Why AI Transformation Success Depends on Tech Savvy Business Leaders Not Just Technology

The Critical Leadership Gap in AI Transformation

The conversation around artificial intelligence transformation has reached a critical inflection point. While organizations pour billions into AI infrastructure, tools, and talent, a sobering pattern emerges across industries. According to multiple studies, between 70 and 87 percent of AI projects fail to move beyond pilot stage or deliver meaningful business value. The culprit, according to recent McKinsey research by Dana Maor, Eric Lamarre, and Kate Smaje, is not inadequate technology but a fundamental leadership gap.

Their December 2025 article argues that successful AI transformation hinges on developing what they call domain owners with a second muscle. These are N-2 and N-3 executives who combine traditional business acumen with sufficient technical depth to oversee AI-enabled transformation. While this thesis offers important insights into a real organizational challenge, it also raises critical questions about feasibility, alternative approaches, and whether we are asking too much of individual leaders.

The Domain Owner Imperative

The core argument is compelling in its simplicity. Organizations have no shortage of AI tools, but they desperately lack leaders who can apply these tools to real business problems at scale. The authors define domain owners as executives responsible for end-to-end business processes or value streams, large enough to deliver meaningful impact but manageable without excessive cross-organizational dependencies. Each domain typically requires five to 15 interrelated use cases to capture transformational value.

The case for prioritizing these leaders is strong. Research consistently shows that most companies derive the majority of their AI benefits from a few deeply transformed domains rather than scattered proof-of-concept projects. A 2024 MIT Sloan Management Review study found that organizations achieving significant value from AI focused resources on transforming complete business processes rather than implementing isolated use cases. This validates the domain-focused approach.

The authors cite revealing LinkedIn analysis showing that only 17 percent of Fortune 500 senior leaders possess technical skill sets, with merely 5 percent having held technical roles during their careers. This capability gap represents a structural barrier to AI adoption. When business leaders cannot evaluate technical proposals, calibrate project progress, or reimagine processes through a technology lens, organizations default to incremental improvements rather than transformational change.

Consider the Adam Boyd example from Citizens Bank. Boyd transformed home equity lending from a 35-day process to just a few days by working cross-functionally to develop use cases encompassing pre-underwriting automation, personalized digital marketing, streamlined applications, and process automation. This required Boyd to learn agile software development, understand the bank's data architecture and technology stack, lead cross-functional development teams, and manage complex change across multiple stakeholder groups. The result was breakthrough customer experience and lower costs.

Such examples illuminate why domain ownership matters. Boyd could not have delegated this transformation to IT or a Chief Digital Officer. The integration of business strategy, customer needs, operational realities, and technical possibilities required someone who understood the domain intimately while developing sufficient technical literacy to make informed decisions.

The Second Muscle Framework

The article outlines four core competencies for AI-capable domain owners. First, they must reimagine their domains with customers at the center, understanding pain points and using creativity to envision AI-enabled possibilities rather than simply automating existing workflows. Second, they develop comprehensive AI-enabled transformation roadmaps with sequenced use cases tied to clear KPIs. Third, they oversee technical delivery with enough depth to prioritize work, solve problems, and challenge team thinking. Fourth, they own end-to-end change management rather than delegating implementation.

Building this second muscle requires specific practices. The authors emphasize that domain owners must become fearless learners, dedicating six to ten hours weekly to technology-related learning through reading, vendor meetings, sprint reviews, conferences, courses, and peer exchanges. They must fixate on value rather than technology, pursuing business problems worth solving with a rule of thumb targeting 20 percent incremental value. They need systems-level thinking to understand how technology integrates with organizational elements. They must get hands dirty working alongside teams rather than delegating. They should develop a nose for tech talent, understanding what differentiates top engineers and data scientists. Finally, they must embrace ambiguity and stage-gate efforts.

Research from Harvard Business School professor Linda Hill supports this continuous learning imperative. Her studies of innovation leaders found that those who successfully drove digital transformation spent significant time experimenting with new technologies, engaging with technical teams, and developing practical rather than merely conceptual understanding. A 2023 Gartner survey found that business leaders who invested at least five hours weekly in technology learning were three times more likely to report successful digital initiatives.

Critical Questions and Counterarguments

While the domain owner model addresses a real gap, several concerns merit examination. First is the feasibility of developing these hybrid business-technology leaders at scale. The authors estimate organizations need 75 to 150 such leaders among N-2 and N-3 populations to transform 15 to 30 core business processes. Developing this bench while simultaneously executing transformation represents a massive organizational lift.

The time requirements alone are daunting. Six to ten hours weekly for learning, deep engagement with cross-functional teams, hands-on involvement in sprint reviews, talent development, and change management could easily consume more than full-time attention. When do these leaders handle their existing responsibilities? Research from Deloitte on leadership capacity suggests that executives already face significant time constraints, with many reporting feeling overwhelmed by current demands before adding technology upskilling.

A second concern involves whether business leaders are the right owners for technical delivery. The article argues domain owners should oversee development, help teams prioritize, and solve roadblocks. However, research from the Standish Group on project success factors consistently identifies experienced technical leadership as critical for delivery effectiveness. While business context matters enormously, there may be tension between having business leaders oversee technical work versus having technical leaders deeply understand business context.

Some successful AI-native companies take different approaches. Netflix, for example, has product managers with strong technical backgrounds who partner closely with business stakeholders rather than expecting business leaders to become sufficiently technical. Amazon's two-pizza teams combine technical and business expertise in small, autonomous units rather than requiring individual leaders to possess both skill sets. These alternative models suggest multiple paths to bridging the business-technology divide.

Third, the article may underestimate organizational and cultural barriers. The authors recommend changing 20 to 30 percent of current leaders, implementing persistent funding models, embedding engineering talent in business teams, and restructuring incentives. These changes threaten existing power structures and career paths. Research by McKinsey itself on organizational transformations shows that cultural and political resistance often derails restructuring efforts. The domain owner model requires not just individual capability building but fundamental organizational redesign.

The Talent Development Challenge

The C-suite recommendations focus on three areas: placing the right people with appropriate incentives, launching strategic upskilling programs, and fixing operating models. Each presents distinct challenges.

Identifying and developing the right leaders requires assessment frameworks that may not exist in most organizations. How do you evaluate whether a business leader has the aptitude and interest to develop technical capabilities? The article profiles successful domain owners like Neesha Hathi, who built technical muscles through roles running a software subsidiary and serving as Chief Digital Officer. Such career paths are uncommon for traditional business leaders. Organizations may lack sufficient candidates with both domain expertise and technical potential.

The upskilling approach emphasizes hands-on practice over coursework, recommending working with consulting partners, two-in-a-box leadership models pairing technical and business leaders, and capstone projects. Research from MIT's Center for Information Systems Research supports experiential learning for technology capabilities. A 2024 study found that executives who participated in hands-on AI projects developed more accurate mental models of AI capabilities and limitations compared to those who completed only classroom training.

However, the two-in-a-box model deserves scrutiny. While pairing business and technical leaders can bridge capability gaps, it also introduces coordination costs, potential conflicts over decision rights, and questions about ultimate accountability. Research on shared leadership models shows mixed results, with success depending heavily on interpersonal dynamics and clear role definition. Organizations should carefully structure these partnerships rather than assuming they will organically succeed.

The operating model fixes—embedding engineering talent, persistent funding, aligned incentives—represent the most challenging recommendations. These changes affect budgeting processes, career structures, reporting relationships, and performance management across organizations. Bain & Company research on agile transformations found that operating model changes took an average of three years to fully implement and often required multiple iterations to achieve desired results. Organizations should not underestimate the time and effort required.

Alternative Perspectives and Complementary Approaches

While the domain owner model has merit, examining alternative and complementary approaches provides a more complete picture. Some organizations successfully pursue AI transformation through centralized AI centers of excellence that partner with business units. This model, employed by companies like Mastercard and DBS Bank, concentrates technical expertise centrally while ensuring...

Implementation Realities

For organizations committed to developing AI-capable domain owners, several practical considerations emerge. First, segmentation matters. Not all business processes require the same level of technical transformation. Organizations should prioritize developing deep technical capabilities among leaders of domains with highest AI potential rather than attempting uniform upskilling across all executives.

Second, career path design requires attention. If the article is correct that technical capability is now table stakes for business leadership, organizations must create development paths that build both business and technical muscles. This might involve rotations through technical roles, formal computer science or data science education, or extended partnerships with technical mentors. PwC research on leadership development found that organizations with deliberate dual-skill career paths developed hybrid leaders more successfully than those relying on ad hoc approaches.

Third, the measurement and incentive systems need careful design. The article emphasizes aligning KPIs with transformation outcomes, but organizations must balance short-term business performance with longer-term capability building. Leaders who spend six to ten hours weekly learning may temporarily underperform on traditional metrics. Organizations should explicitly incorporate capability development into performance expectations and reward systems.

Fourth, succession planning takes on new dimensions. If domain owners with technical capabilities are scarce and difficult to develop, losing them creates acute risk. Organizations should build bench depth, create knowledge transfer processes, and consider retention strategies specifically for this critical population. Research from CEB (now Gartner) found that hybrid business-technology leaders face particularly high external demand, making retention challenging.

The Broader Context of AI Transformation

The domain owner thesis should be situated within broader research on digital transformation success factors. MIT's study of digital transformation found that while leadership capability mattered enormously, it interacted with strategy clarity, organizational culture, technology infrastructure, and change management practices. No single factor, including leadership capability, guaranteed success.

Moreover, the pace of technological change introduces a dynamic element. The technical skills required of business leaders continue to evolve from understanding analytics and machine learning to now encompassing generative AI, agentic systems, and emerging capabilities. This suggests that building the second muscle is not a one-time effort but an ongoing commitment to learning that must be sustained throughout careers.

Research from Oxford's Saïd Business School on absorptive capacity—organizations' ability to recognize, assimilate, and apply external knowledge—provides useful framing. Absorptive capacity depends on prior related knowledge distributed across the organization. The domain owner model essentially argues for distributing technical knowledge more broadly among business leaders to increase organizational absorptive capacity for AI innovation. This framing highlights that developing individual leaders serves the larger purpose of building organizational capability.

Recommendations for Business Leaders

For CEOs and boards evaluating their approach to AI transformation, several recommendations emerge from analyzing the domain owner model and alternatives:

• Conduct honest assessment of current leadership technical capabilities. The LinkedIn analysis suggesting only 17 percent technical skill sets among Fortune 500 leaders provides a benchmark. Organizations should evaluate their own leadership population, identifying gaps and potential.
• Make explicit choices about operating models. The domain owner approach is not the only path to AI transformation. Organizations should consider their context, culture, existing capabilities, and strategic priorities when choosing between distributed domain ownership, centralized AI teams, strong product management functions, or hybrid approaches.
• Whatever model is chosen, invest meaningfully in capability building. Whether developing technical skills among business leaders or business acumen among technical leaders, half-hearted efforts will fail. The six to ten hours weekly learning, hands-on projects, and extended development periods are real requirements, not optional enhancements.
• Address operating model barriers simultaneously with capability building. Developing technical skills among business leaders will not drive transformation if organizational structures, funding models, and incentive systems prevent them from applying those skills effectively.
• Embrace experimentation and iteration. Organizations should pilot different approaches in different domains, learn from results, and evolve their models rather than committing to a single approach across the enterprise immediately.

Conclusion

The McKinsey article makes an important contribution by identifying a critical bottleneck in AI transformation. The shortage of business leaders who can bridge domain expertise with technical capability is real, and organizations that develop these leaders will possess competitive advantages. The examples of Adam Boyd, Neesha Hathi, and Menno Van der Winden demonstrate that building this second muscle is possible and valuable.

However, the domain owner model should be viewed as one important approach rather than the only path to AI transformation success. The model's emphasis on individual leader capability may underweight team-based approaches, complementary functions like product management, and enabling technologies that reduce technical barriers. The significant time, effort, and organizational change required means this approach will not suit every organization or situation.

The most important insight may be the underlying principle rather than the specific model: successful AI transformation requires deep integration of business and technical thinking. Whether that integration happens within individual leaders, through structured partnerships, via strong product management, or through other mechanisms depends on organizational context. What matters is that organizations deliberately address the business-technology gap rather than hoping it will resolve itself.

As AI capabilities continue advancing rapidly, the pressure on business leaders to develop technical literacy will only intensify. Organizations that start now to build these capabilities, whether through the domain owner model or alternatives, will be better positioned to capture value from AI. Those that continue separating business and technology leadership risk being left behind as competitors learn to move faster and execute more effectively.

The question for every CEO and board is not whether their organizations need leaders who can bridge business and technology, but rather how they will develop those leaders at the scale and speed required to compete. The answer will vary by organization, but the urgency is universal.