Bridging AI Strategy and Execution in Customer Service

Omid Razavi

May 5, 2025

I had the opportunity to connect with Nick Clark, Partner and Associate Director at Boston Consulting Group (BCG), during Kore.ai's Reimagine 2025 conference in Orlando, Florida. Nick had just participated in a panel unveiling the joint Kore–BCG whitepaper, "Beyond AI Islands: Why Organizations Must Unify Disconnected Tools to Decode the Art of the Possible."

Our conversation was especially timely as we look ahead to the 4th CSS Executive Forum in Central London on June 18th, 2025. Nick and his team at BCG Europe have been valued partners of our London forums since our inaugural event in 2022.

Omid Razavi: Nick, since we began collaborating, there's been a clear shift in how customer service and support (CSS) organizations approach AI. The conversation has moved from exploring potential to executing at scale. Many leaders juggle multiple initiatives while facing real pressure to transition from pilots to production. Given your broad client work and industry insight, including your contributions to our CSS forums, are you also seeing this shift? And what actions would you recommend for organizations navigating it without stumbling?

Nick Clark: You're exactly right, Omid. The urgency has grown, and leaders are expected to show tangible results from AI investments. That means setting clear, realistic short-term goals that build executive confidence while laying the groundwork for long-term strategic outcomes.

A key insight from our Beyond AI Islands whitepaper is the rise of governance as a design priority, not an afterthought. Over 80% of organizations now prioritize governance early in the process, implementing centralized guardrails, role-based access controls, audit trails, and policy enforcement. These steps help accelerate AI deployment while mitigating risk and ensuring compliance.

We're also seeing a necessary pivot from isolated pilots to platform-based implementations. While pilots help learn and generate momentum, scaling requires consistency, efficiency, and reuse. Organizations that embrace platform thinking can reduce costs by up to 30% and significantly compress deployment timelines. Strong governance and standardized frameworks are essential to driving that consistency and unlocking enterprise-wide value.

Omid Razavi: Exactly. And isn’t there growing scrutiny from boards and the C-suite? They want rapid ROI but remain extremely risk-averse. It creates what I’ve come to call the “AI Catch-22.” On one hand, leaders are being pushed to innovate boldly, adopt generative AI, and leapfrog competition. On the other hand, they are given little room for error—any failure could mean budget cuts, reputational risk, or loss of internal trust. It’s a paradox: “Move fast, but don’t break anything.”

We’re no longer in the era of AI curiosity or sandbox experimentation. It’s about measurable business value—delivered safely, ethically, and at scale. The tension between speed and safety is stalling progress in many organizations. How do you advise CSS leaders to navigate this tightrope?

Nick Clark: It's real tension. The best path forward is to frame a dual narrative: one that delivers quick wins, like better routing or reduced turnaround times, and another that sets expectations for strategic transformation. Transparency is key. Leaders need to demystify the roadmap for their executive teams and anchor it in achievable outcomes that build trust over time.

Omid Razavi: At the conference panel session, you emphasized the importance of platform thinking and warned against the risks of fragmented AI efforts. Why is platform thinking so critical for scaling AI successfully, and what key elements should organizations prioritize to make it work?

Nick Clark: Absolutely. Platform thinking is foundational if you want to scale AI effectively. Many companies rely on disconnected tools and pilots, leading to inconsistent experiences, governance gaps, and stalled progress. Platform thinking brings structure and repeatability across your organization. Instead of reinventing the wheel for every use case, platforms let you reuse components, maintain consistent quality, and accelerate innovation. They also simplify oversight and governance, which are increasingly critical.

Omid Razavi: One of the toughest challenges with AI in customer-facing roles—especially in sensitive sectors like healthcare or finance—is preserving empathy. How can organizations ensure that automation doesn’t come at the expense of the human touch?

Nick Clark: Empathy must be embedded in the design of AI systems from the start. This means building AI that’s contextually aware and emotionally intelligent. Organizations can recognize when a human touch is needed using real-time sentiment analysis and intent detection. Seamless, intelligent handoffs to human agents ensure that empathy isn’t lost, so customers feel heard, valued, and genuinely supported throughout their journey.

Omid Razavi: That resonates deeply. At SupportLogic, I’ve seen firsthand how customers successfully apply real-time sentiment analysis to proactively identify escalations, spot at-risk accounts, and prioritize support cases—not just based on SLAs, but on emotional urgency. This layer of insight, driven by AI and signal extraction, helps support teams respond faster and more thoughtfully, often before a customer even expresses frustration. It’s a strong example of how AI can enhance, not replace, human support.

Do you suggest that AI be held to the same quality standards as human agents when interacting with customers?

Nick Clark: Without question. Organizations need consistent performance metrics across AI and human agents to deliver a seamless, high-quality experience. That means evaluating communication, tone, accuracy, and escalation handling using the same rubrics—because from the customer’s perspective, the standard of service should never change, regardless of who—or what—is delivering it.

Omid Razavi: Let's talk about workforce management. Workforce management has traditionally focused on forecasting and scheduling human agents. Now that AI is becoming a core part of the service workforce, how is the WfM model evolving to reflect this shift?

Nick Clark: We’re seeing a shift from operational scheduling to real-time orchestration. It’s not just about managing shifts anymore—it’s about dynamically allocating work between human and AI agents based on capacity, context, and demand. This orchestration model helps optimize performance, improve responsiveness during peak periods, and align service delivery more closely with customer expectations and business priorities.

Omid Razavi: That's a fundamental change. As AI becomes an integrated part of the service workforce, new roles must emerge to support it. What specialized roles are you seeing, particularly with the rise of agentic AI?

Nick Clark: We're seeing several new roles take shape, which will increasingly become vital to scaling AI responsibly while delivering superior customer and employee experiences.

AI Trainers teach and fine-tune AI models with contextual data and feedback loops. They ensure the AI outputs are accurate, brand-aligned, and continuously improving.

AI Governance Leads establish policies and guardrails to ensure responsible, compliant, and ethical AI use. They focus on access controls, risk mitigation, and trust-building across the business.

AI Journey Designers orchestrate seamless, human-centered interactions between AI systems and people. They define workflows, map touchpoints, and ensure consistency and empathy across digital and human channels.

Omid Razavi: That's a great breakdown, Nick. These roles aren't just supporting AI—they're enabling it to be sustainable and scalable. AI Trainers and Governance Leads ensure AI is effective and trustworthy, while Journey Designers serve as the bridge between great tech and great experiences. As these roles evolve, they're redefining org structures, career paths, and how we deliver value.

One last question. We've long discussed the move from reactive to proactive—and even preemptive—customer service. With the convergence of predictive analytics, agentic AI, and real-time orchestration, this vision finally seems within reach. Do you expect this shift to accelerate significantly over the next 6 to 12 months? And what will distinguish the leaders from those left behind?

Nick Clark: Without question, it's gaining momentum—especially in sectors like financial services, healthcare, and retail, where customer expectations are high, and tolerance for delays is low. The enabling tech is there: real-time data ingestion, intent prediction, and journey orchestration are now feasible at scale.

What will separate leaders is their ability to act on insights, not just generate them. Predicting churn is one thing. Embedding that prediction into a personalized, timely workflow that resolves the issue—that's the game-changer. Companies need a strong operating model, a unified data infrastructure, and collaboration across customer-facing and product functions to do that.

The laggards will stay stuck in pilot purgatory, unable to scale due to fragmented data, siloed teams, or a lack of governance. Those who treat preemptive service as a strategic shift, not just a technology project, will emerge far ahead.

Omid Razavi: Nick, we're aligned in our optimism about AI's transformative potential in customer service and support. This is more than a technological evolution—it fundamentally reimagines how organizations engage customers, empower teams, and deliver meaningful value. Thank you for your thoughtful insights and helping advance the dialogue on what’s coming next.

Nick Clark: Always a pleasure, Omid. These discussions are essential as organizations move from experimentation to scaled, strategic impact. I’m looking forward to continuing the conversation and hearing even more real-world perspectives at the CSS Executive Forum in London next month.

As we prepare for the 4th CSS Executive Forum in London, it’s clear that AI is reshaping service and support at every level—but no single organization has all the answers. That’s why the CSS community is so important. This forum brings together senior leaders to exchange lessons, challenge assumptions, and co-create what’s next.

If you’re leading AI-driven transformation in customer service, success, or support, we hope you’ll join us in London and be part of this pivotal conversation.

👉 If you found this valuable, subscribe to the CCO Perspectives newsletter for more conversations like this.

💬 And don’t miss out—join the CSS Communities and CSS Community Europe groups on LinkedIn to connect with fellow leaders shaping the next chapters of customer service and support.

Comments (0)

Popular

Bridging AI Strategy and Execution in Customer Service

Omid Razavi

Popular

Related