Empowering Customers to Take Control of Their Journeys
Through AI-powered Personalization and Automation
Contributions by Vas Alli, Chief of Staff & CX Strategy Lead for Verizon
Chief of Staff & CX Strategy Lead
The use of Artificial Intelligence (AI) for personalization and predictive automation has been around for some time, but has accelerated to the top of organizational priorities since the global pandemic. Not surprisingly, investments in these technologies have increased exponentially as automation became a necessity overnight. In fact, the call center AI market is forecasted to grow from 800 million USD in 2019 to 2.8 billion by 2024.
Despite this growth, many organizations are still in the early stages of implementation, with three out of four respondents in a recent Harris Poll survey stating they have been working with AI for two years or less.
“CREATING AUTOMATION AND AI-DRIVEN SOLUTIONS SHOULD NOT BE SIMPLY TO DRIVE COST OUT OF THE BUSINESS. IT IS ABOUT EMPOWERING EACH CUSTOMER TO TAKE CONTROL OF THEIR OWN JOURNEY.”
– Vas Alli, Chief of Staff & CX Strategy Lead for Verizon
The purpose of this article is to establish the foundational principles for AI-driven personalization and predictive automation, the underlying key to success in contact centers, CX use cases, and what organizations should prioritize as they continue their digital transformation journey. It includes valuable insights throughout from Vas Alli, Chief of Staff & CX Strategy Lead for Verizon.
In contact centers, cost reduction has often been the primary driver for investing in any type of process optimization, including AI/automation. While lowering the cost-to–serve has historically been the primary investment criteria, this paradigm is shifting. According to the same Harris Poll survey, improving the CX was cited as the leading driver for AI implementation by 53% of respondents, ahead of reducing costs, cited by 48% of respondents.
By putting customer experience ahead of cost reduction, true transformation can and will occur and the efficiencies will follow. This is a major signal for organizations to think differently about their AI strategy, re-focusing their efforts on prioritizing CX, even ahead of straight cost transformations.
Vas emphasizes that anyone on a digital transformation journey should always ensure that the customer is at the center of every decision.
DEFINING AI, PERSONALIZATION, AND PREDICTIVE AUTOMATION
If you asked ten leaders how they define AI, personalization, and predictive automation, you would easily get ten different definitions for each. To be fair, the “definitions” have also changed over the years as the technology and use cases have evolved. What has not changed is that no matter the definition, these technologies
enable better experiences for both customers and employees.
For purposes of this article, we will define each as the following:
ARTIFICIAL INTELLIGENCE (AI) is the means by which machines leverage large (often disparate) sets of data and generate either insights for decisioning or proactive business processes to support customers and/or agents. Simply put, AI is a tool that can do things a human can do – but at a scale, precision, and speed not achievable by humans.
AI-Driven Personalization is the output of leveraging those insights to deliver a unique experience specific to that customer. It is the ability to make offers, provide information, and adjust the interaction based on their specific needs, behaviors, and/or sentiment. This can be accomplished by directly curating the customer’s experience, or via agent augmentation to better assist the customers at any point in their journey.
Predictive & Proactive Automation is the ability to leverage AI (at scale) and “automagically” implement that within the context of the interaction. Predictive automation is exactly as it sounds – using data and analytics to predict what the customer needs and determining how best to leverage that insight. Proactive automation is taking that data and actively taking the next step to augment the customer or agent experience.
What does this look like in practice?
Personalization can mean different things. The following bullets provide a few examples:
- The simplest use case is greeting a customer by their name.
- One-to-many personalization is what one might think of with traditional marketing approaches, in which outgoing marketing efforts are targeted and optimized, based on demographics, customer type, product preferences, and the like.
- At scale, brands can leverage AI to curate data about the customer’s preferences and purchase history to make a personalized offer specific to them. This is 1:1 personalization.
- Or assume product shipping times have been delayed, AI can predict which customers will be contacting the organization. Proactive measures can be taken to notify the customer in advance, and information about the delays can be added to inbound contact channels such as IVRs, an app, websites, and chatbots.
- Similarly, perhaps a mobile customer has added a new service to the account, and this is the first billing cycle since the account change. AI anticipates the customer may call about their bill and serves up information to the customer or agent. As the agent receives the call, they immediately anticipate the reason for the call so they can quickly address the issue. Or, in anticipation of that call, an outbound contact is made to the customer. For example, “I know you recently added services to your account. Since this is the first billing cycle, I wanted to ensure you understand your invoice.”
Across all these examples data is leveraged to personalize the interaction and ultimately create a better overall experience for the customer.
“YOU ARE CURATING AN EXPERIENCE THAT IS FOR THE CUSTOMER, NOT YOUR BRAND … THE WINNING ASPIRATION IS TO ENSURE THE EXPERIENCE IS SIMPLE, PERSONALIZED, AND EFFORTLESS.”
– Vas Alli, Chief of Staff & CX Strategy Lead for Verizon
THE KEY TO SUCCESS
Vas reinforces that the intent of any implementation should be to create truly exquisite experiences that reduce friction and make it easier to do business.
To make this a reality through a customer-first strategy, Vas explains that there is a delicate balance that must be considered when developing use cases.
This balancing act must take into consideration three vectors when designing these solutions:
The level of each will dictate which channel of interaction is best suited and how the experience is orchestrated to deliver the least amount of friction for the customer.
Assume a transaction is high value, complex, and potentially very emotional. During the pandemic for example, tragically many were out of work, impacting their ability to pay their credit card bills. Customers contacting their credit card companies likely had a high degree of stress, and the transaction itself was clearly high value and imbued with complexity. Forcing these customers through a digital self-serve channel is not the ideal solution. But what would be helpful is for AI to provide information to agents to make this as smooth and delicate as possible for the customer. This is a much different journey than one with less emotion or complexity, such as wanting to know the minimum payment due.
Developing and maintaining customer trust is the lynch pin to the success of any strategy. Customers with high value, complex, and emotional transactions are likely to distrust an AI-powered solution. Even if it is a top call driver, it does not mean customer-facing automation is the right answer. Rather, enabling agents to better serve those customers will go much further and be more appropriate.
Customer trust can erode quickly and be difficult to regain, so customers must be given a reason to trust the brand. The way to accomplish this is understanding the emotion, complexity, and value of interactions before determining what and where to automate.
IMPLEMENTING FOR LONG-TERM SUCCESS
There is increased pressure by organizations to stay ahead of the competition.
Many brands were at some stage of their digital transformation journey prior to the global pandemic, but this black swan scenario acted as a catalyst for accelerating these strategies. There is no reason to believe that this trend will not continue, and in fact, digital transformation will likely continue to accelerate.
Organizations should have a clear rhyme and reason for why they are deploying AI and which interactions they are automating—never losing sight of their intent: to empower customers to own their own journeys.
With that being said, how do organizations ensure they are set up to succeed?
- Align the Culture With the Vision — Peter Drucker’s well-known phrase, “Culture eats strategy for breakfast,” certainly applies in this context. The organization must have the right culture, with leadership behind it, to drive this type of transformation. If the organization is not ready for the required investments, or is not deeply committed to ensuring the customer’s journey is an elegant one, the best strategy will not succeed, and may do more harm than good.
In this Forbes article, Kartik Hosanagar, Professor of Technology, Digital Business and Marketing at the Wharton School and author of the A Human’s Guide to Machine Intelligence, was quoted
“What separates the AI projects that succeed from the ones that don’t often has more to do with the business strategies organizations follow when applying technologies than the ability of the technology itself to transform the business. Many of the problems are less about the tools and more about leadership. Most of the failures to harness the power of AI lies in human behavior, management understanding, and the failure to mesh algorithmic capabilities into organizations, business models, and the culture of the business.”5
So, organizations must be honest with themselves and determine if the culture and organization is ready. If they are not, there is a very real risk that their transformation will not even get off the ground and possibly even fail.
- Specify the Problem — Ensure the organization has clarity and agreement on the actual problem that needs to be solved. A common pitfall is organizations attempting to address a problem that should not be solved. For example, take the high emotion / high value / high complexity transactions discussed above. While a customer-facing solution may seem like the best implementation, consider this from the customer’s perspective. The more appropriate solution may be to use AI not to remove redundancy but to augment call center staff and help the agent be more focused on providing thoughtful/empathetic responses.
Customers are increasingly demanding more empathetic brand interactions and AI can play a pivotal role in giving your agents superpowers, by providing them with relevant information faster or even automatically running diagnostics/troubleshooting. Either way, the goal is the same – to provide the most elegant experience for your customer, how you go about doing that is the key.
- Clearly Articulate the Vision and Strategy — Once the problem statement is articulated, clearly define the vision for the desired experience. In other words, what does that new and improved experience look like for the customer? Once there is alignment on the vision, develop a clear articulation of how that vision will be realized. No organization has unlimited budget or resources, so developing a clear strategy with success measures is critical. The challenge, which will be discussed more later in this article, is to ensure the strategy is not overly complex, so much so that the implementation is delayed or, worse, never started.
- Create an Organizational Structure that Supports the Strategy — The days of operating in silos are over, and AI automation efforts are not for the faint of heart. It requires the right leadership and organizational structure to manage the complexities of developing and launching a truly transformational digital program. In this Forbes article and accompanying survey, only 5% of companies surveyed currently have a chief AI officer. However, two-thirds of respondents expect their organizations will create a C-suite role who owns the AI strategy for the company within one to three years. Additionally, many companies are creating supporting roles, such as AI ethics managers. The figure below shows the roles leaders are prioritizing. In short, ensuring the right organizational and leadership structure is essential.
- Establish Specific and Measurable Success Metrics — For any strategic initiative, establishing what success looks like is critical. How does the organization know if the AI/digital transformation efforts are successful? If the vision is simply to deflect calls out of the contact center, one measure of success may be the actual call deflection rate of that specific call type, successfully diverted via automation. Note the word “successfully.”
For any personalization or automation initiative, Vas reminds leaders, “The goal should be to reduce the cognitive load of the consumer and alleviate any points of friction in that consumer’s journey. Ask yourself, ‘Did we make it easier to do business with us?’” This would be measured via voice of the customer to understand satisfaction and level of effort, in addition to internal data/analytics that measure the success of the solution.
- Start off small, remember it is a marathon, not a sprint — It can be tempting to “go big” with these transformation efforts. Senior leadership for example may issue aspirations to “revamp the E2E customer support journey.” When starting off, the best approach is to start small, gain experience, and realize quick wins. Begin with low-risk investments, creating pilots to experiment before rolling out to the broader organization and prior to transforming more complicated tasks. Continually monitor, measure, calibrate, and adjust to refine that experience when it is evident that customer adoption is low, it is not working as designed, or the result is not meeting objectives. This does not mean organizations should not “think big,” but instead select a few meaningful use cases that will demonstrate success and prove the value. By gaining quick wins with small projects relative to the organization, confidence is built in the strategy and key learnings are gained for tackling larger-scale transformations.
Leveraging AI to empower customers to own their journeys through personalization and automation is here to stay. While most organizations are early in their journey and it may be a daunting endeavor, the longer organizations wait to evolve, the harder it will become. Professor Hosanager was also quoted in the Forbes article as saying, “The key message is leaders need to understand enough about how AI works to strategically align it with value creation and make smart investment decisions.”
He sums up the message, reiterating what is outlined earlier, advising leaders to:
- View AI as a tool, not a strategy
- Take a portfolio approach to AI projects that balances quick wins with fundamental process redesign
- Grow the talent base by re-skilling existing employees and hiring new talent
- Focus on the long term by sticking with AI through inevitable early failures
- Be aware of new risks AI can pose and manage them proactively
Throughout all of it, Vas reminds leaders that these transformational agendas are not for the faint of heart, and those brands who maintain a long-term commitment to the customer experience are the ones who will ultimately succeed.
- https://www.marketsandmarkets.com/Market-Reports/ call-center-ai-market-263925467.html
- https://www.globenewswire.com/news- release/2020/02/27/1991835/0/en/Study-Reveals- Hidden-Drivers-of-AI-Adoption.html