Having led technology projects and teams for decades, I’ve heard the conversations around AI and its potential for the future. What could AI do? Where could it create efficiencies? This groundswell of excitement had businesses in nearly every industry quickly pivoting to integrate AI into operations without a clear strategy.
Today, AI has significantly evolved, and so has the conversation. Now, boards and executive teams are asking a much simpler question: What did AI actually deliver? The AI honeymoon phase is over, and the prove-the-ROI phase is heating up.
The Rising Pressure to Prove AI
In a recent independent research study of 500+ global business leaders, 91% of executives said pressure to prove AI ROI has intensified. AI is now a real business investment with real expectations attached to it.
What concerns me is not the pressure itself. It’s how organizations are responding to it.
Across industries, I am seeing companies move too quickly to scale AI before they have the foundation required to support it. Leaders are setting ambitious goals while teams are still working through the realities of execution. In many cases, organizations are mistaking speed for readiness.
One gap in our research stood out immediately. While 56% of executives believe AI is fully integrated across the product life cycle, only 18% of managers agree. That disconnect matters because it reveals how differently leadership and delivery teams often view progress.
Executives believe the organization is scaling AI successfully while teams on the ground are still dealing with fragmented systems, unclear processes, and operational bottlenecks.
And the result is predictable. Expectations are set ahead of execution readiness, and organizations commit to outcomes they are not yet equipped to deliver.
When the Rush to ROI Is a Trap
AI is accelerating development and output across industries. Teams can prototype faster, automate more work, and increase production speed. But speed alone does not create business value. AI is an enabler, not a substitute for the foundational steps of product development.
In our research, 75% of product leaders said they still struggle to execute on strategy. Organizations are producing more while alignment around priorities, ownership, and outcomes struggles to keep pace.
This is where ROI pressure can become dangerous. When organizations feel pressure to prove value quickly, they often prioritize work process efficiencies before systems and teams are ready. And in some cases, companies immediately look to reduce headcount or replace expertise with AI. This sends the wrong message about AI. There will always be a critical need for human oversight and depth of expertise.
AI is most effective when it amplifies expertise, not when it attempts to replace it. Without the people who understand how systems operate, how products are developed, and how governance needs to function in real-world environments, AI initiatives become fragile. Organizations end up with unreliable outputs, security concerns, and systems that struggle once they move beyond controlled pilots.
AI does not replace product and engineering fundamentals. In many cases, it makes gaps more visible.
Without Modernization, AI Can’t Deliver
The real work of the AI era is legacy system modernization.
AI models may perform well in isolation, but production environments are far more complex. Data silos, disconnected systems, and aging infrastructure quickly become blockers once organizations attempt to scale AI across the business. Failure to modernize will result in the continued accrual of legacy tech debt.
Time after time, I see ambitious organizations come to the realization that AI success depends on much more than the model itself. And in our research, 95% of leaders said they are modernizing systems, with AI and analytics as the top driver.
Organizations cannot treat AI as a shortcut around existing processes. AI needs to be built into how products, systems, and workflows are developed and operated. Alignment across teams, clear outcomes, and strong execution will always matter. In fact, they matter even more when AI enters the picture.
What Successful AI Looks Like
When a global life sciences company partnered with Modus Create to modernize clinical trial strategy design, the business problem was clear: analysts were relying on slow, spreadsheet-driven manual processes to build country-specific trial strategies. The solution combined predictive enrollment models with a constraint optimization engine. This gave teams a way to quickly generate strategies while still allowing analysts to apply their own judgment.
That discipline paid off. The platform delivered more than $5 million in clinical trial cost savings and helped compress strategy design from weeks to minutes.
That story reflects how successful AI adoption actually happens. Not through shortcuts or rushed deployment, but through clear ROI goals, disciplined execution, modernization, and steady refinement.
The Foundation Comes First
Organizations that will succeed with AI are the ones investing in the right foundation upfront. They align leadership expectations with execution realities. They understand the why behind modernizing legacy systems—so that AI can operate in real-world conditions. And they build teams that combine technical capability with operational expertise.
AI can absolutely deliver measurable business value. But long-term success depends on whether organizations are willing to invest in the systems, processes, and people required to support it at scale. The ones that get this right won't be the ones that moved the fastest. They'll be the ones willing to be brutally honest about where they actually stand, and build from there.
