In the rapidly evolving world of MarTech, few are as deeply immersed in the intersection of technology and customer experience as Aisha Amaira. With a rich background in CRM marketing technology and customer data platforms, Aisha brings a unique perspective on how businesses can harness the power of artificial intelligence to not just solve problems, but to build lasting relationships. In our conversation, we explore the stark divide emerging between companies that are scaling AI effectively and those falling behind. We delve into how “agentic AI” can deliver surprisingly empathetic interactions, the practical steps to escape “pilot purgatory” and achieve real impact, and the profound organizational shifts required to truly unlock AI’s potential. Finally, we look to the future, envisioning a new role for human agents and a fundamentally transformed customer journey.
Your research identified a growing divide between “leaders” scaling AI and “laggards” sticking to human-only models. What specific capabilities are these leaders building to separate from the pack, and what metrics are they using to prove the ROI of moving towards full-service AI care?
It’s a fascinating and crucial distinction we’re seeing. The leaders, which our recent survey identified as the top 10 percent of organizations, are truly separating from the pack. They’re not just dabbling; they are aggressively building what I call an “AI muscle” to scale across their operations. They see AI as the future of what full-service care can be, pursuing tangible gains not just in efficiency but in customer experience and even sales growth. In stark contrast, the laggards, the bottom 30 percent, seem almost hesitant to let go of the familiar, what they see as the safety of the human-only model. The metrics leaders are using are incredibly clear and impact-focused. They’re looking at goals like automating up to 70 percent of all customer contact or ensuring that 80 percent of all tasks are fully resolved in under 120 seconds. These aren’t vanity metrics; they represent a radical difference in speed and efficiency that directly translates to both customer satisfaction and a healthier P&L.
The concept of “agentic AI” has been compared to LEGO blocks building complex workflows. Beyond simply resolving issues faster, how can this technology create genuinely empathetic experiences? Please share an example of how an AI can demonstrate understanding and adjust its tone based on a customer’s needs.
That LEGO block analogy is one of my favorites because it perfectly captures the essence of agentic AI. Each block is a small, specialized AI worker designed to do one thing exceptionally well—fetch a document, schedule a technician, check for compliance. When you connect them, they form these incredibly complex and intelligent workflows. But the magic, the empathy, comes from how they operate. True empathy in AI isn’t about feeling emotion; it’s about demonstrating understanding. Imagine a customer calling about a sensitive issue for the second time. An agentic AI can instantly recall the previous interaction and adjust its approach. It can change its pitch, its tone, and its speech patterns to reflect that it understands the customer’s history and potential frustration. It might say something like, “I can see you called about this last week, and I want to make sure we get this right for you today.” It never gets frustrated, never has a bad day, and can always maintain a calm, even keel. This creates a surprisingly delightful experience, one that makes a customer pause and think, “This is different,” and begin to trust the interaction.
Many organizations find their AI initiatives get stuck in “pilot purgatory” with little measurable impact. Based on the DoorDash example of setting a clear 70% success metric, what is the right way to structure a pilot for success, and when should a company bypass the pilot phase altogether?
“Pilot purgatory” is a very real and frustrating place for so many companies. We’ve seen reports that as many as 95 percent of generative AI pilots deliver little to no measurable impact, and that’s often due to poor structure. The first question I always ask a company is, “Do you even need a pilot?” Sometimes, the technology is proven, and you can move straight to implementation. A pilot is truly for experimentation, when there’s a genuine question about whether something can be done. The DoorDash story from 2019 is a perfect example. They came to us with a clear, measurable goal: “Can you successfully automate 70 percent of our outbound food order calls?” We didn’t know if we could, and they didn’t know either. That’s the ideal scenario for a pilot. It had a razor-sharp success metric—70 percent—and it was a true experiment. We ran it for two weeks and hit 74 percent. That clarity removes all ambiguity. Without a specific, business-impact-focused KPI like that, a pilot becomes a vague science project that’s doomed to fail or just fade away.
It’s suggested that the primary barriers to large-scale AI automation are organizational inertia and a lack of open APIs, not the technology itself. What practical steps should a CIO take to prioritize API development, and how can leaders foster a culture that embraces redesigning core processes?
Absolutely, the technology is no longer the main limiting factor. The real roadblocks are internal. I cannot stress the importance of APIs enough; it’s a mundane topic, but it is the absolute foundation for advanced automation. If I were a CIO today, my top priority would be creating a comprehensive roadmap to open up every single user interface field and action that a human agent currently has access to. Every action an agentic AI needs to take on behalf of a customer, whether it’s updating an address or processing a return, requires an API. These are often deprioritized, but without them, you’re stuck with simple FAQ bots that nobody wants. This needs to be a C-suite conversation. Fostering the right culture requires top-down leadership. It means leaders must be prepared to challenge long-held norms, ask tough questions about risk tolerance, and combat the mindset that “customers won’t adopt this.” You have to bring together leaders from business, product, and tech who truly believe this is the future and are excited to collaborate on redesigning processes from the ground up.
Companies often focus on using AI to mimic tasks humans already perform. How can leaders shift their mindset to innovate entirely new customer interactions, such as proactive check-in calls? Could you outline a process for identifying and launching these previously unscalable, value-add services?
This is where the real transformation lies. Simply automating what humans already do is an incremental improvement, not a revolution. The key is to ask, “What valuable things could we do for our customers if we had unlimited, elastic, and inexpensive capacity?” For example, a large moving company would never have the human capacity to call every single customer 30 minutes into their move to check on their experience. But an AI can. Imagine the impact on customer satisfaction and the reduction in post-move complaints from that single, proactive, intelligent conversation. The process starts with that mindset shift. First, identify your core business goals—like improving retention or reducing complaints. Then, brainstorm new interactions that could achieve those goals, completely ignoring the limitations of a human-powered model. Finally, map those new interactions to the capabilities of agentic AI. Don’t just digitize old processes; invent entirely new moments of value that were never before possible.
As AI automates most transactional queries, the role of the human agent will fundamentally change. What does the new “customer success manager” role look like in a B2C context, and what new skills and career paths must organizations build to support this critical evolution?
This is one of the most exciting shifts happening. As AI handles the transactional, the repetitive, the mundane, the human role becomes elevated. The vision I’m hearing from forward-thinking retail CEOs is to transform their agents into something more akin to a B2B customer success manager. In this model, a human agent isn’t just answering random calls. Instead, they are assigned a dedicated portfolio of, say, 100 to 150 of the brand’s top customers. They get to know these individuals, understand their history, and build genuine connections that an AI can mimic but not truly create. This requires a profound organizational change. You have to start now to identify the agents who have the potential for this kind of relationship-building role. You need to create new training programs, new career paths, and new compensation models to support them. People-changes take time, so you have to begin investing today in hiring and training the kind of talent you’ll need in your 2027 contact center.
What is your forecast for customer care? Looking ahead to 2027, what will be the most significant, tangible difference a customer will experience when interacting with a leading company versus a laggard?
Looking ahead to 2027, the difference will be night and day. When a customer interacts with a laggard, they’ll still experience the friction we know today: waiting on hold, being transferred, repeating their issue to multiple people. It will feel outdated and frustrating. In stark contrast, when that same customer contacts a leader, their experience will be almost instantaneous and effortless. The vast majority of their needs will be resolved in under two minutes by an intelligent, conversational AI that understands their history and speaks to them naturally. If they do need to speak with a human, it won’t be a random agent in a queue. It will be an elevated “customer success manager” who knows them, is empowered to solve complex issues, and is focused on building a long-term relationship. The tangible difference will be the complete absence of friction and the feeling of being personally known and valued, whether you’re interacting with AI or a human expert.