Is ROI on AI Investment the CEO's Biggest Challenge?
Technology is rewriting strategy, yet returns are uneven. PwC's latest Global CEO Survey, presented at the World Economic Forum in Davos, shows revenue confidence at a five-year low-while AI adoption accelerates.
Mohamed Kande, PwC Global Chairman, calls 2026 a decisive year for AI. His point is blunt: a small cohort is already translating AI into measurable financial outcomes, while many are stuck in pilots. That gap is widening.
The ROI gap: a widening AI divide
PwC surveyed more than 4,000 chief executives across 95 countries and territories. Only 12% say AI delivered both cost savings and revenue gains in the past year. Over half (56%) saw neither, while 30% report revenue up and 26% report costs down.
The leaders are doing something different. They've laid strong AI foundations and deployed at scale-across products, services, demand generation, and decision-making. PwC notes they are two to three times more likely to have embedded AI extensively across the business.
The payoff shows up in the P&L. Companies applying AI extensively to products, services, and customer experiences achieved nearly four percentage points higher profit margins than those that did not. For 42% of CEOs, the top concern is whether they're transforming fast enough to keep pace with AI.
As Kande puts it: "2026 is shaping up as a decisive year for AI. A small group of companies are already turning AI into measurable financial returns, whilst many others are still struggling to move beyond pilots. That gap is starting to show up in confidence and competitiveness, and it will widen quickly for those that don't act."
What the leaders do differently
- Treat AI as an enterprise capability: Move beyond pilots into production across customer journeys, operations, and decision support-not just experiments.
- Make data the product: High-quality, well-governed data underpins secure, responsible, scalable AI. Without clean data, ROI stalls.
- Tie AI to the P&L: Prioritise use cases with direct margin impact-pricing, churn prevention, sales assist, claims automation, fraud, procurement.
- Engineer for adoption: Bake AI into workflows where people already work. Measure usage, quality, and business impact-not model accuracy alone.
- Ship small, ship often: Time-boxed releases with clear exit criteria. Kill or scale based on evidence.
- Build in security and risk: Cyber, compliance, and model risk management move in lockstep with delivery.
- Upskill line leaders: Equip managers to select use cases, interpret outputs, and coach teams through new ways of working.
Sachin Agrawal of Zoho UK makes the core point: strong digital health and end-to-end transformation drive resilience, and AI's success depends on disciplined data management. High-quality, well-governed data ensures advanced technologies operate securely, responsibly, and at scale.
Reinvention beats hesitation
CEOs are adjusting course. Forty-two percent say their companies started competing in new sectors over the past five years. Among those planning large acquisitions in the next three years, 44% expect to pursue deals outside their current industry.
PwC's data shows that companies with higher revenue shares from new sectors report larger margins-and their CEOs are more confident about growth. The signal is clear: business model reinvention is tied to performance, not just story.
Innovation discipline, not slogans
Half of CEOs say innovation is central to strategy. Execution tells a different story. Only one in four agree their organisation tolerates high-risk innovation, has disciplined processes to stop underperforming initiatives, or runs a defined innovation centre.
Fewer than one in 10 (8%) say they've implemented at least five of six innovation-friendly practices at scale. PwC's numbers link those practices to higher sales from new products, faster revenue growth, and higher margins. Innovation is a system, not a pitch.
90-day plan to turn AI into ROI
- Pick three high-ROI use cases: Prioritise pricing uplift, cross-sell/retention, and a cost-out workflow (e.g., claims, AP, customer service). Define owners and hard targets.
- Launch data quality sprints: Identify critical tables for each use case. Assign data owners. Fix freshness, accuracy, lineage, and access in weeks-not quarters.
- Ship to production: Move from demos to deployed MVPs inside existing tools (CRM, ERP, service desk). Set adoption milestones per role.
- Instrument impact: Baseline costs and conversion before launch. Track lift with A/B or phased rollouts. Report weekly on usage, quality, and P&L impact.
- Create a kill/scale cadence: A monthly forum that cuts underperformers and doubles down on winners. No zombie projects.
- Build guardrails: Add model monitoring, access controls, and incident response. Involve security and legal early.
- Fund talent and training: Cross-functional pods (product, data, engineering, ops). Equip managers and frontline teams to work with AI.
- Communicate the why: Share the business case, new workflows, and expected outcomes. Celebrate wins publicly.
Kenny MacAulay, CEO of Acting Office, is direct: AI and tech spend has become a make-or-break decision for growth. It's about revenue, costs, and the systems that improve customer communications, compliance, and invoicing. Those who commit to digital change will pull ahead; those who ignore the signals may struggle.
What it means next
The pattern in PwC's data is consistent: companies moving fastest to reinvent their operating models-and deploy AI extensively-are outperforming slower movers. The AI divide is less about access to models and more about execution, data, and adoption.
As Kande says: "In periods of rapid change, the instinct to slow down is understandable, but it's also risky. The value at stake across the global economy is increasing, and the window to capture it is narrowing. The companies that succeed will be those willing to make bold decisions and invest with conviction in the capabilities that matter most."
Further reading: PwC Global CEO Survey and the World Economic Forum Annual Meeting.
If your leadership team is building AI capability this quarter, explore targeted upskilling paths at Complete AI Training (Courses by Job).
Your membership also unlocks:
