There’s no question that generative AI and agentic systems are reshaping business. Leaders everywhere are eager to deploy these technologies, expecting a surge in productivity, cost savings, and competitive edge. But here’s the problem: just installing AI tools doesn’t automatically create value. In fact, many companies that rushed to adopt AI are struggling to see meaningful returns on their investments. [mckinsey.com]
The Myth of “Deploy and Profit”
One of the most striking insights from McKinsey’s recent analysis is this: too many leaders have taken the approach of, “Let’s take this technology, deploy it, and watch good things happen.” But the truth is that while technology is important, it’s the business transformation around that technology that determines value — not the tech itself. [mckinsey.com]
Eric Kutcher, McKinsey’s North America Chair, puts it plainly: this era of AI is 80 % business transformation and only 20 % technology transformation. In other words, the technology is the easy part — changing how the organization thinks, works, and creates value is the hard part. [mckinsey.com]
This aligns with broader industry data suggesting that while a majority of companies claim to use AI, only a fraction actually realize measurable business impact. Rapid experimentation and pilot projects abound, but few scale to deliver profits or strategic advantage. [itpro.com]
The ROI Gap Is a Strategy Problem
AI doesn’t deliver value in a vacuum. It delivers value when it’s embedded thoughtfully into business processes that matter. Simply automating an inefficient or outdated process usually just amplifies inefficiency. For AI to drive results:
- You need a clear vision for how AI connects to your business outcomes. Leaders who are seeing real progress articulate what success looks like in five years and work backward to design the change.
- Workflows must be redesigned, not just automated. That means rethinking how work gets done and where AI fits into customer value creation, not just handing a tool to IT and hoping it spreads.
- Leadership must own the transformation. CEOs who embrace AI personally, learn about its capabilities, and align it to business goals are far more likely to bridge the ROI gap.
What planning for change looks like
To move from experiments to measurable results, leaders should plan for change in these four dimensions:
1. Strategy tied to outcomes.
Define where AI will change revenue, cost, risk, or capital intensity. Set a three to five year ambition, then work backward to today’s capabilities and constraints. McKinsey’s work shows that organizations making this shift from tools-first to outcomes-first are more likely to report impact. [mckinsey.com]
2. Process and workflow redesign.
Do not use AI to speed up a process that already works poorly. All that does is make the bad parts happen faster. First, take the time to look at how the entire process actually works from start to finish. Remove steps that do not add any value. Make it easier for work to move from one person or team to the next without confusion. Once the process is clean and logical, then bring in AI to make it faster or smarter.
Research from McKinsey shows that companies get much better results when they fix the process before adding AI. ITPro also found that many AI projects fail to scale because companies try to drop AI into old, messy processes instead of rebuilding those processes first. [mckinsey.com] [itpro.com]
3. Skills development and fluency.
Real adoption requires more than giving employees access to new tools. Teams need training, hands-on experience, and confidence in how to use AI responsibly and effectively within their roles. Building fluency helps AI become part of everyday work instead of an isolated experiment.
4. Leadership engagement.
AI transformation cannot be delegated entirely to IT. Leaders must stay involved, set priorities, and communicate a strong vision for how AI supports the company’s future. When leadership provides alignment and accountability, AI efforts are far more likely to scale and deliver ROI.
Two quick illustrations
- Customer support: what not to do
A company adds a chatbot to its old customer service process. The bot answers basic questions faster, which seems helpful at first. But the rest of the system is still messy. Information is scattered, teams do not have clear ownership, and problems still bounce around. So even though customers get quicker replies, their issues are not solved any faster. In the end, there is no real improvement in satisfaction or cost savings.
- Forecasting: what to do instead
A finance team steps back and fixes how forecasting should work before adding AI. They clean up the data, define who decides what, and remove confusion in the process. Once the workflow makes sense, they bring in an AI tool to help with forecasting. Suddenly the team can work faster, predictions become more accurate, and analysts get more time for higher value planning. The financial impact shows up within a couple of quarters.
Bottom Line
AI is not a magic bullet. It is a powerful enabler when it is part of a broader plan to change how value is created and how work flows. The organizations that will break through the ROI barrier are not the ones with the most pilots. They are the ones that tie AI to outcomes, redesign the work, build fluency, and lead from the top.
Ready to bridge the AI ROI gap in your organization? Start by defining a clear strategy, redesigning processes, and aligning leadership around outcomes. For guidance on building a plan that works, get in touch.










