A partial view of a businesswoman pushing wooden dominoes while an AI-driven robotic hand prevents the cascading dominoes - AI challenges concept
PHOTO: Shutterstock

This article is part 3 of a 4 part series, sponsored by Genesys.

Blending artificial intelligence (AI) and machine learning (ML) tools with human contact center agents can help you deliver the best possible customer experience. But getting started with AI and ML can feel like a daunting challenge, especially if your company relies on customer support tools developed years ago.

As with any technology transition, begin by asking key questions to help you identify goals and desired outcomes. What does the solution do? What problems will it solve for your business? How will solving those problems help the bottom line? How will you measure success?

Once you’ve answered those questions, you’re ready to begin the journey toward a blended AI-enabled customer experience platform. Here are the first three steps many businesses take to get started, beginning gradually with AI/ML basics and progressing into analytics and insights.

1. Find Out Why Customers are Calling

If you own a restaurant, you probably don’t ask the servers to greet patrons, find out if they have a reservation, and show them to an empty table. A host usually takes care of all that, so that your wait staff is free to do the most important work of serving customers.

A chatbot or voicebot can perform a similar role. For example, instead of human agents spending their time learning who the caller is and the reason for the call, your voicebot can greet callers and ask, ‘How can I help you today?’

Known as interaction steering, this AI-enabled feature helps you determine customers’ needs up front and then use that data to direct them appropriately to the next step. It saves customers’ time. It saves your human agents’ time by taking basic, mundane tasks off their plates. And it sets up the call for a successful outcome.

By analyzing the metrics behind these interactions, such as where incoming calls and chats are directed, you can help agents be more productive in the future. For example, of 10,000 chats in a given period, you might discover that 6,000 went to customer support, 2,000 were directed to your customer FAQ page, and so on — information that can help you more effectively optimize your contact center operations.

Related Article: Why Chatbots Are Worth Talking About

2. Route Callers to the Best Available Agent

Once you have a better sense of where callers need to be routed, the next step is to get them there with predictive routing.

Predictive routing forwards a call to the best agent in relation to the customer at that particular moment in time. For example, the system can look at more than the skills required to service that customer (such as language and technical competency) and also take into account the human agent’s disposition and how that has performed in past interactions with other customers.

One way to measure success of your predictive routing solution is to provide a quick survey for customers at the end of a call. If one of your goals is to improve your Net Promoter Score, your voicebot might ask callers if they would recommend your company after their interaction with an agent.

You might also ask if the agent answered your caller’s questions or met their needs. The data collected and analyzed can help determine if your automated interactions are achieving the desired result of improving the customer experience.

Related Download: Top Strategies for Workforce Optimization

3. Gain Insights From Real-Time Data

With interaction steering and predictive routing in place, your AI-enabled customer experience platform gathers real-time customer interaction data, providing valuable interaction analytics that show how to better optimize your workforce and the customer experience.

An interaction analytics software tool pulls in real-time data from all your customer engagement channels such as webchat, email, phone calls, SMS, instant messages, and social media and analyzes those conversations.

With interaction analytics, you can automatically identify interaction types and conversations with topics and categories, including sentiment and intent; dynamically understand if conversations have met with appropriate regulatory compliance requirements; gather employee and customer feedback; and discover conversational trends as they’re occurring.

The analytics can be accessed via an easy-to-use dashboard of interaction intelligence that lets you see, among other things, how (and if) desired keywords are being used in conversations, enabling you to drill down to specific agents and specific customer interactions.

Insights gleaned from analytics can provide many optimization opportunities. For instance, real-time data gathered from individual employee-customer interactions can help you identify where additional training is needed, so you can provide highly targeted, brief training modules. Training videos can be 15 seconds or less, which are easier to fit into busy schedules and are often preferred by the growing number of millennials and Generation Z employees in the workforce. In addition, if every agent only has an hour for training each week, it would be wasteful to show each person the same training content. By surfacing meaningful insights of each agent’s performance, you can customize that hour to the training each agent needs to provide better service.

Start With the Fundamental AI Building Blocks

There’s no need to dive head first into AI and ML. Just stick a toe in the water with the basics and don’t try to accomplish everything from the beginning. Simply identify your biggest business challenges that the platform can address and focus relentlessly on solving them. Start with the fundamental AI building blocks of interaction steering and predictive routing. And over time, start leveraging the platform’s real-time data and analytics to make real, tangible improvements to your customer experience.