How to fast-track AI in customer experience

Service leaders need to start with the basics first—such as using AI in CX analytics—before moving to complex problems

Blog
Todd Krieger

Todd KriegerSenior Editor at Freshworks

Jul 25, 20244 MINS READ

Customers today expect instant response and a premium experience—regardless of industry, product, price, or channel of communication.

Yet despite dramatic advances (and adoption) of chatbots and other AI-powered tools, the latest Forrester survey reveals that overall quality of customer experience in the U.S. has sunk to its lowest point ever. 

What gives? We sat down with Murali Krishnan, senior VP of customer service at Freshworks, who explained why the future of customer service is brighter than ever but requires service leaders to think systematically about how, when, and where they apply AI to improve experience. 

Here are edited highlights of the conversation.

Why do so many customers consider CX quality so low at a time of huge improvements?

The biggest single reason is rising customer expectations. Customers expect seamless and efficient service, similar to what companies like Amazon provide. They want everything available at their fingertips and expect minimal contact. This means companies need to adopt new tools to meet these expectations.

However, many companies are still getting it wrong despite the availability of great tools specifically designed for customer experience. One major issue is that some companies have been applying AI as a blanket solution without properly triaging between simple and complex issues.

Some companies have been applying AI as a blanket solution without properly triaging between simple and complex issues.

This approach often leads to poor customer experiences. For example, when AI is applied indiscriminately, it might handle initial calls well but fail to address more complex issues. Customers end up frustrated when their problems are misrouted or when AI solutions are insufficient.

Read also: Why your gen AI projects are stuck at the starting line

It's crucial to apply AI in a systematic way, starting with simpler issues and then gradually addressing more complex ones. AI can do excellent analytics and quickly understand customer issues, but when it gets it wrong, it can be very wrong. Therefore, human supervision is necessary to ensure that AI solutions are accurate and emotionally intelligent. This means allowing AI to handle straightforward problems while human agents manage more nuanced and challenging issues. 

What can companies who are lagging behind in CX do to catch up quickly?

Companies need to first understand their unique challenges. For instance, newer SaaS companies might face complex issues that need careful handling. Generative AI can help rapidly build a strong knowledge repository and set up an effective triage system using AI. This means categorizing issues based on complexity and ensuring that simpler problems are resolved quickly while more complex ones are flagged for human intervention.

Second, they can implement clear policies and procedures. In a B2C environment, this is essential for AI systems to provide effective support. For example, AI can handle straightforward issues by following predefined protocols, but human agents should step in when the AI encounters more nuanced problems.

Subscribe to The Works

Insights on the business impact and ROI of AI. Sign up today.

Third, companies should maximize how they use AI for CX analytics. AI is excellent for initial analytics and understanding customer issues. Companies should use AI to analyze customer interactions and identify common problems. This data can then be used to continuously improve both AI and human performance.

It’s critical to remember, though, that while AI can provide quick and accurate solutions, it can also make big mistakes. Therefore, human oversight is necessary to review and validate AI-generated solutions before they reach customers. This collaboration between AI and humans leads to better customer experiences and ongoing refinement of support processes.

How, specifically, will generative AI improve customer experience over the next few years?

Several ways. First is providing much greater personalization in customer support. Today, we often group customers by company or user type, like admin or power user. In the future, AI will be able to personalize support to individual users within a company. For example, if a customer raises a query, AI can remember his or her past interactions and connect them with the same support agent or predict their next question based on similar interactions. This leads to a more tailored and satisfying customer experience​.

Next is efficiency. Gen AI can help create the next best action for human agents, making their work more accurate and efficient. By analyzing multiple symptoms and providing precise solutions, AI allows human agents to focus on more complex issues. This drives overall efficiency, reducing the time needed to resolve issues. AI will also support proactive customer service by predicting potential problems and addressing them before they escalate​.

Human supervision is necessary to ensure that AI solutions are accurate and emotionally intelligent.

Gen AI will also enhance multilingual support, providing resolutions and content in multiple languages, which is crucial for a diverse customer base. That, in turn, will help democratize knowledge by aggregating data from various sources like customer forums, communities, and support interactions.

What is your long-term outlook for AI in customer support?

The future of customer service with AI involves continuous improvement, where AI and human agents collaborate to provide the best possible experience. This synergy leads to better resolutions, increased efficiency, and highly personalized support. AI allows for proactive support by predicting potential issues before they arise.

It also facilitates continuous improvement through feedback loops, refining its problem-solving capabilities over time. This results in a constantly evolving and improving customer support environment. The insights will help the whole organization benefit from improving their products and services. Additionally, AI will enable the creation of more accurate and timely content for knowledge bases, making support processes more responsive and effective.