Volume 1 of this project established the philosophy of Vanguard Leadership, the three forces that now shape the NEO era (networked, exponential, orchestrated) and the three capabilities a leader builds within it, the Cotruglian architecture that keeps honest exchange possible inside networked environments, and the diagnosis of what leadership is now up against. Volume 2 takes the philosophy and converts it into an operating system.
A reader who has picked up this book expecting it to promise victory should be told, on the first page, that victory is not what the book promises. I am honest about that because the alternative, promising something the doctrine cannot deliver, would discredit everything that follows. What the book promises is narrower, harder, and more useful than victory.
What it promises is an operating system you can run when the environment is structurally resistant, the odds are unfavourable, and the straightforward path is closed. An operating system that continues to function at five per cent probability as readily as at ninety per cent. An operating system whose outputs are not wins but something smaller and more durable: continued operation, a narrowing interval before erasure, disciplines that compound across campaigns, residue that outlasts any single one of them.
That is the whole claim. Everything in the rest of the book is in service of making the claim honest.
Who This Book Is For
The reader I have in mind is already living inside the conditions this book describes. She may be a founder whose product is correct for a market that has been pre-empted before she arrived. He may be an executive whose transformation programme has been running for three years under shifting mandates that never actually close. They may be a university leader watching institutions that once produced their own adulthood now produce procedural replicas of it. They may be a minister, an industrialist, a banker, a general, a regulator, a civic leader, a teacher, a founder with a single instrument trying to enter a market that incumbent capital spends heavily, every year, to keep closed.
The reader I have in mind is not waiting for a theatre in which her instrument will be welcomed on its merits. She has already learned that such theatres are rarer than the literature implies. She is trying to work out how to keep operating inside the theatres that actually exist.
This book does not exclude readers who have not yet had the experience. Readers newer to serious institutional life will still find the material useful. What the book assumes is not experience. It assumes a willingness to read for accuracy rather than for comfort. A reader who wants to be told that the instrument will eventually prevail on its merits if only the leader is true to herself should close the book now and save the time.
This is also a book for the Vanguards. The alumni of the programmes I have run across Southeast Europe, the Gulf and Africa, the African fellows, the executive participants in the cohort of senior women executives, the partners and students and practitioners who have entered this doctrine through whatever door carried them into it. I wrote much of what follows with the Vanguards specifically in mind. They know what the pressure of this environment feels like, and they have earned the right to expect from a book in their field an accounting that is not decorated.
What Volume 2 Adds
Volume 1 described what Vanguard Leadership is. Volume 2 describes how it runs.
The move from description to mechanism is the core work of this book. Volume 1 named the Cotruglian operating system and the five axioms that hold it in place. It named the three capabilities — Fear as Data, AI as Force Multiplier, Tribe as Coordination — that a serious leader now has to build. It named the human domain of fortitude: what remains to the leader once AI optimises everything else (judgment, beauty, meaning, trust and mastery). Those claims remain in place. Volume 2 does not replace them; where it refines them, it says so in the section that follows. Volume 2 takes each claim and asks the more demanding question: by what machinery, through what routines, and with what aids does a leader actually operate at this level of difficulty, in this environment, under these constraints, across the decades of a working life?
The chapters ahead answer that question across several registers at once. At civilisational scale, the book names the governance infrastructure the machine economy requires in order not to destroy the trust on which commerce depends. At cognitive scale, the book adopts the tradecraft lineage that runs from Heuer backward into the history of disciplined analysis and forward into the practices VL OS asks its leaders to run. At empirical scale, the book uses the testimony of executives who have been living this environment for years, and of alumni who have been running versions of the doctrine for two decades. At mechanism scale, the book names the specific tools that make the operating system run, and the four ways a reader can engage with them: to use, to learn, to experience, to build. At the scale of the human leader, the book names the disciplines and the fortitude that make everything else above possible.
Volume 1 was a diagnosis and a framework. Volume 2 is an operating manual.
How This Volume Stands to the First
Volume 1 was written to build awareness of the need to adapt fast to the NEO era, and of the need for a new kind of leadership. This volume was written after more campaigns, some of them lost, and in six places it is more careful. None of the six withdraws a claim of the first volume. Each states the conditions under which the claim holds, and a reader of both should know where they are.
Advantage, not victory. Volume 1 promised advantage: the leader who builds will, over years, hold a position the extractor cannot match. This volume keeps that promise and adds what it does not cover. Advantage compounds across a career; it is not promised for any single campaign, and the fortitude chapters are about the campaigns lost on the way.
Open networks and captured theatres. Volume 1 held that the NEO era structurally favours the Cotruglian builder, because networks expose deception and amplify what works. It has never been easier to build, and any network is worth having, because it will be harder and harder to sell anything. This volume adds the theatres Volume 1 called Machiavellian domains and thought rare: committees, eligibility rules and markets held by incumbents, where the network mechanism is blocked and a sound instrument can lose for years. In institutional work they are more common than Volume 1 assumed for markets.
Virtue and structure. Volume 1’s third axiom, that virtue is competitive architecture, stands. This volume adds that at the scale of institutions and machines, virtue becomes durable trust only when it is recorded and can be verified. Inside a tribe, people act on each other’s judgment without asking for approval case by case; between institutions, conduct is verified after the fact. The Trust Architecture instruments the first volume’s claim. It does not replace it.
What AI is to the leader. Volume 1 said that AI does not replace human judgment, though it does replace routing and routine roles, and that AI should be designed so that good conduct is its best strategy rather than relying on constraint alone. This volume works out what that means for the middle layers of an institution, and adds that today’s agents are governed by declared authority, receipts and review. Partnership is the aim; declared authority is the means. The machine supports orientation; the human owns it.
Tempo. Volume 1’s rule that the faster coherent cycle wins governs the decision loop against competitors. This volume adds pacing for the replies an institution demands: fast loops, slow replies.
One model, or a map. Volume 1 presented Mission Command as the leadership model built for the NEO world, in which the human role becomes more and more the oversight of processes run by people and by AI. This volume keeps it at the centre of the Command layer and draws a map of the conditions under which it, and the alternatives, hold.
The first volume’s frameworks keep their place. Its five axioms (existence precedes essence, but craft defines it; reputational capital compounds exponentially; virtue is competitive architecture; long-term orientation dominates; positive-sum coordination beats zero-sum competition) are the values every layer carries. Its four pillars are capabilities the layers build: critical strategic thinking in the Intelligence and Command layers, AI augmentation in the Interface layer, resilience in the Capability and Performance layers and the fortitude arc, tribal collaboration in the Command layer, the Trust Architecture and Federated Continuity. Of its three capabilities, AI as Force Multiplier is carried into the Interface layer and Tribe as Coordination into the infrastructure of trust; Fear as Data is taken as given, and the fortitude chapters build on it.
Why This Is Not Only a Manual
The reader may reasonably ask why the manual arrives with ancestry chapters, philosophical grounding, and a long arc on fortitude attached to it. The answer is not that the manual needs decoration. The answer is that a manual without the grounding beneath it is, in the NEO era, a document the machine economy is about to make free.
Manuals are the layer of knowledge AI commoditises most quickly. Any reasonably capable AI can now describe a decision protocol, walk a leader through a hypothesis-scoring matrix, or produce a battle card against a set of questions. Within a small number of years, the operational content of most leadership manuals will be available to any leader with access to a capable AI, which is already every leader, and will soon be every aspiring leader. A book that consists only of mechanisms has no moat against that commoditisation. Its instruction will be replicated and delivered at lower cost by systems that do not have the book’s author, the book’s publisher, or the book’s intended reader in any meaningful relationship.
What AI does not commoditise is the philosophical grounding that teaches a leader when not to trust the mechanism. Not the mechanism itself — the disposition that judges the mechanism from outside. A leader who has absorbed Heuer’s warning about her own mind, Cotrugli’s insistence on trade as governed practice, the military tradition’s acceptance that the first coherent story is usually wrong about something load-bearing, and Dante’s identification of fortitude as the virtue the other virtues need in order to keep functioning after defeat — this leader brings to every mechanism a standing question the mechanism cannot answer for her. Is the mechanism producing the right output, or producing an elegant output against the wrong frame? Is the AI supporting my judgment or silently shaping it? Is this manual, this book, this doctrine, still correct under conditions it did not anticipate?
The question is the book’s most important teaching. It applies to the manual you are now reading. No single source should be trusted fully, not the author, not the mechanisms, not the AI the mechanisms are run through, not this book. Scepticism of the manual is part of the manual. Leaders who cannot hold that stance will, over the coming years, be the leaders whose judgments their AI tools silently shape and whose strategic positions erode without their ever noticing that it was their tools that decided them.
The ancestry chapters exist because the stance cannot be installed by assertion. It has to be absorbed from sources that have earned the right to teach it across centuries. The fortitude arc exists because the stance cannot be sustained across decades without the human capacity to continue after the mechanism has failed. The manual is the middle of the book. The philosophy is why the manual still works when the environment is trying to commoditise it.
The Central Dependency
An operating system is not a machine. It is a set of disciplines, tools, and protocols that produce outputs only when a competent human leader runs them with intent.
This is the central honesty of the book. Vanguard Leadership has tools. You will meet many of them in the chapters ahead. The tools are real. They have been built across years of serious work, some over decades, and several of them are now operating in live institutions under real pressure. But the tools do not work on their own. A leader who expects the tools of VL OS to produce leadership the way a spreadsheet produces a sum has misunderstood what the book is offering.
The tools become load-bearing when a human with intent and fortitude operates them. Without that human, the tools are inert. The intelligence system is a system that helps a leader think; it does not think for her. The decision loops are loops a leader runs; they do not run a leader. The trust infrastructure is infrastructure that records what a leader does; it does not tell her what to do. The practice environment is an environment in which a leader trains; it does not train a leader who refuses to show up. The institution is an institution that a leader builds and maintains; it does not maintain itself.
This is not a weakness of the doctrine. It is a feature. A leadership doctrine that claimed to produce leadership without the leader would be a fraud, and the first sign that the claim is fraudulent is that the doctrine starts promising victory without the leader. Volume 1 promised advantage and meant it: advantage that compounds over years for the leader who does the work. What no honest doctrine can promise is victory that arrives without that work, or in every campaign.
Intent and fortitude are what the reader brings. The book brings the rest.
What Vanguard Leadership promises is an architecture inside which the reader’s intent has a better chance of producing usable outputs than it would without the architecture. What VL OS promises is that a leader who brings the right kind of attention to the right mechanisms at the right moments, across years of practice, will accumulate the kind of residue a leader without the architecture cannot accumulate. The architecture raises the ceiling of what intent and fortitude can produce. The architecture does not replace them.
A reader who takes nothing else from this book should take this: the tools will not save you. The discipline of using the tools will compound your own capacity. The disciplines, practised long enough, will produce the ability to keep operating at five per cent. That ability is the most valuable thing the book can give you. It is also the only thing the book can give you that is not a tool.
How to Read This Book
The book has three parts and a set of cases, organised in the order a serious reader will find most useful.
Part A is the doctrine. It opens with the ancestry, Cotrugli Reforged, and with The Architecture, the one chapter that maps the layers of the operating system: command, intelligence, interface, diagnostic, performance, environmental, capability, infrastructure, and the human at the centre. The rest of Part A elaborates the layers one by one, axis by axis. The Decision axis covers how the leader thinks, how judgment becomes the action of the institution and how the leader works with the machine, together with the intelligence tradition that grounds it and two chapters in the voice of the author’s AI companion. The Integrity axis covers how the operating system measures itself. The Execution axis covers the governance, training, evidence and trust infrastructure the system runs on, and closes with the four modes of engagement: Tools You Can Use, Methods You Can Learn, Environments You Can Experience and Governance and Trust You Can Build. Part A ends at the human centre of the system: will and telos, the reader’s own resistance, the doctrine running on its authors, and three chapters on fortitude. Not fortitude as a virtue to admire, but fortitude as the operational capacity to keep running the whole system when no individual campaign is producing a satisfying outcome.
Part B is the practitioner’s manual Volume 1 promised: hiring, the first ninety days, managing performance, rewarding, communication and battle rhythm, every ninety-day plan in the doctrine, the Vanguard Scorecard, the institution that outlives its founder, failure modes and tools. It is written for Monday morning.
Part C holds the cases: PayPal in 2000, Haier and Bayer, an engineering firm in Bosnia and Herzegovina, Reply and the Medici bank, DeepSeek, Kela, the field evidence from executives living the NEO environment, and the map of where this operating system lands among the frameworks a leader already carries.
The book can be read end to end. It can also be read by part. A reader wanting to act on Monday can begin with Part B, using Part A as reference. A reader facing a specific institutional difficulty can begin with the fortitude chapters and work back to the disciplines they require. The architecture of the book is linear by default but non-exclusive by design.
What You Will Get, What You Will Not
A reader who works through this book will leave with a specific set of things.
A diagnostic language that lets you read institutional environments for direction rather than for vocabulary. After a few weeks of practice, you will find yourself reading documents in a way that produces accurate conclusions about where authority actually sits, and you will find that those conclusions often disagree with what the documents claim.
A triage discipline for the routes you spend your time on. You will start distinguishing dead from fake-open from usable, and you will redirect a meaningful fraction of your calendar accordingly.
A register of disciplines for entering institutions without either collapsing into them or being rejected by them. You will learn to hide teeth under the tongue without losing the teeth.
A pacing instinct that converts time into information rather than time into anxiety. You will answer messages later. The decisions you make after twenty-four hours will be better than the ones you used to make inside twenty-four minutes.
A fuel arithmetic for your own resources. You will spend capital, attention, and relationships at rates that are sustainable across decades, not rates that would produce a successful single campaign at the cost of every campaign afterwards.
A triage for the people in your rooms. Carrier, owner, presence. Most of the people in most rooms are presence. Recognising which is which will compound your relational efficiency.
A practice of self-query at the end of every serious engagement. Four questions that, asked privately, give you a calibrated reading of your own position without requiring external validation.
A specific language for what you are already experiencing, which many readers will find is the most unexpected gift the book offers. The environment I describe is not a theoretical environment. It is the environment a significant number of serious leaders are operating inside right now, often without adequate language for it. Naming it accurately reduces its cost.
You will not get a promise that your instrument will win on its merits inside a theatre built to keep it out. You will not get a formula for institutional persuasion. You will not get encouragement to believe that the system will turn towards adulthood if you simply explain adulthood more clearly. You will not get the comfort that a book in this space is often designed to sell. You will also not get a warning about how hard the environment is; you already know.
What you will get is the accounting.
Ancestry and Acknowledgement
This book is single-authored by choice. Its intellectual lineage is not single-authored, and I want to name the ancestry clearly before any of the chapters begins.
The Cotruglian tradition is the longest. Benedetto Cotrugli wrote his Libro dell’arte di mercatura in 1458; it was printed in Venice in 1573 as Della mercatura et del mercante perfetto. In 2016 I produced the first English translation of his work, published privately as a gift to alumni and partners; Palgrave Macmillan’s edition followed, and I was invited to Venice for its launch. His argument that virtue, kept in honest records, is the structure trust stands on remains the oldest load-bearing claim behind everything in this book.
The tradecraft lineage enters through Richards Heuer, whose Psychology of Intelligence Analysis I treat as equipment rather than ornament. Behind Heuer is the broader discipline of intelligence analysis that grew up during the Cold War, and behind that the post-war sciences of decision and control: Herbert Simon’s work on bounded rationality, on which Heuer explicitly builds, and Norbert Wiener’s cybernetics. The disciplines of evidence-register integrity and hypothesis falsification that sit inside VL OS derive from that lineage.
The military-doctrinal line runs through Boyd’s OODA work, through the Alberts-Hayes writings on Power to the Edge and Network-Centric Warfare, through the Mission Command tradition that begins in Prussian military theory and extends into contemporary practice. Where the doctrine in this book insists that speed matters less than orientation, that intent matters more than order, that decentralised execution under clear Commander’s Intent outperforms centralised micromanagement, it is drawing from this lineage.
The machine-economy architecture that runs through the NCTE and PAC protocol families is the product of a multi-year collaboration. The published work is co-authored with Mario Brčić, Matjaž Gams and Tali Režun. The VIS framework is co-authored with Damir Stručić. The EU-DUCCI proposal is co-authored with Stručić as well. Wherever this book draws on the technical architecture, the co-authors are credited in the endnotes. The doctrinal synthesis in the book itself is mine, which is why the book is single-authored; but the technical substance that supports the synthesis is collective.
The empirical ground is provided by the executives, alumni, and practitioners whose testimony and experience shaped the diagnostic chapters. Many of them are anonymised. Some of them have requested not to be named even indirectly. The alumni archive, from which much of the bottom-up evidence is drawn, is accessible separately for readers who want to engage with the lived voice behind the doctrine.
What I owe all of these ancestors, co-authors, and contributors is the same thing the book owes its readers: honest accounting. I have tried to give it.
The Reader’s Task
The book has been written to be used, not only read. A reader who reads it without ever running one of the disciplines it names will not benefit from it. A reader who runs even two or three of the disciplines deliberately across a quarter will find that the disciplines begin to produce residue that compounds.
Begin wherever the pressure is greatest in your own environment. If you are currently inside a structurally closed institutional campaign, begin with the fortitude chapters and work backwards. If you are making calendar and coalition decisions, begin with the triage disciplines in the practice chapters. If you are wondering which tool to adopt first, begin with the tools section and match the tool to the problem. The book is designed to reward this kind of use.
One last thing. The book is not the doctrine. The doctrine lives in the people who run it. If the book is doing what I want it to do, the reader will eventually close it and go back to her work with a clearer sense of what she has been doing and what she can do next. That is the highest service a book of this kind can provide. Everything written here is in service of that one outcome.
The first operation in any serious life is the one the reader is already running. The second begins when she decides.
