Check out these reader favorites:
Shifting Ground
I’m noticing a shift in how serious people talk about running companies. After two decades of leadership literature celebrating vision-setting and empowerment through distance, the evidence is piling up that the best-performing firms are run by CEOs who actively shape daily execution. Delegation has diminishing marginal returns when the delegater loses awareness of the work that’s actually being executed.
Simultaneously, the AI gold rush is producing its first wave of expensive failures: companies running dozens of disconnected pilots, accumulating technical debt faster than productivity gains, and discovering that 55% coding speed improvements reverse into liabilities when inexperienced developers deploy AI-generated code into legacy systems without oversight.
The market is starting to distinguish between operators who architect how work gets done versus those who simply exhort people to work differently. That distinction matters because it determines which capital expenditures actually generate returns versus those that are merely expensive theater.
The Builder-Operator
Harvard Business Review’s research on high-performing companies exposes an uncomfortable pattern for conventional leadership thinking. The CEOs creating sustained competitive advantage aren’t the ones giving inspirational TED talks about empowerment, they’re those obsessing over customer-value metrics, designing how work gets done, using experiments to make decisions, teaching organizational toolkits, and driving continuous improvement. Imagine that. Not much time to prepare for your close-up when you’re in the business of actually getting work done.
There’s a meaningful difference between a leader who demands approval for every decision and one who architects the frameworks that enable better decisions at scale. The former creates bottlenecks; the latter fosters the development of new organizational capabilities.
In terms of capital allocation, this distinction matters enormously. When you evaluate a company, you’re not just betting on current cash flows, you’re betting on the system’s capacity to generate improving cash flows over time. A CEO who coaches teams on how to identify and eliminate operational waste is building an appreciating asset. A CEO who delivers quarterly speeches about “operational excellence” while the actual work systems remain unchanged is not.
This connects directly to culture and organizational effectiveness. MIT Sloan Management Review’s research shows that high-performing organizations under flexible work arrangements share characteristics unrelated to physical attendance policies, they measure results rather than presence, provide genuine autonomy, and redesign spaces to drive superior outcomes. Organizational culture isn’t what leadership says matters; it’s what the operating system of the company rewards and enables.
When companies constantly celebrate their teams’ extraordinary efforts to meet deadlines or resolve crises, they’re inadvertently revealing broken systems rather than demonstrating cultural strength. Organizations that require heroics to function are admitting their processes don’t work. Resilience shouldn’t depend on individuals repeatedly going above and beyond, it should be embedded into how inputs become outputs.
The practical test is simple: if remote work policies or inspirational all-hands meetings are the primary tools leadership deploys to address cultural challenges, the organization has a system design problem it’s attempting to solve with words. That rarely works, and is expensive. When evaluating management teams, look for evidence of system-building rather than just strategic vision.
Can they articulate how work actually flows through their organization?
Do they have frameworks that scale decision-making rather than centralize it into disconnected silos?
The answers to these questions distinguish leaders who build valuable operating assets from those who preside over slowly deteriorating companies.
Discipline, Not Proliferation
The AI adoption playbook most companies are running (launch pilots across multiple departments, measure productivity gains, scale what work) sounds reasonable until you examine what’s actually happening. Harvard Business Review warns companies pursuing generative AI adoption against launching numerous scattered pilots across departments. Instead, organizations should concentrate on deep implementation in one strategically important domain where AI can create interconnected value.
The technical debt problem becomes readily apparent. MIT Sloan Research documents that while generative AI tools can boost individual developer productivity by up to 55%, rapid deployment without proper oversight creates dangerous downstream costs. In environments with legacy systems, inexperienced developers who use AI to generate code often compound existing problems rather than solving them. The short-term productivity gain can reverse into a long-term liability when the code requires maintenance, integration, or debugging.
The philosophical confusion runs deeper. MIT SMR’s most-read article this year argues that organizations adopting AI need to first clarify whether their decisions reflect their core operating philosophy or inadvertently adopt the philosophies embedded in the large language models they’re purchasing. How AI defines knowledge, represents reality, and creates value shapes outcomes in ways most companies don’t examine until it’s too late.
A company that values deep expertise and judgment but deploys AI tools trained on internet-scale text averaging might find itself systematically devaluing the very capabilities that created its competitive advantage.
The capital allocation error ties these together. Running dozens of disconnected AI pilots is expensive, not just in software licenses and consulting fees, but in management attention, organizational change capacity, and opportunity cost. Each pilot consumes resources that could be deployed elsewhere. More importantly, scattered experimentation sends a signal about management’s strategic clarity. When a company announces broad AI initiatives without articulating where they’re focusing deeply, it reveals either strategic confusion or communications theater aimed at investors rather than operational reality.
The alternative framework emphasizes concentration over proliferation. Deploy AI deeply in one strategically critical domain where it creates interconnected value, rather than superficially across multiple functions. This approach treats AI adoption as a capital allocation decision: what’s the expected return on incremental investment, what risks are we assuming, and how does this compare to alternative uses of the same resources? From an operating perspective, this means most companies should be running fewer AI experiments, not more. Organizations need to build the long term know-how necessary to truly, foundationally transform how they produce their goods and services. This is a long game, not a momentary, aesthetic choice. The week’s research converges on a framework for evaluating operating companies that differs meaningfully from conventional approaches.
For Operators
Focus capital allocation on system-building, not activity. Every dollar spent on leadership development, process improvement, or technology deployment should have a clear theory about how it builds organizational capabilities that generate improving returns over time.
Treat AI adoption as a capital allocation decision requiring concentration; run fewer experiments with more intensity in strategically critical domains.
Design an execution-centric culture. If you find yourself implementing mandates or delivering inspirational messages to change behavior, step back and ask what system design changes would make the desired behavior natural rather than forced.
For Investors
Evaluate management teams on evidence of system-building. Look for CEOs who can articulate how work flows through their organization, what frameworks enable decision-making at scale, and how they develop organizational capabilities over time. This matters much more than strategic vision statements.
Discount broad AI initiatives lacking strategic focus: companies announcing numerous pilots across departments without articulating where they’re deploying deeply are either strategically confused or performing for investors.
Value operational discipline over growth narratives. Companies demonstrating rigorous operational discipline, clear data monetization strategies, and system-designed cultures are building compounding advantages that show up in improving returns on incremental capital over time.
Final Thought
Nobody knows exactly how AI will reshape competitive dynamics, or which leadership frameworks will prove most valuable in the next decade, or even whether current operating strategies will survive the next economic cycle. However, we do know that companies that build robust operating systems, make disciplined capital allocation decisions, and treat organizational capabilities as appreciating assets have historically generated superior returns across multiple economic cycles. This week’s research suggests that the pattern is reasserting itself, the builder-operator is returning, and the market will stand up and take note.
- j -
John Brewton documents the history and future of operating companies at Operating by John Brewton. He is a graduate of Harvard University and began his career as a Phd. student in economics at the University of Chicago. After selling his family’s B2B industrial distribution company in 2021, he has been helping business owners, founders and investors optimize their operations ever since. He is the founder of 6A East Partners, a research and advisory firm asking the question: What is the future of companies? He still cringes at his early LinkedIn posts and loves making content each and everyday, despite the protestations of his beloved wife, Fabiola, at times.
Sources
Hands-On Leadership and System-Building
Scott Cook & Nitin Nohria, “The Surprising Success of Hands-On Leaders,” Harvard Business Review, November–December 2025.
Focuses on CEOs at Amazon, Danaher, RELX, and Toyota who actively shape how work is done, acting as teachers and system-builders rather than distant delegators.
https://hbr.org/2025/11/the-surprising-success-of-hands-on-leadersHarvard Business Review November–December 2025 issue table of contents.
Includes related pieces on C-suite effectiveness, strategic hibernation, data monetization, and a Gen AI playbook that inform the system-building and capital allocation lens in the memo.
https://hbr.org/archive-toc/BR2506
AI, Technical Debt, and Philosophical Alignment
“The Top 10 MIT SMR Articles of 2025,” MIT Sloan Management Review (roundup).
Highlights pieces such as “Philosophy Eats AI” and work on the hidden costs of coding with generative AI, technical debt, and leadership challenges around AI decision-making.
https://mitsmr.com/4pGBjYw
Culture, Hybrid Work, and Resilience
“The Top 10 MIT SMR Articles of 2025,” MIT Sloan Management Review.
The same roundup also surfaces research on hybrid work, return-to-office debates, meeting effectiveness, and resilience design—background for the memo’s “culture-as-system” framing.https://sloanreview.mit.edu/article/the-top-10-mit-smr-articles-of-2025/
Data, Capital Allocation, and Operational Discipline
Harvard Business Review November–December 2025 issue description.
Includes articles such as “The Gen AI Playbook for Organizations,” “What Every Company Can Learn from Private Equity,” and “How to Monetize Your Data,” which inform the sections on data as a monetizable asset and PE-style operating rigor.
https://store.hbr.org/product/harvard-business-review-november-december-2025/BR2506
Love it! TRUE CEOs are not giving inspirational TED talks... they're laying efficient foundations for growth.
Thanks, John, for another extremely well-written and researched post. Capital allocation seems to be a really big problem in universities as well, as many of them are investing in lots of different pilots for lots of different tools without really taking the time to instead take stock and work out which direction we should be going in.
Thanks again for this article. Definite pause for thought.