The AI-Powered Enterprise. Seth Earley
Читать онлайн книгу.
are typically tasked with improving customer experience (for example, call center operations) do not always have the resources to get to the root of the problem and fix issues upstream. Call centers are incented to maximize productivity, not to actually solve the problems that are causing customers to call in the first place.
To solve the problem, eventually the root cause needs to be addressed. But all too often, the organization lacks the discipline to make the investment in finding the root causes of experience problems. Or the organization has tried solutions before but has been burned due to the complexity of the problems, the number of systems involved, interdepartmental dependencies, a lack of overarching governance, incorrect metrics, outdated or poorly selected technology, or an insufficient application of resources to the problem.
Many customer experience improvements are rooted in reducing the cognitive load on the customer—by simplifying menus, understanding customers’ intent, and finding a way to cut directly to the chase. If I had been presented with hundreds of choices for chef’s knives rather than one or two options, I likely would not have purchased any of them. The “paradox of choice” states that providing too many options increases anxiety and leads to second-guessing decisions and unrealistic expectations.
This is why simple interfaces presenting just the information needed in a particular situation reduce the mental work and allow for faster, easier decisions with higher levels of satisfaction. But it’s not at all easy to figure out how to best simplify the experience. The key lies with exercises that map the customer’s journey with the company and attempt to understand the customer, including their decision criteria, values, and considerations, along with the features, functions, and characteristics that are important to them. These components become a subset of the data attributes that represent the customer; those attributes must be included in the ontology so that systems can align the correct content with the signals gathered during their journey.
Just as a great salesperson reads verbal and nonverbal cues and body language to gauge what a customer needs, our digital customer engagement technologies read the customer’s digital body language and present what they need—no more, no less. We understand those needs and digital signals through journey mapping.
TOWARD A HIGH-FIDELITY MAP OF THE CUSTOMER JOURNEY
Executives believe they understand their customers and their customers’ needs. They most likely have groups of employees tasked with understanding and researching this exact topic. They may run focus groups, conduct user studies, analyze data in voice-of-the-customer surveys, and organize research teams.
One goal of all this activity is to develop a journey map: a high-level description of the steps the customer takes on the way to achieving their objectives. The customer journey:
•traverses multiple channels and touch points;
•interacts with every part of the business (throughout the product or service lifecycle);
•is supported by multiple departments;
•includes transactions across many systems and applications;
•is governed or managed through various processes and organizational structures;
•leverages models of the customer to varying degrees, including attributes, characteristics, and preferences from the ontology; and
•extends well beyond marketing, sales, and support and depends on all the other parts of the organization.
Maps of customer journeys are typically developed at a high level; the details of moment-to-moment needs are difficult to analyze, understand, and serve. Many customer journey maps are based on what the organization thinks its customers experience rather than what they actually experience. The reason for this is that primary research (talking to, interviewing, and observing customers in their actual environment) is expensive. Even when organizations undertake costly customer research, there can be biases in what the researchers are looking for or interpreting.
One revealing exercise is to compare a hypothetical understanding (getting executives in a room to chart out how they believe their customers interact with their organization) with one that is validated by actual users in as close to a real-life circumstance as possible. Many organizations employ usability testing labs, and although those environments attempt to simulate real-world conditions, they are, by nature, a simulation.
AI-powered customer experience takes this approach to an entirely new level. It does so by considering customers’ needs in their specific contexts as they go about their tasks and building from this information data models that represent moment-to-moment needs of the customer. We all play different roles throughout the day—parent, employee, boss, colleague, and customer of a range of businesses. Our needs change as our role and our immediate objective changes.
The High-Fidelity Journey Map Is the Keystone of a Better Customer Experience
Preparing for AI-powered customer experience demands a new kind of journey map: a high-fidelity journey map. While you may think you know your customer, you’ll know them a lot better with a high-fidelity journey map informed by AI.
No doubt your organization has developed customer journey maps at some level of detail from your research activities, testing, user focus groups, and internal working sessions. To keep things manageable, people who create these maps tell me things like this: “We are keeping it simple. We don’t want to overcomplicate things.” That is not a bad idea. “Don’t make it the Mardi Gras,” as a colleague of mine was fond of saying.
However, in practice, many customer journey maps lack enough specificity—they are oversimplified. Not only do they fail to include enough detail in the tasks that the customer needs to execute, but they lack a way of representing the customer stages and objectives in ways that computers can understand and act upon.
The solution is to create what I call high-fidelity customer journeys. Why “high-fidelity”? Well, beyond sounding cool and different and buzzwordy, the term actually means something. High-fidelity customer journeys are representations of the customer’s needs in data terms. The map of the high-fidelity customer journey models customers’ “attributes”—the descriptors and identifying features indicate their role, buying stage, interests, demographics, goals, and even state of mind.
Developing customer journeys that can convert what the customer is trying to achieve into things that the technology can present back to the customer presents a problem. We have to think through the “So what?” question. What does it mean when we say that the customer is from a particular industry or that they are trying to select one product or another? How do we represent (and how does it matter) when our customer is at the “choose” stage in trying to decide what to purchase? The process is logical and objective, but there is art along with the science.
Oftentimes, internal research groups and external agencies can miss the critical linkages of the customer experience to attributes that your systems can interpret and act upon. These linkages are mechanisms that allow the journey map to come alive within the systems and technologies within your organization. They inform the machine learning and AI tools and are defined as elements within the enterprise ontology.
To complete the high-fidelity journey map, you must validate the journeys through primary research—that is, through actual observations of customers, interviews, and simulations that prove out your assumptions and insights about what the customer wants and how they think about the world.
High-fidelity journey maps are validated models. They are distinct from other customer journey representations because they include a detailed, nuanced, and multidimensional understanding of various aspects of the customer. The high-fidelity journey map requires new and evolved ways of thinking. It also takes into consideration variations in use cases and makes those part of the ontology. This data allows AI and machine learning programs to assemble and optimize offers by recognizing signals that indicate the customer’s intent and context.
When high-fidelity journey maps are combined with customer attribute models (descriptors that represent the customer’s interests, needs, tasks, objectives,