The promise of artificial intelligence in customer experience has never been clearer—or more urgent. Yet many organizations find themselves paralyzed between the fear of falling behind and uncertainty about how to proceed. According to Daniel Ziv, Global VP of AI and Analytics at Verint, this hesitation itself poses the greatest risk.
“The pace of change impacts the need for much, much faster insights,” Ziv explains. “We need to analytics that could provide either real time or near real time insights.”
Two Critical Frameworks for Understanding AI Analytics
Macro vs. Micro Analytics in Customer Experience
Ziv introduces a useful distinction between two complementary types of analysis that organizations need:
Macro analytics examines patterns across all customer interactions to identify emerging trends and issues at scale. As Ziv notes, “You need a system that can look scalably at the entire volume of stuff and give you a statistically valid sample across that trend.” For instance, one Verint customer discovered they were receiving three times as many calls about tariffs compared to recent periods, with these calls lasting twice as long.
Micro analytics applies these insights to individual interactions in real-time. “I need to infuse that insight and analyze the call as it’s happening, provide real time guidance or apply this information to my website or to my IVA or IVR,” Ziv explains.
The gap between these two creates competitive advantage: “Companies that have both have a real advantage,” he observes. Organizations that excel at macro analysis often fail to operationalize insights quickly enough, while those focused on micro analytics may be optimizing “on last year’s insights.”
Generative vs. Agentic AI in Customer Experience
Understanding the evolution from generative to agentic AI helps organizations plan their investments more strategically.
Generative AI responds to prompts with content, understanding, and analysis. It excels at analyzing unstructured data like voice conversations and text interactions. As Ziv describes it: “AI can generate content, can understand unstructured conversations, including voice and text and video and images and respond to it.”
Agentic AI takes this further by breaking down complex goals into multiple steps, executing them autonomously. “I can set a goal and say, hey, I want to increase revenue for my business. Go figure this out. And it’ll itself generate questions. Go to the Web. Go to my database. Go to different places. Come up with answers,” Ziv explains.
Verint’s evolution of its Genie Bot illustrates this progression. The initial generative version analyzed calls about complaints or churn on request. The agentic version now breaks down questions independently, scales analysis across larger samples, quantifies results, and exports findings to presentation formats—executing entire workflows with minimal human intervention.
The Flawed Decision-Making Patterns Holding Organizations Back
Ziv identifies two common failure modes that prevent organizations from capitalizing on AI:
Pattern 1: Gut-Based Decisions Without Data
“A lot of executives tend to be in this position because they’ve been in a space for so long, they feel their gut feeling is usually accurate,” Ziv observes. While experience matters, rapid change undermines intuition’s reliability. Leaders are making decisions without data “because I don’t have the data, and I have to make a decision.”
Pattern 2: Analysis Paralysis
The opposite problem occurs when leaders recognize they need data but wait too long to get it. “I’m making the decision too late. By the time I make the decision, I’ve missed the opportunity. I’ve lost money and I’m behind compared to the competition,” says Ziv.
Both patterns stem from the same root cause: analytics systems that can’t deliver insights fast enough to support decision-making at the speed of business change.
Quantifiable Outcomes: What’s Actually Possible
When organizations implement both macro and micro analytics effectively, the financial impact is measurable and significant. Verint provided the following examples:
- A large international telecom saved $80 million and increased sales by 11% using macro insights combined with real-time coaching capabilities
- Organizations typically find $5-10 million in insights within one to two days of deploying generative AI analysis
- A UK financial services company generated $5 million in additional revenue by using AI to identify and resolve process inefficiencies in loan applications
- An energy supplier saved $2 million through increased agent capacity after using AI to identify processes suitable for self-service
“When we deploy our latest (Verint) Geniebot, we see that within one or two days, we typically find five to ten million dollars of insights just off the bat.” Ziv reports.
Practical Guidance for Getting Started
Start Small, Move Fast
Rather than comprehensive transformation initiatives, Ziv advocates for incremental deployment: “What can you turn on without changing anything that will add value to your environment? Similar to your phone. I can download an app and turn on capability that now is transformative.”
One Verint customer piloted automated call wrap-up notes with 300 agents, validated the impact, then scaled to 30,000 agents. This approach reduces risk while accelerating time-to-value.
The Hybrid Architecture Advantage
Organizations with on-premises infrastructure face a false choice between wholesale cloud migration and missing out on AI innovation. “You can have telephony on-premises and we send recordings in a secure way to cloud services with your existing deployment. You get immediate outcomes with very little cost, very little effort, very little risk,” Ziv explains.
This hybrid approach allows organizations to access cloud-based AI compute power and models while maintaining existing infrastructure, avoiding the complexity and risk of full migration projects.
Focus on Outcomes, Not Infrastructure
“Ninety percent of the time and the effort is focused on moving infrastructure but only 10 percent is about outcomes,” Ziv observes about traditional cloud migration approaches. By inverting this ratio—focusing first on turning on AI capabilities that deliver business value—organizations can fund subsequent innovations through realized savings.
The Strategic Imperative
Ziv frames AI adoption as fundamentally strategic: “If you think about the biggest companies in the world, they are data companies that have better analytics, whether it’s Amazon or Google.”
The competitive advantage lies not in which models to use, but in whether organizations can operationalize their data: “Your competitive advantage is the data that you have that your competitors don’t have. More importantly, how well are you using the data that you have access to?”
His assessment of the current moment is unambiguous: “These are make or break years and down to minutes. Every minute where you’re not taking advantage, learning and embedding AI, it’s either a great opportunity or a huge risk.”
Conclusion
The path forward requires organizations to abandon both gut-based decision-making and analysis paralysis in favor of rapid, iterative deployment of AI capabilities. By implementing both macro and micro analytics, organizations can identify opportunities at scale while operationalizing insights in real-time customer interactions.
The technology has matured to the point where organizations can achieve meaningful business outcomes—measured in millions of dollars—within days of deployment. The question is no longer whether AI analytics can deliver value, but whether organizations can move quickly enough to capitalize on the opportunity before their competitors do.
As Ziv puts it: “The biggest risk is inaction.”
You can see the whole interview on Youtube here or subscribe to the podcast at here.
