The productivity gains are real. But for many firms, the results aren’t adding up.
Despite growing excitement—and rapidly rising budgets—few organizations can point to meaningful returns from their GenAI investments. According to Gartner, at least 30% of GenAI projects will be abandoned by the end of 2025 due to poor data quality, escalating costs, or unclear business value. Meanwhile, nearly all executives surveyed by KPMG said they plan to increase spending, with nearly half anticipating investments of over $100 million.
So why isn’t the ROI keeping pace?
GenAI Works—But Only When It’s Used
At the individual level, GenAI is already making a difference. Many professionals report saving 20–40% of their time on tasks like drafting, summarizing, and organizing. These are real, repeatable improvements.
But isolated productivity gains don’t add up to transformation. To see returns at the organizational level, GenAI must become part of how work gets done, firm-wide.
The Real Challenge: Adoption at Scale
If only 10–20% of employees use GenAI tools regularly, it’s unlikely the firm will achieve a measurable return. The math simply doesn’t support it.
At the same time, there's no going back. Once people experience the benefits of GenAI, they tend to keep using it. The genie’s out of the bottle. So while it's possible to slow down or rethink your strategy, the momentum is already underway. The question is no longer whether to move forward, but how.
Where GenAI ROI Begins: Aligning Time With Priorities
The biggest opportunity for GenAI ROI may not lie in helping people work slightly faster, but in reclaiming time currently lost to low-value work.
Insights shared by Ryan Alshok (CEO, Laurel) during a Skills webinar suggest that knowledge workers may spend 60–70% of their time on administrative tasks, context switching, or coordination, rather than strategic priorities. In contrast, top-performing firms like Nvidia reportedly align up to 60% of employee time with operational and strategic goals.
That gap represents an enormous opportunity. Even a modest 1–5% productivity improvement across the firm could justify significant investment, if the technology is thoughtfully integrated.
GenAI Is a Opportunity for Change - Not Just a Set of Tools
A common pitfall is starting with the question: What tool should we roll out next?
That’s understandable - but often unproductive. In practice, those conversations rarely lead to the kind of transformation firms are aiming for.
A more useful starting point might be:
- Where are we losing time?
- What bottlenecks or handoffs slow things down?
- What gets duplicated unnecessarily?
GenAI ROI depends less on which tools are selected and more on how work is redesigned around them.
The firms seeing traction tend to approach GenAI as infrastructure. They invest in:
- Process redesign aligned with GenAI capabilities
- Training programs that focus on application, not just access
- Metrics that track improvements at the team or organizational level
- Change management that supports sustained adoption
- Scalable technical architecture that supports enterprise use
A Phased Approach to GenAI ROI: Start Broad, Then Go Deep
There's no single roadmap, but for many firms, a phased strategy has shown promise. It allows you to build momentum without trying to change everything at once.
Phase 1: Broad Access, Low Risk
Start by offering access to GenAI tools across the organization. Let individuals experiment. Let early adopters emerge. This builds familiarity and begins identifying where the real opportunities lie, without significant risk.
Phase 2: Department-by-Department Redesign
Once awareness and comfort grow, focus on one department at a time. For example:
- Sales might streamline proposal generation
- Legal could optimize contract review
- Marketing might accelerate content workflows
This targeted approach allows you to refine what works and build credible internal case studies before scaling more broadly.
Phase 3: Cross-Functional Integration
Only after department-level maturity should you focus on cross-functional workflows. That’s where the network effects—and real firm-wide GenAI ROI - start to emerge. But by then, you’ve laid the groundwork with tested practices, trusted champions, and measurable results.
The Bottom Line: GenAI ROI Is Possible - But It Takes a Different Mindset
It’s not that GenAI can’t deliver ROI. It’s that most firms are trying to capture it before they’ve created the conditions to support it.
Rather than trying to boil the ocean, firms might consider starting cooking with one pot and then scaling based on what works.
With 65% of IT leaders expecting higher tech spending in 2025 (source), and Generative AI cited as a top driver, the urgency is real. But that investment doesn’t have to be a leap of faith.
Start with access.
Build competence.
Redesign workflows where it matters most.
That’s how GenAI moves from potential to performance.
Not sure why your GenAI efforts aren’t delivering ROI?
Take the AI Execution Scorecard to find out where the gaps are - and how to close them.