The 79/24 Problem: Why Executives Know AI Will Drive Revenue but Can't Say Where

Executive Summary

• IBM IBV and Oxford Economics surveyed 2,007 senior executives across 33 geographies and 23 industries (Q3–Q4 2025). The finding that anchors every CIO and CFO conversation: 79% of executives expect AI to significantly contribute to revenue by 2030 — but only 24% can articulate where that revenue will come from.
• The expectation-to-roadmap gap is not a planning failure. It is a structural one: 68% of executives worry their AI efforts will fail specifically due to lack of integration with core business activities — meaning the ambition exists but the connective tissue does not.
• AI investment is projected to surge 150% (as a percentage of revenue) between 2025 and 2030. Executives expect a 42% productivity gain. But only organizations that redesign workflows around AI — not alongside it — are positioned to capture that gain: multi-workflow AI adopters anticipate 24% greater productivity and 55% higher operating margins than peers.
• The skill gap is sharper than most workforce plans assume. 57% of executives expect most current employee skills to become obsolete by 2030. By end of 2026, executives expect 56% of the workforce will require reskilling due to AI-driven automation. The window is shorter than most HR timelines account for.
• The model clarity gap mirrors the revenue gap: 57% say competitive advantage will come from AI model sophistication, but only 28% have a clear view of which models they’ll need. Buying the most powerful model is not a strategy.

Source credibility: MEDIUM-HIGH. IBM IBV is IBM Consulting’s research arm — the 68% “integration failure” finding aligns with IBM’s commercial thesis. Oxford Economics partnership provides independent methodology. n=2,007 across 33 geographies and 23 industries, Q3–Q4 2025 fieldwork is large and current (Tier 1). Weight the directional findings; apply vendor context when citing absolute percentages.

The Expectation-Roadmap Gap

The 79/24 split is the most useful number in this dataset. Nearly four in five executives believe AI will materially contribute to their revenue in the next five years. Fewer than one in four can describe the pathway.

That asymmetry is not confusion — it’s compression. Executives see the destination (AI-driven revenue growth) but haven’t done the architecture work: which workflows change, which revenue streams scale, which customer behaviors shift. The confidence is real. The plan is not.

The 68% who worry about integration failure are being precise about the cause. AI tools that sit next to existing workflows — augmenting tasks without restructuring how work gets done — generate productivity measurements that look like rounding errors. The organizations capturing meaningful gains are redesigning the workflow first, then deploying AI into the redesigned process.

IBM IBV’s data shows what that redesign looks like at scale: organizations that scale AI across multiple workflows (rather than isolated use cases) anticipate 24% greater productivity gains and 55% higher operating margins than peers by 2030. The difference is not tool selection. It is architecture.

The Productivity Flywheel — and Who Captures It

Executives expect AI to increase productivity by 42% by 2030. That is a large number. The less-cited companion stat is more actionable: 67% expect to have captured most of those gains by 2030 — meaning a third of organizations will invest heavily and still be chasing the return.

The divergence tracks a specific pattern. AI-first organizations — those redesigning operations around AI, not just adding AI to existing operations — anticipate:

• 70% greater productivity improvement vs. peers
• 74% greater reduction in process cycle times
• 67% greater improvement in project delivery times

Those are not marginal differences. They suggest the productivity question is less about whether AI delivers and more about whether the organizational architecture is built to receive it.

The 70% of executives who plan to reinvest AI-generated productivity gains into growth (rather than cost reduction) are describing the flywheel: efficiency gains fund AI capability expansion, which funds further efficiency gains. The organizations that capture 2030 gains are mostly the ones already running that loop.

Today, 47% of AI spend targets efficiency. By 2030, executives expect 62% to target product, service, and business model innovation. That shift — from efficiency spending to growth spending — is the inflection point. Companies still optimizing existing workflows in 2028 will be two cycles behind.

The Workforce Reckoning Is Sooner Than HR Plans Assume

The workforce data in this study is more urgent than the headline AI investment figures.

Executives expect 56% of the workforce to require reskilling by end of 2026 due to AI-driven automation. That is not a 2030 figure — it is a 14-month horizon from the fieldwork date. Most corporate reskilling programs are structured as multi-year initiatives. The timeline assumption embedded in those programs is wrong.

The skill profile shifting is equally important. 57% of executives expect current employee skills to become obsolete by 2030. But 67% say mindset will matter more than skills — meaning the executives themselves have concluded that training for specific technical skills is a losing game. The durable investment is in problem-solving capacity and adaptability, not tool certification.

Two structural signals point the same direction:

• AI-first organizations are 48% more likely to create net-new job roles and 46% more likely to redesign organizational structure — suggesting the workforce change is not replacement but reconfiguration.
• 74% say AI will redefine leadership roles. 68% expect to have a Chief AI Officer by 2030. The management layer is changing faster than most succession plans acknowledge.

The Model Clarity Problem

The AI model landscape will look fundamentally different by 2030 than it does today. 82% of executives expect multi-model AI capabilities — meaning organizations will run combinations of large and small models tuned for specific tasks, not one vendor’s platform across everything. 72% expect small language models to surpass large language models in enterprise prominence.

The procurement implication: buying a single enterprise AI platform today and assuming it remains the right architecture in 2027 is a poor bet. The organizations with durable advantage are those building model-agnostic integration layers — not those optimizing for one vendor’s current capabilities.

The model clarity gap is real: 57% say competitive advantage will come from model sophistication, but only 28% have clarity on which models they’ll need. That 29-point gap represents the next phase of vendor lock-in risk. Signing long-term enterprise agreements before the multi-model architecture stabilizes creates exit costs that will be painful by 2027.

Key Data Points

What This Means for Your Organization

The 79/24 gap is a diagnostic, not a data point. If your organization is in the 79% — if AI is expected to contribute to revenue — but you cannot name the three specific workflow changes that will generate that revenue, you have an ambition without an architecture. The gap between expectation and roadmap is where budgets get spent and results don’t arrive.

The integration finding sharpens the problem. 68% of executives who worry about failure name the same root cause: AI that doesn’t connect to how the business actually runs. That failure mode is specific and preventable. It means AI tools deployed in isolation — without redesigning the workflows they touch — will generate activity metrics and not business outcomes.

The workforce timeline is the most underreacted finding in this dataset. If 56% of your workforce needs reskilling by end of 2026 — 14 months from the survey fieldwork — and your reskilling program is a 2027 initiative, you have a sequencing problem, not a capability problem. The answer is not to accelerate a broad reskilling program; it is to identify which specific roles and workflows are changing first and resource those specifically.

The multi-model architecture shift will create procurement decisions in the next 12–18 months that are difficult to reverse. Long-term single-vendor commitments signed now will look constraining by 2027 when multi-model, task-specific deployment becomes the baseline rather than the advanced approach.

If any of these gaps are live conversations in your organization, the questions worth pressure-testing are specific: which workflows are changing, which skills are actually durable, and which vendor commitments are building flexibility rather than foreclosing it. I’d welcome that conversation — brandon@brandonsneider.com.

Sources

Primary:

• IBM Institute for Business Value / Oxford Economics. “The Enterprise in 2030.” January 2026. n=2,007 senior C-suite executives, 33 geographies, 23 industries, Q3–Q4 2025 fieldwork. https://www.ibm.com/thought-leadership/institute-business-value/en-us/report/enterprise-2030
• IBM Newsroom. “IBM Study: AI Poised to Drive Smarter Business Growth Through 2030.” January 19, 2026. https://newsroom.ibm.com/2026-01-19-ibm-study-ai-poised-to-drive-smarter-business-growth-through-2030

Credibility note: IBM IBV is the research arm of IBM Consulting. IBM has commercial interest in AI consulting and transformation engagements. Cross-reference against independent evidence. The integration failure finding (68%) directly supports IBM Consulting’s commercial positioning. Oxford Economics partnership provides independent survey methodology. Treat as vendor-sponsored research with strong methodology; weight directional findings, apply caveat when citing in client-facing context. Cross-reference against: BCG AI at Work 2025 (5% substantial gains gap), METR RCT (experienced developers 19% slower without workflow redesign), Wharton/GBK Collective 2025 (executive-manager perception gap).

Temporal tier: TIER 1 — Q3–Q4 2025 fieldwork, published January 2026. Cite directly, no caveat needed.

Brandon Sneider | brandon@brandonsneider.com April 2026