UBCNews - Business - How AI Agents Boost Business ROI: Making The Shift To An Agentic Workforce

Episode Date: January 5, 2026

So here's something wild: only ten percent of organizations right now are seeing significant ROI from agentic AI. But almost everyone expects returns within one to five years. That gap tells us something important about where we are in this transformation, doesn't it? Spotlight on Startups City: Laguna Niguel Address: 110 Chandon Website: https://spotlightonstartups.com

So here's something wild. Only 10% of organizations right now are seeing significant ROI from agentic AI. But almost everyone expects returns within one to five years. That gap tells us something important about where we are in this transformation, doesn't it? Absolutely. And I think it's because we're in the middle of a fundamental shift. This isn't just about automation anymore. AI agents are autonomous systems that can think, adapt, solve multi-step problems. They're making decisions, learning from outcomes, and operating across entire workflows without constant human oversight. Right. So when we talk about agentic AI versus traditional automation, or even generative

AI, we're talking about a completely different beast. Can you break that down for us? Sure. Traditional automation follows predefined scripts. You set rules, it executes them. Generative AI creates content based on patterns it's learned, but agentic AI, it orchestrates end-to-end workflows, makes dynamic decisions in real-time, and performs independently across systems like HRIS, CRM, financial platforms. It only escalates to humans when truly necessary. That's powerful, and the numbers back it up. 79% of organizations report at least some AI agent implementation, and by 2020, Eight, a third of enterprise software is expected to incorporate Agentic AI, automating 15% of work decisions. So, why are traditional ROI models falling short here?

Because they focus too narrowly on cost savings and time save per transaction. Agenic AI delivers so much more. Agility, scalability, improved employee experience, strategic impact. If you only measure direct efficiency gains, you're missing the bigger picture. Mm-hmm, I hear you. You need a multi-dimensional strategy. Look at operational efficiency, sure, but also employee productivity, strategic impact, cross-department scalability, and adoption experience. For example, lift cut customer service resolution times by 87% with a Claude-based system.

In customer service broadly, some strategic implementations have achieved ROI reaching 800%. Those are impressive results, and I'm curious how does this play out in other areas like IT operations or financial services? In IT operations, AI can process alerts, logs, and metrics data at scale, significantly reducing ticket volume, a global telecommunications provider using IBM Watson assistant reduced call volume and average handle time while improving employee morale. In finance, Agentic AI monitors transactions for fraud, automates loan processing, and streamlines reconciliation. J.P. Morgan Chase achieved a 450% increase in ad click-through rates by using AI to optimize digital advertising copy.

That point about multidimensional strategy sets up our next piece, how leaders actually implement this across functions. But first, a quick word from our sponsor. Spotlight on Startups is a dedicated platform celebrating bold entrepreneurs and early stage companies. Their mission is to shine a light. on the entrepreneurial path, the wins, the challenges, the lessons, and the human side of building something new. They produce audio, video, and articles to reach audiences across platforms,

creating an evergreen library of startup stories that inspire and inform. Go to spotlight on startups.com to learn more. Picking up on multidimensional strategy, how do you actually measure ROI when Agentic AI touches multiple departments? Great question. You want to track a self-service rates, AI handled requests divided by total request times 100. Also, look at velocity of adoption across departments and multifunctional workflow completion. The wider the coverage, the more compounding value you get in time savings, data consistency, and employee satisfaction. And speaking of employee satisfaction, that's often overlooked in ROI calculations, isn't it? Definitely, fewer handoffs, faster resolutions, smarter automation, all of this increases

job satisfaction. You can measure it through CSAT scores, net promoter score, and reduction in follow-up tickets. These aren't just sentiment metrics. They correlate closely with retention and engagement. I actually remember working with a client who saw employee morale improve dramatically after implementing AI agents in their support function. The team finally had time to focus on the strategic work instead of repetitive tasks. That human element is vital. Now let's talk about the elephant in the room. Investment timelines.

Most organizations report taking two to four years to see satisfactory ROI on typical AI use cases. That's significantly longer than what's often expected for many tech investments. Why the disconnect? Because AI rarely delivers value in isolation, it's introduced alongside efforts. to improve data quality, reconfigure teams, streamline operations. Plus, many organizations overestimate their data maturity and invest in AI before addressing core infrastructure gaps. That delays results.

Right. So what separates the leaders from everyone else? The top 20% seeing real returns. What are they doing differently? AI ROI leaders treat AI as an enterprise transformation, not a tech upgrade. They prioritize revenue-focused ROI, make strategic bets on both generative and agentic AI, and invest substantially in their AI programs. Most importantly, 65% of organizations now say AI is part of corporate strategy, signaling a fundamental shift in how businesses view these

investments. In other words, they're making AI central to how they operate, not just a side experiment. That's leadership commitment at the highest level, and I imagine they're measuring things differently too? Exactly. Leading organizations understand that generative AI and agentic AI require different measurement approaches. Generative AI can deliver short-term impact and build momentum. Agentic AI requires longer-term thinking because of its complexity and transformative potential. They understand that not all returns are immediate or purely financial. So to everyone listening, if you're a startup founder or business leader thinking about this transition, what's the practical first step? Have you thought about where your organization might start?

Start with high-impact use cases. Focus on areas like IT support, onboarding, or procurement requests where you can demonstrate quick wins, then continuously measure and optimize using analytics dashboards, track automation rates, resolution accuracy, customer satisfaction. And here's the thing. Drive adoption through change management. Train employees, create internal champions, integrate AI smoothly into daily operations. Right, because even the most powerful AI system doesn't deliver value if people don't engage with it. What about avoiding common pitfalls? Don't calculate ROI before pilot validation, don't overlook strategic and qualitative value, and don't fail to optimize post-launch. Many organizations also face challenges.

like integration with legacy systems, data quality issues, security concerns, and cultural resistance. You know, I joke that sometimes the hardest part of implementing AI isn't the technology. It's convincing Bob from accounting that the system won't replace him. That's true for any major change, isn't it? Have you thought about how this shift compares to historical technology transitions? Some say it's like the move from steam to electricity. That's a good analogy. When factories switched from steam power, they had to reconfigure production lines, redesign workflows, invest in new infrastructure, reskill their workforce.

The full benefits only emerge once they fundamentally changed how they operated. The same is true for AI. 78% of C-suite executives believe realizing maximum benefit fromogenic AI requires a new operating model. Mm-hmm. Interesting. By 2030, IDC forecasts that 45% of organizations will orchestrate AI agents at scale across various business functions. Gartner predicts AI agents in customer service will manage over 80% of standard support tasks by 2029. We're going to see agentic AI managing entire workflows, identifying inefficiencies, adapting to changing business needs, all while continuously learning from past experiences. That's a future where AI augments human capability rather than just replacing tasks.

Leading organizations expect agentic AI will enable employees to spend more time on strategic and creative work. How do you think that changes organizational culture? It requires a human-centered approach. Position AI as a tool that strengthens rather than threatens. Research shows that companies investing in generative AI are seeing significant returns. an average ROI of $3.7 for every dollar invested, with top performers achieving $10 for every dollar invested. But the real transformation is cultural.

Organizations are recognizing that AI fluency needs to become a core competency across their workforce. So the organizations that thrive will be those that view AI as both a strategic capability and a source of agility. Final thought, what's the one thing you want founders and leaders to remember as they work through this transition. Remember that agenic AI delivers value across five dimensions. Operational efficiency, employee productivity, strategic impact, cross-department scalability,

and adoption experience. Measure all of them and understand that while investment is rising, with the vast majority of organizations planning to increase their AI spending, patience and strategic thinking are key. This transformation takes time, but the compounding benefits are worth it. Couldn't have said it better. Thanks so much for breaking this down today. My pleasure. This is an exciting time to be building and scaling businesses with AI.