people to talk to humans. For instance, Alfredo is an AI virtual assistant used by an Italian banking group called Credito Valtellinese — also known as Creval. Alfredo gives approximately 1,000 automated responses based its training on all 14 of the brand’s knowledge domains. Alfredo receives all service desk requests and can connect customers to human agents via Skype when needed.
This use of AI in customer service resulted in 92% of customers giving positive feedback about Alfredo, partially due to the faster customer journey. People can now get the answers they need without going through so many separate channels. Even when the chatbot switches a user to a human worker via Skype, everything happens in one interface.
5. Shorten ticket resolution times
Many people who agree to take surveys after contacting customer service are asked how satisfied they are with the resolution process and whether the interactions with the company took care of the problem. A business typically marks an unresolved issue as “unresolved” or leaves it open. Then, it’ll likely calculate the number of open or unresolved tickets to see if they go up or down.
That method of measurement makes sense, mainly because people don’t like to repeatedly get in touch with customer service professionals to sort issues out. It’s becoming common for companies to boost their customer service efforts and use AI to accelerate the time required to close customer support tickets.
Financial services company Nordea wanted to minimize the time needed to resolve claims from its life insurance customers who had disability claims. Before the brand invested in AI, the average claims-handling time was 75 days. That’s consistent with industry averages, but Nordea wanted to do better. They relied on AI to get results.
The company used AI solutions such as computer vision, which can read handwritten letters and convert them to digital text — and machine learning algorithms that separated disability claims into three categories. These processes separated claims according to risk and complexity while helping to detect fraud. Nordea representatives believe their fully trained model will handle up to two-thirds of cases, freeing up 50% of the company’s claims capacity.
Uber takes a similar approach in applying AI to English-language support ticket resolution. Besides categorizing the tickets by using machine learning and natural language processing (NLP), the brand’s system recommends responses for workers to give as they talk to customers. This approach reduced resolution times by more than 10% while achieving customer satisfaction levels that were equal to or better than non-AI-driven communications.
6. Improve real-time emotional intelligence
Skilled customer service representatives are emotional intelligence experts. The best ones know what to say to soothe tempers and avoid making fed up customers more irate. But phone conversations don’t allow agents to pick up on things that could help them detect distress faster, such as crossed arms or an angry facial expression.
AI can’t see a person’s responses through the phone, but a company called Cogito offers intelligent software to help pick up the non-verbal cues workers may miss. Before one health insurance company began using the technology, many of their current or potential members perceived a lack of empathy from the support agents. As a result, they had consistently high dropout rates and lower-than-desired numbers of new customers.
The company’s call center agents using Cogito received instant feedback that alerted them when customers showed signs of needing extra empathy. The resultant increased engagement between call center representatives and callers cut the membership dropout rate by 27% and saved the company a projected $53 million annually in unnecessary claims costs.
7. Slash return rates
Businesses can also use AI in customer service to reduce the likelihood that people will want to return the items they buy. A top German e-commerce brand Otto primarily depends on AI to reduce returns. It uses predictive analytics to analyze 3 billion past transactions and 200 variables ranging from weather information to what people search for on the Otto website.
Since Otto doesn’t stock all the products it offers, its use of AI fits well with its business model. The solution could help the company keep track of which items customers bought most often, then keep only those things on hand. AI predicted what customers would buy a week before they made those purchases. The system can also tell with 90% accuracy what products the brand will sell within 30 days, which enables representatives to purchase the right amounts of products from suppliers. This reduced Otto’s surplus stock on hand by one-fifth.
The AI system reduced returned products by more than 2 million items a year. Plus, by using the predictive powers of AI to pinpoint the most-demanded products and keep them easily accessible and ready to ship, customers received their orders faster.
Exciting possibilities for using AI in customer service
If companies have a customer service obstacle to overcome, AI can likely help them tackle it. The examples here are diverse concerning the industries covered and the ways companies used AI. The common thread between them is that AI leads to more content customers and better-performing companies.
Want to read more related AI content? Download our artificial intelligence stats roundup for 2019!