Required Reading: AI Has Complete Power Over the Customer Experience
AS COVID-19 SPREAD across the country, customer service became more important than ever. Anxious customers require personalized, caring service, and companies that don’t deliver it will face even greater losses.
But no matter how well-trained they are, contact center reps can’t provide the service today’s customers demand if they don’t have the information they need at their fingertips. Poor data and slow, disorganized systems can prevent even the best employees from creating good customer experiences.
Seth Earley, founder and CEO of Early Information Science, argues in his new book, The AI-Powered Enterprise, that the key to providing customer service professionals with the tools they need begins with artificial intelligence, and the companies that get AI right are more likely to keep their customers coming back. CRM Editor Leonard Klie recently spoke with Earley to learn more.
CRM: In the book, you say AI is going to transform how companies create and deliver value to customers. How is this going to happen?
Earley: This will happen in many ways:
• greater understanding of user needs by reading the digital signals about how products are used;
• faster resolution of problems;
• more effective customer support;
• intuitive, voice-guided self-service options;
• wider use of the best approaches for customer solutions;
• finely tuned experiences that cater to the tastes, wants, and desires of customers and prospects; and
• lower costs and higher quality of products and services from new efficiencies in design, manufacturing, and delivery of solutions.
Many of these are less about big giant transformations than they are incremental improvements across multiple areas that ultimately add up over time to a transformation.
What areas of business will most be affected by AI? How will it affect marketing, sales, and customer service?
The greatest and most visible impact will be customer support and the customer experience. Most AI is in the background doing things like optimizing warehouse operations or personalizing offers based on recent purchases. But every aspect of how customers interact with the company will be affected.
If you have developed a customer journey map to illustrate how your customers interact, you will find that AI can be applied in each stage. The customer experience will improve any time we can make the information for the user more targeted and relevant. It’s as simple as that.
The customer experience can be summed up by getting answers to a series of “How do I…?” questions: How do I find Milwaukee tools on the website? How do I troubleshoot the Nest doorbell I just installed? How do I cancel my order? AI improves the user experience by serving up exactly what customers need when they need it. While this might not sound new, AI can do this when correctly configured and supported by the right information.
Think of the best customer experience you have ever had. It was probably from someone who knew you well, who you trusted, and who gave you exactly what you wanted, possibly before you knew you wanted it. It is impossible right now to scale that level of service to hundreds of thousands or millions of customers, but the future of AI will be exactly that—one-to-one relationships with all of your customers and prospects. Massive and effective personalization at scale.
How can AI turn customer service from a cost center into a revenue center?
In many organizations, customer service falls to people specifically assigned to that role or to others who have frequent customer interactions, such as members of the sales team. Ideally, when someone calls, the rep should have a complete understanding of the customer, including which products they already own and which prior issues have come up. Having that 360-degree view of the customer will provide opportunities for cross-sell and upsell as long as the company has earned the right to do so. Critical to this is trust. When customer service reps solve problems and provide the right information in a timely fashion and demonstrate a good understanding of both customer needs and the company’s solutions, they have a greater likelihood of being considered trustworthy and can offer alternatives and purchase recommendations. AI impacts each part of this scenario. Faster solutions can be accessed via a knowledge bot. A 360-degree view can be achieved using graph data with machine learning. A knowledge architecture supports faster information access and can make purchase recommendations that the customer will appreciate. Every interaction is an opportunity to improve the relationship. Getting the right products to a customer is a high-value service that can come through what is routinely considered a cost center—the call center. The people who answer questions for customers in a service role could also be selling them solutions as a trusted resource.
You also say that despite the billions spent so far on bots and other tools, AI continues to stumble. What’s holding it back?
AI has been successful in many cases, but there have also been many disappointments. Here are the key reasons:
• Unrealistic expectations: The market is hyped, and a lot of promises have been made that cannot be kept.
• Lack of supporting processes: Personalized content is required for a personalized experience, but many companies don’t know enough about the customer to truly differentiate content.
• Disconnected technologies: Many stand-alone efforts are not well integrated, further fragmenting the user experience.
• Lack of appropriate data and architecture: AI systems have to be taught about products, services, solutions, etc. The training content must be in the right structure, and the data must be good quality.
• Lack of success measures or unclear business objectives: Not knowing what success looks like and how it will be measured means project success will be a matter of opinion rather than determined by objective measures.
Other factors include needed skill sets, complicated tech stacks, and more.
What do you see as the major dos and don’ts with AI and customer service?
Do:
• Clearly define your objective—greater customer satisfaction? Increased revenue per customer?
• Measure baselines and project the expected benefits.
• Continually make course corrections based on measures and user feedback.
• Have a narrow scope and modest goals for initial projects.
• Carefully examine data sources to ensure they will be readily available at scale.
• Use an executive sponsor with a track record of success and organizational clout.
• Ensure that there is adequate funding given the scope.
• Build detailed use cases with personas representing various customer archetypes and their tasks.
• Test against those use cases.
Don’t:
• Set unachievable goals for initial projects.
• Underestimate required budgets.
• Ignore cultural issues, including socialization and change management.
• Outsource core capabilities to an AI vendor.
• Rely on non-specialists for the information architecture needed to support your effort.
• Learn as you go for mission-critical programs.
What will companies need to do differently to capitalize on AI?
It’s the basics of good information hygiene. If your salespeople don’t use the CRM system, is that the tech organization’s fault? It could be that the technology was poorly configured, but the bigger issue is getting buy-in and communicating the benefits of the technology to the sales team. Similarly, pay attention to exactly which process you are trying to address and enlist a range of impacted stakeholders. Companies have to spend more time and attention on where they get their critical data, who owns it, who manages it, and who impacts and is impacted by it. Providing users with insights about the impact of good and bad data on their jobs, their customers, and especially the organization will make the issue of data quality more meaningful. The business side needs to own its data (in partnership with IT) and take responsibility for its quality. Demonstrate this so it sinks in. They need to understand that this is a shared responsibility. Show what is possible with the right applications and quality inputs and get them excited about the possibilities. Data needs to be treated as the true business asset that it is. The entire digital experience is comprised of data.
For some applications that have AI in their inner workings, the benefits can be quite immediate. In the majority of cases where organizations are not simply deploying something out of the box, it can range from months to years. In most cases, the biggest holdup is from data issues.
What is the main point you want people to take away from reading this book?
I want people to understand that AI success requires certain things to be in place that are less about AI and more about IA (information architecture). People need to understand that AI is a shared journey between business stakeholders, IT, senior leadership, and the customer. I want to demystify AI and show how its success is predicated on tried-and-true information management principles and that there is no getting around those principles. AI is complex but not mysterious. Even though the inner workings will be delegated to technical staff, businesspeople and senior leadership can and must understand the business value, objectives, functions, and success factors for AI.