Article

AI for customers

Machine intelligence is shifting customer service from reactive to proactive

  • Proactive support systems can prevent service outages, not just minimize downtime when they happen
  • Virtual chatbots use AI algorithms to contextualize customer issues and offer solutions
  • AI-powered support tools at ServiceNow have improved customer satisfaction scores 10–15% and slashed resolution times threefold

While customers sleep, modern customer service systems are often busy with their most important work. Digital support platforms now routinely detect and resolve issues before customers are even aware of them—averting outages, minimizing slowdowns, and building an increasing level of customer satisfaction, around the clock.

The critical enabler of proactive customer support is artificial intelligence.

Customer service has been, by tradition, reactive: An issue occurs, and in response a customer sends an SOS by phone or email, or seeks help from a chatbot. But AI has greatly improved the outcomes of such reactive measures. Virtual chatbots, for instance, use AI algorithms to contextualize the issue, offer basic solutions, or direct the customer to a human agent, if necessary.

Proactive customer service holds even more promise. At ServiceNow, publisher of Workflow, we already use it to help customers avoid costly shutdowns. As the technology matures, we’re continually exploring how it can become more powerful and impactful. Much of that innovation is driven by AI, whose predictive powers can help create seamless, successful customer experiences.

Since we introduced AI-powered proactive support, our customer satisfaction scores are up 10–15%. We’ve reduced customer resolution times threefold, allowing service agents to focus on more complex or urgent issues.

Here are four AI-based capabilities we’re developing that could transform customer service over the next three years:

Proactive customer communication

Pattern recognition is AI’s superpower. When one customer solves a support issue, that fix can be applied automatically to other customers by analyzing the root cause and searching for matches across the customer spectrum.

Consider this example: A customer’s credit card gets hacked. AI will analyze the specific usage patterns and behaviors, such as where and when the card was used. Applying that analysis predictively allows support staff to flag other cardholders who may face the same security risk, and automatically trigger communications that warn them of the potential problem.

[Read also: How digital technologies & workflows are reinventing the customer experience]

Intelligent FAQs

Customers hate dead ends, especially when they need answers fast. But that’s what they get when you force them to search a long list of FAQs to find the information they need.

Rather than directing customers to FAQ pages that may not solve their problems, proactive support uses AI to pinpoint the exact sentence or paragraph within a help article that provides the information they need, when they need it.

This “wisdom of the crowd” can help train AI to get better at this over time. The more customers who have a similar issue and search for the same answers, the smarter the AI becomes at locating the relevant data.

Auto-scanning of case backlogs

Inevitably, cases arise that need a human touch. They end up in line for a support agent to investigate, which creates backlogs. Here too, AI can play a big role in speeding up solutions.

AI can scan case backlogs to provide automated solutions whenever possible. Its algorithms can also prioritize each case based on factors such as service agreement levels or the severity of outages for affected customers. That allows AI to escalate the most urgent issues.

Support integration across channels

Today, access to AI support is limited to proactive monitoring and virtual chat. Ultimately, it will be there for the customer any time, at any touchpoint. It’s about integrating support AI into all channels and extending its visibility across your entire ecosystem.

Say, for example, that a customer is upgrading to a new product. Support AI will know they bought the product two weeks ago. It will also know that other customers who bought the product have called with specific questions after two weeks. The next time the new customer logs onto the portal, a targeted FAQ pops up to address the issue.

AI can also assist human agents as they chat with customers online, providing all the context needed to resolve issues fast.

These examples are just the beginning of the customer service revolution. AI is continually improving. We’re excited to keep innovating with the singular goal of making customer service support faster, smarter, and more proactive.