How Organizations Are Losing Budget on AI Implementation: Key Insights from Industry Leaders

Contact center leaders invested $2.1 billion in AI tools last year, but a troubling reality has emerged: a significant portion of that investment is actively degrading productivity rather than enhancing it. At the 2026 ICMI Digital Event, three industry veterans—Jim Iyoob, President of ETS Labs and Chief Revenue Officer; Manu Dwievedi, AVP of Product Strategy & Innovation; and Shawndra Tobias, Chief Data Strategy Officer at Etech Global Services—cut through vendor promises to reveal exactly where AI investments go wrong and how to fix them.

The Hidden Costs of Poor AI Implementation

The session opened with a stark reality check about how organizations are hemorrhaging budget through critical operational gaps. Agents are navigating 5-6 screens simultaneously while outdated "agent assist" technologies add more distraction than value. Leaders spend 40% of their time searching for data and insights without the training to interpret it effectively. Meanwhile, generic AI models trained on internet data rather than contact center-specific interactions produce false positives that erode trust among agents and supervisors.

Perhaps most concerning: 90% of customer interactions aren't being analyzed at all in most organizations. "Your customer will tell you everything you want to know, good, bad and ugly, when they're on that conversation with your agent," Iyoob explained. "The problem is if you're not extracting that data to understand what could have prevented it, that's where the risk becomes." Organizations are making critical business decisions based on biased information from analyzing only 1-3% of calls, potentially losing $4-5 million in missed opportunities depending on size.

Domain-Trained AI: The Game Changer

The speakers made a compelling case for specialized, domain-trained AI over generalized models. While generalized AI offers 60-70% classification accuracy with high false positive rates, domain-trained AI delivers 85%+ accuracy by learning from millions of actual contact center interactions. This specialized training enables the AI to understand industry jargon, emotional cues, and context in ways that generic models simply cannot match.

Dwievedi demonstrated their Q Eval platform, showcasing real-time transcription and sentiment analysis, automated note-taking that saves 60 seconds per call, and contextual understanding that identifies call intent and resolution. Most impressively, the platform provided specific, actionable coaching: "When the customer was actually trying to book for their family, you should have actually offered interconnecting side by side rooms, which we have." This level of specificity stands in stark contrast to generic AI responses.

Proven Results and Implementation Strategy

Three case studies demonstrated measurable ROI: a healthcare program achieved 16% increase in conversion rates and 22% increase in critical behaviors; a global drug manufacturer saw 45% decrease in escalations and $542,000 in cost savings; and a hospitality leader achieved 33% revenue growth. These results came from identifying and replicating top performer behaviors rather than creating entirely new processes.

The most valuable insight concerned implementation strategy. Dwievedi revealed that successful AI adoption follows an 80/20 rule: "80% of that implementation is not the AI tool at all, but how you manage change." Organizations that achieve success start with clear problem definitions, follow a crawl-walk-run methodology focusing on three priorities rather than 100, and invest heavily in building trust through data transparency. Following this approach, customers achieve 94% adoption rates and see ROI within 90 days.

The Path Forward

The session's clear message: AI investment without proper evaluation and implementation strategy is worse than no investment at all. Organizations must shift from chasing shiny new technologies to solving specific operational problems with specialized AI that enhances rather than hinders experience. "It's not about replacing people," Dwievedi emphasized. "It's about solving that problem for your customer."

The speakers recommended starting small—pick three things that keep you up at night, work on those until successful, then move to the next priorities. By focusing on domain-trained AI, prioritizing agent experience, and investing in change management, organizations can transform their AI investments from budget drains into powerful drivers of customer satisfaction and business growth.

Session Title: Half Your AI Budget is Going to Tools That Make Your Agents Slower
Speakers: Jim Iyoob, President, ETS Labs | Chief Revenue Officer, Manu Dwievedi, AVP of Product Strategy & Innovation & Shawndra Tobias, Chief Data Strategy Officer, Etech Global Services
Event: ICMI Digital Event 2026

Want to dive deeper into these insights? Watch the full session recording and discover the complete 25-point audit framework, additional case studies, and detailed implementation strategies here.

Interested in more cutting-edge contact center strategies? Explore other recorded sessions from the ICMI Digital Event here.