How to Make AI Work With the Right Data Points

Before working to onboard new clients and ensure their continuous success and growth at Next Caller, I worked in the ultimate customer service role: as a Middle School Teacher and Dean of Students. Building school culture requires being a servant leader. You need to exhibit authentic listening, empathy, problem-solving, understanding, and seeking to understand on a daily basis. A strong school leader must truly believe that they serve the students, the families, the teachers, and the community in which they work. These skills and mindsets are immediately transferable to the world of customer service, and no skill moreso than a teacher’s ability to take in dozens, if not hundreds, of data inputs to make a real-time, personalized decision.

Take, for example, the moment a child walks into a classroom. In that moment, the teacher is noting all of the following things (among others):

• What is the child’s posture?
• Are the child’s clothes clean?
• What is their demeanor?
• What is their reading level? What is the reading level required for today’s lesson?
• What do I know about this child’s special needs?
• Do I believe this child has eaten today?
• What makes this child laugh? What brings them joy?

…and then, that teacher makes a series of instantaneous decisions that will impact that child’s entire day. What kind of greeting to use. How to alter the opening activity. The frequency of one-on-one check-ins during the lesson. This personalized, real-time experience is possible because of the teacher’s ability and capacity to isolate the most important inputs regarding a child’s day, and use that information to alter teacher behavior to meet the needs of the student. The portrait of perfect customer service.

As AI begins to play a larger role in the customer service arena, it is more important than ever to make sure that we are maximizing the impact of revolutionary technologies by ensuring that they are making decisions based on the right inputs. AI in customer service often gets used as a catchall phrase. To drill down, when I say AI, I mean the full spectrum of assistive automated services from chatbots to machine learning to speech analytics to agent coaching. You don’t need to look any further than the victory of DeepMind-built AI over Go Champion Lee Sedol to see the incredible strides that have been made in this technology, and how this tech can not only refine, but fundamentally alter our practice.

However, to be a viable solution in the contact center, we need to be certain that we are making valuable, high-stakes decisions based on the correct inputs. So what data points are essential for maximizing AI’s application and impact in the contact center?

Identification

Just as a teacher takes all of their background knowledge about a child into account when deciding how to greet them, so too must any agent, AI or otherwise, identify who they are talking to before engaging in any decision-making. It is important to identify a customer in the channel within which you’re communicating with them as well as across all channels.

Is this call related to an inquiry that just came through Twitter? Can this email be linked to a Facebook interaction we just had?

Before any advanced analytics or intelligent decision-making can take place, it is imperative that businesses identify to whom, exactly, they are speaking.

CRM Communication

How robust is your CRM, and can your AI offerings leverage that information for advanced, real-time decision making? In order for AI to use your desired inputs, you have to be sure that those inputs are there. While this sounds intuitive, too many businesses put the cart before the horse, investing in AI services without the necessary data that serves as the fuel for the fire. The most effective AI offerings are able to use CRM information in real-time: how can this inquiry be tied to past reviews, past purchases, or past offerings. Essentially, accelerating the rate at which you can use the information you have painstakingly collected to personalize in-the-moment.

For example, how much impact could an airline make if it builds business rules surrounding a customer who calls within X hours of a flight? Or if a shipping logistics company were to do the same for a customer who tweets within X days of an expected delivery? AI doesn’t replace the human element – it allows the human element to execute faster, and potentially for the human element to learn completely new and creative ways of approaching a customer need. As industries begin to engage in even deeper data collaboration or leverage enormous common data repositories, the opportunities for personalizing based on past experiences will grow exponentially.

Analytics

Integrating AI into your contact center isn’t the endgame – it’s not a box to be checked on the “21st Century Thought Leader” scorecard. Any successful contact center leader who is leveraging AI will be able to answer this simple question: Is it working? Is it allowing our team to get smarter about interactions and drive customer happiness?

In planning out how to assess the efficacy and impact of any number of AI offerings, reflective leaders ask the following questions:

• What is our feedback loop? How are we integrating customer feedback on their experience with AI?
• How are we “checking the work” of our chatbots or AI offerings?
• When we brag about “machine learning”, are we certain we’re “learning” with the right inputs?

The teachers I discussed before weren’t born with the ability to read a child and respond immediately – they built that skill over years of incorporating feedback, honing their craft, and searching relentlessly for the right data points to maximize their warmth, love, care, and ultimately, student outcomes. As contact centers increase their use of AI, the goal shouldn’t be automation or cost reduction, but rather agents and AI offerings that have the same empathetic, customer-focused decision making abilities that teachers have.

Learn more about the latest innovations in customer service. Register now for ICMI Contact Center Demo!