AI at Scale: Moving Beyond Pilots to Realize ROI 

AI’s promise to transform business is no longer in question – but how to unlock tangible value at scale remains a pressing challenge for executives. Many companies have run dozens of small AI pilot projects, only to see minimal bottom-line impact. Leading strategists now argue that the key is to stop scattering efforts across endless experiments and instead double-down on a focused AI initiative that can truly move the needle. 

From Experiments to Impact – A Strategic Shift 

Despite billions in investment, an estimated 95% of corporate generative AI programs have so far failed to deliver significant financial returns. Too often, firms fall into an “experimentation trap” – deploying AI in isolated pilot use cases that never scale up to enterprise-level impact. The result is a flurry of demos and proofs-of-concept with only marginal efficiency gains, but no meaningful competitive advantage. To break out of this trap, boards and C-suites are rethinking their approach to AI. Instead of running a multitude of shallow pilots, the new mandate is to focus AI on one domain and go deep in that area to drive transformational impact. 

Harvard Business Review’s November–December 2025 feature, “Stop Running So Many AI Pilots,” illustrates this strategy through the lens of consumer goods giant Reckitt. Initially, Reckitt’s leadership saw dozens of potential gen AI applications – from automating slide decks to customer support bots to procurement optimizations. Each promised some time savings and quick ROI, but spread across disparate functions. Executives soon realized that this laundry list of pilots would yield only incremental benefits, not the dramatic strategic boost they were seeking. In response, they chose a single high-impact arena – marketing – and concentrated their AI efforts there, leveraging the company’s rich customer data and tech-savvy talent base in that function.

Deep Focus Yields Real ROI 

By redeploying AI at scale in one core business area, Reckitt achieved striking performance gains that validate the focused approach. After partnering on a bespoke generative AI “marketing insights engine,” the firm reports three notable improvements: 

• 60% faster product concept development – accelerating the creation of new campaign ideas and content.

• ~30% reduction in time to adapt and localize ads across markets – with more consistent, highquality output. 

• Up to 90% less time spent on routine analytics tasks (e.g. post-campaign analysis) – effectively eliminating grunt work while doubling the quality of insights generated. 

These results were not achieved by off-the-shelf tools alone. Reckitt invested in tailored AI solutions – from custom-trained GPT models on its proprietary data to multi-modal content generators – all integrated into new marketing workflows. The lesson is that reaching scale requires more than technology; it demands organizational change. “GenAI offers huge opportunities for growth, but only if harnessed the right way – by rethinking how the entire marketing team operates,” notes Reckitt’s former CMO Fabrice Beaulieu. “Successful adoption comes down to effective change management and intentional leadership”. In other words, executive sponsorship, talent upskilling, and process redesign were as vital as the algorithms in converting AI pilots into real performance gains. 

Enterprise Value at Scale 

Reckitt’s experience reflects a broader shift among leading enterprises in 2025. Many firms are concluding that merely deploying AI tools here and there isn’t enough – the real value emerges when they reengineer business processes end-to-end to be AI-powered. In fact, recent industry research finds fully half of companies (especially in tech and finance) are now moving beyond isolated use-cases to redesign workflows at scale with AI. These “AI at scale” pioneers report their employees saving significantly more time and refocusing on higher-value tasks, sharpening decision-making, and driving measurable improvements in productivity and quality. Crucially, they also rigorously track the value created by AI – aligning projects to clear business KPIs and ROI metrics – rather than chasing hype. For boards and strategists, the takeaway is clear. Sporadic pilot projects will not deliver transformative ROI. To create real shareholder value, companies must prioritize AI investments that scale, concentrate on domains where AI can bolster strategic advantages, and commit to the organizational shifts needed to support them. Done right, a single well-chosen AI initiative – implemented deeply and supported from the C-suite down – can unlock outsized gains, turning AI from a series of science experiments into a robust driver of enterprise value.

Sources & Editorial Note:

This article represents Boardroom Strategy’s original analysis, informed by publicly available research, case studies, and executive commentary on enterprise AI adoption.

Referenced Sources:

Harvard Business Review — Stop Running So Many AI Pilots (Nov–Dec 2025)

BCG — AI at Work 2025

Copyright:

© Boardroom Strategy 2026. All rights reserved.