E-Handbook: Advanced self-service platforms open new customer frontiers Article 3 of 4

phonlamaiphoto - stock.adobe.com

This article is part of our Essential Guide: Customer service technologies revolutionize CX engagements

The pros and cons of customer service AI

AI is growing deep roots into cloud-based support services, but there remains a question of whether it can provide the empathy and understanding that brand loyalty requires.

More and more, chatbots are being used as service desk utilities, replacing human support personnel. This is visible in the marketplace, as many support pages on commercial websites provide bot assistance and the industry integrates more self-service platform technology.

Consumers frequently are put off by efforts to obtain support over the phone, only to become lost in an endless interactive voice response phone maze asking them to "press this button" and "listen to these options," when all they want is a human. While automation often saves money for the enterprise, there remains a question of whether customer service AI is positive for the customer.

An indispensable human touch?

For some, chatbots are a step in the wrong direction -- a move away from human empathy and compassion that keep customers loyal even when things have gone south.

For instance, in a survey of 1,000 customers and service professionals, Cogito Corp. found that 80% of those responding said a personalized experience is crucial, and 97% considered it important that customer service take an authentic interest in solving their problem.

The survey concludes that better conversations lead to greater customer satisfaction and loyalty, and a competitive advantage.

However, vendors of chatbot platforms are incorporating personalization features into their bots to address this deficit. Salesforce, in particular, with its new Einstein Bots, is able to connect a service bot to customer profile data, giving the bot access to the customer's purchase and service histories, as well as enabling it to register customer sentiment to aid in fine-tuning its responses.

Those capabilities aren't yet common among chatbots, but may soon be -- after all, it's already routine for CRM service software to put customer profile and service history data at the fingertips of human service personnel.

But is that enough? To remain loyal to the brand, the customer seeking help needs that help to be more than just knowledgeable -- they need for it to be sympathetic.

chatbot chart

AI can make a chatbot seem more human

The question of whether the customer wants to chat with another human or a bot is by no means settled. Opinions abound: A PwC study said customers want more humanity and less automation, "especially when something goes wrong." And in the U.K., a MyClever customer survey rated human service over bot service, with the complaints that bots often have trouble with even simple answers and fail to meet expectations in friendliness.

For some, chatbots are a step in the wrong direction -- a move away from the empathy and compassion that keep customers loyal even when things have gone south.

On the other hand, a 2017 report from the International Customer Management Institute showed that 85% of survey respondents said they want to see their organization adopt or expand the use of AI. Moreover, in the 2018 State of Chatbots report -- a joint effort between Drift, SurveyMonkey Audience, Salesforce and MyClever -- 69% of consumers said they prefer chatbots for quick communication with a company.

Can AI help the bot take that final step, combining availability and speed with a humanlike touch?

The answer to that is yes, with qualifications. Bots must have a clear set of rules about when to elevate service to a human, and be given real-time sentiment analysis tools: With this emerging technology, chatbots' actions can be cued by a customer's anger, stress or confusion to shift conversational styles -- from cool and technical to warm and friendly, from humorous to snarky -- depending on what will set the customer at ease.

Most chatbot platforms now enable flexibility in configuring highly variable response styles, and for the most part, programming skills are not required. The sentiment analysis piece is not so easy: It's a third-party component, making compatibility a consideration.

There's no turning back. Customer service AI is here to stay -- and it is rapidly becoming the default.

Dig Deeper on Marketing and sales

Content Management
Unified Communications
Data Management
Enterprise AI