The Coherence Standard — The Manifesto

Prologue

A Word You Didn't Know You Needed

On the thing you've been feeling but couldn't locate

You were in the room. Maybe last quarter. Maybe last week. The strategy was sound, the language confident, the plan agreed. Everyone walked out aligned.

And somewhere between that room and Monday morning, something essential disappeared. Not the plan itself — but whatever made it intelligible to the people who now had to carry it.

Nobody mentioned it. Nobody ever does.

That disappearance has a cost — and not the kind your dashboard measures. Not the overrun, the missed deadline, the attrition number. Something underneath all of those. The thing that explains why a project that succeeded still felt hollow. Why a team that stayed stopped believing. Why a strategy that should have changed everything quietly changed nothing.

You can feel it. You just don't have a word for it yet.

This manifesto doesn't start with a framework. It starts with a word — and once you have it, you won't be able to unsee what it names.

The word started as a physical property. Latin: cohaerere. To stick together. Not align. Not coordinate. Stick. The kind of thing you only notice when it stops happening — when parts that were supposed to behave as one thing start quietly behaving as several, each making perfect sense on its own, none adding up to anything whole.

We ship outputs and assume the meaning travels with them. It doesn't. Meaning has no tracking number. We assemble brilliant people and call the assembly a team. Stack decisions and call the stack a strategy. And then we wonder why the gap between what we intended and what landed keeps showing up, quarter after quarter, wearing a different name.

So the word grew up. Cohere becomes coherence. Not just closeness but sense. Not just connection but consistency — the condition that what you said on Monday doesn't contradict what you did on Thursday unless there's a reason you can actually explain.

We used to treat coherence as elegance. A nice-to-have for organisations with the luxury of caring. Now it's a survival trait. And most organisations don't have it.

Here's what's strange. A company can look completely healthy while coming apart at the seams. The calendar stays full. The language stays confident. The deliverables keep arriving. People keep "syncing." And yet something subtle is failing — the same discussion returns with a new title, the best people spend their best hours translating and smoothing rather than building, and nobody can answer the question surfacing in the back of everyone's mind: why does this keep happening to us?

Not because they can't do the work. Because the work can't hold together long enough to become cumulative.

This has a name: Fragmentation.

It isn't chaos — chaos is at least visible. Fragmentation is orderly. It's what happens when a system keeps moving after it's stopped making sense. Every new tool reasonable. Every new workflow justified. Every new layer of process designed by someone who genuinely wanted to protect the work. And in aggregate, the whole loses its stickiness. The internal logic stops hanging together. Motion becomes the perfect disguise.

Motion is easy to measure. Sense is harder. So we produce motion.

Here's the structural reason: every person in your organisation operates from their own partial, quietly diverging picture of reality. What matters. What's true. What was decided and why. Without architecture that keeps those pictures aligned, every transfer becomes a negotiation between incompatible versions of the same situation. The gap accumulates invisibly. The cost of that gap is what we call the reset economy — and you're paying into it right now, whether you know it or not.

You've felt this. The project somehow exhausting despite going well. The strategy day that produced genuine energy and then — nothing changed. The new person who arrived full of capability and left eighteen months later, quieter than they came. Nobody quite understood why.

It would be comforting to blame people. But your people aren't the problem. Meaning erodes in motion. Decisions survive but not the reasoning. Plans exist but no longer make sense inside the world that changed around them. Work arrives clean and complete, except for the part that mattered.

So your smartest people become couriers of context. They write long messages not to communicate but to prevent relapse. They schedule meetings not to decide but to reconstruct. They over-document as self-defence. Nothing breaks loudly. People compensate quietly. And the organisation mistakes the compensation for the work.

If this sounds familiar, it's not because you're cynical. It's because your organisation is paying a debt nobody named — and unnamed costs accumulate the longest.

Now here's where it gets urgent. Speed is cheap. AI is sold as the great accelerant — and it is. But when speed is available to almost everyone, it stops being an advantage and becomes a baseline. The differentiator won't be who moves fastest. It will be who can move without losing the plot.

Coherence is the refusal to accept that loss as normal. Not as a mood, not as a culture poster. As a standard. If we act, we can account for what that action meant. If we change direction, we can show the line from then to now. If we ask people to move fast, we owe them a world where the work still holds together when it lands.

Coherence is unfashionable because it exposes cheap tricks. It makes performative busyness look small. It makes "complexity" sound less like sophistication and more like an excuse. Good. Once you start noticing what doesn't endure, you start seeing why so much effort feels heavier than it should.

So here's the invitation. Not to agree, not to admire. To ask, wherever effort is being spent: does this hold together, or does it merely move?

If that question lands — you already know what comes next.

Topic 1

The Bill You’re Already Paying

Why the most expensive problems in your organisation look exactly like normal work

Every leadership team has a theory for why things aren’t quite working. The meetings run long. The handoffs fail. The deadlines slip. Fix those, and everything improves. It’s a compelling theory. It’s also wrong.

The actual cost isn’t in the failures. It hides inside the successes. Inside the competent, well-intentioned, completely ordinary working week: the meeting that opens with ten minutes of “just getting everyone up to speed.” The project that restarts because nobody documented why the first version failed. The customer who explains their situation again — not because the record is empty, but because what it holds isn’t quite enough to act on.

None of this announces itself as a problem. It announces itself as work. And because it looks like work, it accumulates unnoticed — year after year, inside companies that are otherwise performing well.

We have built entire professions around measuring the wrong thing with extraordinary precision. We can tell you the cost-per-click of a campaign down to four decimal places. We cannot tell you the cost of asking a loyal customer to explain themselves for the fourth time. One of these numbers is in the dashboard. The other is quietly deciding whether that customer comes back.

Your organisation is paying this tax. The only open question is whether you know what it’s called — because you cannot design against a cost you haven’t named. The name is Coherence Debt.¹

Coherence Debt accumulates every time a human being is required to carry what the system should carry — to re-explain, re-remember, re-build what was already built. It is the organisational cost of a system designed for output rather than continuity.

Like financial debt, it accrues interest. Unlike financial debt, it carries no due date and appears on no balance sheet. It hides inside the work itself, indistinguishable from normal effort — until the organisation notices it is spending a growing share of its capacity maintaining what it has already built rather than building what comes next.

The people absorbing this cost are almost always your best people. They’re the ones capable enough to bridge the gaps, translate between teams, hold the thread in their heads when no architecture exists to hold it. They do it because they care. They do it until they don’t. And when they leave, they take with them the memory the system was never designed to hold — and the organisation discovers, too late, that its continuity was a person, not a property.

Customers pay this tax in a different currency — not money, but belief. Every time a system forgets them, something small but irreversible happens. They update their expectations downward. One day, without drama, they leave — not because of one bad moment, but because the accumulated weight of being forgotten finally outweighed the inertia of staying.

Coherence Debt is what individuals pay. But it produces a systemic condition that accelerates it: the reset economy. Every transition clears the shared understanding — not because anyone intends it, but because the architecture was never designed to carry meaning forward.

Companies track burn rate religiously. Very few can tell you their reset rate. You don’t pay it all at once. You pay it in instalments — one extra meeting, one extra clarification, one extra workaround — until the company is moving fast and going nowhere with incredible consistency.

The Machine That Scales It

Until recently, this tax had a natural ceiling. Human systems are slow enough that the gaps between intent and execution were survivable. The capable people absorbed them. The organisation compensated. There was always someone who remembered where the thread was.

AI removed that ceiling.

In 2014, Amazon began building an automated recruiting engine. The idea was efficient and logical: train a system on a decade of hiring data and let it surface the best candidates, faster than any human team could. The tool learned. It learned which résumés had historically led to hires. It learned which patterns correlated with success inside the company. And because the decade of data it trained on reflected a workforce overwhelmingly composed of men, it learned something else too: that being female was a predictor of rejection. The system began penalising résumés that contained the word “women’s” — a women’s college, a women’s athletics team. It downgraded candidates not on the basis of competence, but on the basis of a pattern the company had never examined and the machine had no reason to question.²

Amazon disbanded the project in 2017. The lesson reported was about bias. But the deeper lesson — the one most organisations still haven’t absorbed — was about coherence. The bias didn’t originate in the algorithm. It had accumulated across a decade of human decisions, each one locally reasonable, none of them examined as a system. The machine didn’t create the incoherence. It inherited it, and then it scaled it — at a speed and confidence no human process could have matched and no human review cycle could catch in time.

Every organisation deploying AI into fragmented operations is running the same experiment. The tool is different. The dynamic is identical. A system that couldn’t carry its own meaning at human speed is now failing to carry it at machine speed — more often, with less opportunity to catch what’s going wrong before it accumulates.

Speed is not the problem. Speed applied to a system that cannot hold its own logic is the problem. And right now, every company in the world is being sold the cure — before anyone has agreed on what the condition actually is.

Topic 2

Coherent or Merely Confident

What intact actually looks like — before the pressure is applied

There’s a peculiar problem with most organisational thinking: it describes what’s broken without ever establishing what healthy looks like. It catalogues symptoms without a baseline. Offers remedies without being able to say what recovery would feel like.

Before we talk about how things break, we need to talk about what intact means.

Coherence is the condition of a system that can carry its own meaning across distance, time, and the boundaries between people. Not because it’s protected from pressure — because it’s designed to hold its shape under pressure.

Coherence is not harmony. A room where everyone nods might simply be a room where nobody is willing to name the contradiction. The most harmonious organisations we’ve ever seen were the ones most violently committed to their own narrative. Harmony is easy to manufacture. Coherence is not.

It’s not alignment either. Alignment is a photograph. Coherence is what allows the direction to survive the journey from the photograph to Monday morning. And it’s not culture — culture is what people actually do when the pressure is on; coherence is whether what they do connects to what the organisation is trying to become.

The most dangerous error — and the most common one — is mistaking insulation for coherence. Full alignment. Confident language. Everyone nodding. And underneath, the constraints are being quietly ignored, the difficult feedback filtered before it reaches anyone who could act on it.

Insulated systems aren’t protecting coherence. They’re freezing their picture of reality and refusing to let what’s actually happening correct it. They don’t fragment slowly. They fragment suddenly — because they’ve been accumulating unacknowledged reality until the distance between what the system claims and what is true becomes unsurvivable.

Ask leaders if they communicate clearly, and the answer is almost always yes. Ask their teams, and a different answer lives just beneath the polite one. That gap — between the confidence at the top and the translation happening below — is not a failure of honesty. It’s a failure of architecture. No amount of open-door policy closes it.

Real coherence is permeable. It lets reality in. It is tested continuously by the decisions the organisation makes under pressure. And it is only real if it can be disrupted.

What we mean by a coherence field is not a boundary you draw or a culture you declare — it’s the live tension between what the organisation intends and what it actually does, held visible, held honest, and actively maintained.

Where those two planes align, the field is strong: decisions are predictable, transfers carry meaning, contradictions surface before they become structural. Where they diverge silently, the field weakens — not because anyone decided to let it, but because no architecture existed to surface the gap before it accumulated.

The coherence field is the thing worth building for. Everything that follows — structure, culture, empathy, time — exists to keep it real.

Structure: The Architecture of Relationships

Walk into a fragmented company and you’ll find an elaborate hierarchy. Clear reporting lines. Defined ownership. A chart that would make an MBA proud. Walk into a coherent one and you’ll find something that looks similar and is entirely different underneath. The hierarchy may be identical. But what’s actually running the place isn’t the chain of command. It’s the network of relationships through which understanding travels.

A hierarchy tells you who decides. A relationship tells you how that decision travels — who carries it forward, what reasoning must survive the journey for it to still be intelligent in other hands. Fragmented companies design for authority and hope relationships emerge. Coherent ones design for relationships and let authority serve them.

Culture: What Travels Between Meetings

Every company runs on two cultures simultaneously. The stated one — walls, handbooks, all-hands language — and the practiced one: what actually governs behaviour when the choice is between carrying the story and meeting the deadline.³

In fragmented companies, these two cultures diverge quietly. Nobody announces it. The gap accumulates between what the company claims to value and what its structures actually reward. Over time, the values on the wall have nothing to do with the work being done beneath them.

In coherent companies, culture isn’t a collection of stated values. It’s the mechanism by which shared understanding persists — the unwritten contract that says: when work moves between us, we don’t drain it of its story. We pass the thread.

Empathy: How Reality Enters the System

Empathy is almost universally misunderstood in organisational life. It gets treated as a personal quality — a soft skill, something that belongs in HR conversations. In coherent companies, it’s calibration: the discipline of asking what this looks like from a different position in the system, and what it costs if we don’t account for that.

Think about the last product launch that made complete internal sense and confused the people it was built for. That’s not a communication failure. It’s what happens when a system stops letting external reality correct its internal logic. The organisation became coherent with itself. It lost coherence with the world. Empathy is the mechanism that prevents that collapse. When it exits, the coherence field stops being permeable and starts being managed.

Time: The Dimension Everyone Forgets to Design For

Every coherent system must hold together across time. Context decays. Rationale fades. The “why” behind a decision becomes progressively harder to reconstruct the further you move from the moment it was made. And yet almost every organisation architects for the present — how decisions are made now, how teams collaborate now — without asking the question that time makes urgent: will the understanding that animates this today be intelligible to those who inherit it tomorrow?

Designing for time, structurally, means treating certain context as load-bearing rather than incidental. The reasoning must survive alongside the conclusion — or the conclusion eventually can’t be acted on intelligently by anyone who wasn’t in the room.

These four dimensions are integrated, not parallel. Structure determines whether meaning can travel. Culture determines what happens to it when it arrives. Empathy determines whether the picture being carried is an accurate one. Time determines whether the system can hold coherence beyond the tenure of the people who currently maintain it. Pull on any one of them and the others move.

You already know which one is weakest in your system. That knowledge is the beginning of the work.

Topic 3

What Has to Survive the Journey

The three elements that hold meaning together — and the five places they break

Coherence doesn’t break everywhere at once. It breaks somewhere specific. At a crossing. A transfer. A moment where something that was whole on one side arrived diminished on the other — and nobody designed that moment, so nobody noticed when it started failing.

The question is: what was supposed to survive the journey?

The Continuity Bundle

As meaning moves through a company, three elements must travel together. Separately, they’re data. Together, they’re understanding.

The continuity bundle is the minimum unit of meaning that must survive every transfer for work to remain intelligent in other hands. It is not a document, not a briefing, not a handover note. It is the living combination of why something was decided, what surrounds it, and where things actually stand.

Without all three, whoever receives the work inherits the conclusion without the reasoning, and begins making decisions on a foundation they cannot fully verify.

The three elements are Intent, Context, and State — together forming the ICS triad at the core of every coherent transfer.

Each element is irreplaceable. They are not three ways of saying the same thing. They are three distinct properties of meaning — and the loss of any one produces a failure mode that the other two cannot compensate for.

Intent — The Direction Vector

Intent is the rationale that transforms a decision from an instruction into a principle. It answers not what was decided but why — the values trade-off that was made, the alternative that was rejected, the future the decision was designed to produce. Without intent, work can be executed but not extended. The team that inherits a decision without its intent can follow it precisely in conditions identical to those that produced it. The moment conditions change — and they always do — they have nothing to navigate by.

Intent is the most fragile element of the bundle because it is the least required for immediate execution. Work gets done without it. Decisions get enforced long after anyone can explain them. What remains is motion that can no longer account for itself.

Context — The Surrounding Reality

Context is what makes information intelligible rather than merely accurate. A fact without context is a trigger without a target. A decision without context is a conclusion the next person cannot verify — they can only comply with it. What was true at the time? What constraints shaped the choice? What did we know, and what did we not yet know?

The failure mode of missing context is insidious precisely because it is invisible at the moment of transfer. The sender rarely knows what context they carry that the receiver does not have. The receiver rarely knows what context is missing until they act on the information and discover, too late, that something essential was assumed rather than stated.

State — The Ground Truth

State is where things actually are — not where the roadmap claims, not where leadership hopes, but the honest current condition that the whole picture must be calibrated against. It is the hardest element to maintain because it changes continuously and because the pressure to make it convenient is constant.

When state drifts from honest to optimistic, the organisation begins making decisions based on a version of reality it has chosen to believe rather than one it can verify. This is not deception — it is the natural tendency of systems to preserve their own coherence by adjusting their representation of reality rather than their behaviour. The result is a picture of the organisation that is internally consistent and externally wrong. Decisions built on it are locally rational and globally unpredictable.

The ICS triad is not a checklist. It is a structural requirement. Lose Intent and you produce execution without wisdom — teams that follow rules into conditions the rules were never designed for. Lose Context and you produce confidence without grounding — decisions that are correct in the abstract and wrong in the specific. Lose State and you produce strategy without calibration — organisations optimising toward a version of reality that no longer exists.

Remove any one of them and the transfer produces a locally rational actor operating from an incomplete picture of reality. At scale, those incomplete pictures diverge. That divergence is fragmentation — not caused by failure, but by architecture.

Memory Versus Storage

Storage captures what happened. Memory preserves why it mattered.

Fragmented companies have built infrastructure that is extraordinary at the first and almost incapable of the second. The reason is structural: storage is entity-based — it records what exists and what occurred. But a coherent picture of reality requires the relational layer too — how things connect, what depends on what, why those connections were forged. Storage can never be memory because it only ever captures one plane of a two-plane problem.

You can retrieve a ticket from three years ago in seconds. You cannot explain, with any confidence, why the decision it records was the right one at the time — or whether it still is. The record exists. The reasoning behind it is gone. And so the organisation does what organisations always do when understanding runs out: it calls a meeting, rebuilds the context from scratch, reaches roughly the same conclusion, and files another ticket.

We have named the problem “knowledge management” — which is a bit like calling a leaking roof “precipitation management.” The systems we have are extraordinarily good at storing what happened and constitutionally incapable of preserving why it mattered. Storage grows. Memory evaporates. And the distance between what the company knows and what it can actually act on keeps widening — invisibly, politely, one reset at a time.

A remembering system doesn’t keep artefacts. It keeps the threads between them.

That is the difference between a company that merely moves and one that actually progresses.

Five Coherence Boundaries

The continuity bundle doesn’t degrade randomly. It degrades at specific, predictable locations — boundaries where the transfer is structurally at risk. There are five. Each has a failure signature. Together they form the organisation’s coherence maintenance map.

Aspirational Coherence — The Source

The boundary before anything travels. It breaks when the intent is internally inconsistent before it is ever communicated: strategy that contradicts itself, priorities that conflict without acknowledgement, direction that cannot be carried because it was never coherent to begin with. Nothing that follows can compensate for a bundle that was already fragmented at origin.

Transitional Coherence — The First Crossing

The boundary at the moment of handover between people, teams, or layers. It breaks when both sides believe the transfer succeeded and neither is right. The sender carried context they didn’t know they were carrying. The receiver interpreted through assumptions they didn’t know they were making. The bundle arrived, but diminished.

Local Coherence — Within the Work

The boundary inside the work itself — not at a crossing, but over time. It breaks quietly, as people rotate and the reasoning behind decisions evaporates while the decisions themselves remain. Eventually the organisation is maintaining something it can no longer fully explain.

Compositional Coherence — Between Parallel Efforts

The boundary where separately coherent work must combine into a coherent whole. It breaks when every team is locally sound and the whole is globally incoherent: initiatives that contradict, dependencies never made visible, strategies that compete without acknowledgement. Nobody failed. The architecture did.

Enactment Coherence — Through to Outcomes

The boundary at the moment of result — the last and most honest test. It breaks when the results arrived, the metrics moved, and something essential was lost so gradually that no one noticed until the distance became structural. By the time enactment coherence fails visibly, the degradation has usually been accumulating across the other four for some time.

These are not five ways things break. They are five coherence boundaries — each structurally distinct, each generating a specific maintenance obligation, each requiring a designated owner. An organisation that maintains all five by design is rare. An organisation that discovers all five by crisis is everywhere.

Topic 4

Relational Intelligence

Where intelligence has moved — and what it costs to keep looking in the wrong place

Here is the question nobody asks after an organisational failure.

Not what went wrong. Not who let it happen. The question underneath those questions — the one that would actually be useful: where was the intelligence supposed to live?

We didn’t have language for that question, so we blamed people instead. Leadership failure. Communication breakdown. Cultural misalignment. We treated it as a talent problem and hired accordingly — hoping individuals could supply what the system refused to provide. That substitution worked for a while. It doesn’t anymore. Because the intelligence being asked for is no longer the kind that lives inside individuals. It’s the kind that lives between them.

Intelligence Moved. The Org Chart Didn’t.

Intelligence has always migrated. First it lived in people — the craftsman, the foreman, the expert who carried the knowledge in their hands. Then it moved into systems — processes, workflows, standardised operations. Then into data — the dashboard, the algorithm, the belief that if you measure enough you understand enough. Each migration changed what winning required. The companies that noticed first redesigned accordingly.⁴

Now it’s moved again. Not into any single thing. Into the relationships between things — between decisions and actions, between intent and understanding, between what a company has learned and what it is about to do with that learning.

Intelligence didn’t disappear. It relocated. And the companies that can’t find it are still looking in the wrong places.

It lives in the relationship with customers, who carry the full weight of every prior interaction and notice immediately when the company has forgotten them. In the relationship with teams, whose output depends on shared constraints and candour. In the relationship with every stakeholder who is not an audience for outcomes but a participant in the meaning those outcomes are supposed to create.

This is where the work actually lives. Not in the tasks, not in the systems — in the quality of connection between them.

The companies winning in ways competitors can’t reverse-engineer aren’t doing it with smarter people or cleaner processes. They’re doing it because meaning survives movement inside them — and that property lives between the parts, not inside any one of them.

What Translation Does to Meaning

Every transfer is a translation. And translation is where meaning degrades — not because people are careless, but because the person passing the work carries context they don’t know they carry, and the person receiving interprets through assumptions they don’t know they’re making.

Meaning degrades across boundaries the way a signal degrades across distance. It isn’t failure. It’s physics. And no amount of individual effort reverses it.

Relational Intelligence is the architecture that reduces that degradation — not to zero, but enough that what arrives still carries what mattered. Still holds the thread. Still gives the next person enough to continue rather than restart.

You can’t solve this locally. You can’t make an individual better at it. It requires a designed relationship between human judgment and continuity — one where meaning survives under pressure rather than only when the most capable person in the room happens to remember.

Which raises the question every organisation now has to answer: in that designed relationship, what does each side do?

The Division That Makes It Real

The decisions about what matters and why — the setting of direction, the naming of trade-offs, the judgment of when the gap between what was intended and what is happening has grown wide enough to require correction — these are irreducibly human responsibilities. They involve values, not just calculation. They require the capacity to hold two things in tension and decide which one yields. No machine makes that call without corrupting what it is supposed to protect.

Humans preserve meaning.
Machines preserve memory.

This is the division that most AI strategies get backwards. They automate judgment and wonder why coherence deteriorates. They deploy tools to decide and are surprised when meaning disappears. The machine is not a replacement for the human who holds the thread. It is the infrastructure that makes holding the thread possible at scale.

What machines carry is what human memory cannot hold at scale: the thread across translations where it would otherwise degrade, the connections between past decisions and present reality, the signal when coherence is slipping before the distance becomes structural. They do not decide. They make the right human decision possible by ensuring the right understanding is present when it’s needed.

Get this division right, and coherence becomes a property of the system rather than a function of who happens to be in the room. Get it wrong, and the result is the most sophisticated fragmentation engine in organisational history — a system eroding meaning faster than any review cycle can catch, while reporting clean metrics to leadership.

Think about the AI deployment your organisation is planning right now. Ask which side of that division it is being built for. The answer will tell you more about your coherence field than any culture survey.

When the Architecture Is Finally Right

Leadership changes character

When the company no longer depends on specific individuals to hold its continuity together, leadership can become what it was always supposed to be: the authorship of meaning. The work of deciding what matters and refusing to let it be traded away for the convenience of the present.

Scale stops destroying coherence

The reason growing companies fragment is not that people stop caring. It is that the volume of translations required to hold a large enterprise together exceeds what human systems can carry. When continuity is designed into the architecture, new people inherit the reasoning of those who came before — not just their outputs. New initiatives connect to what was built rather than restarting the story under a new name.

Artificial intelligence becomes what it should be

Not a productivity tool applied to fragmented work, accelerating outputs that have lost their connection to intent. A steward of continuity: the organisational memory that humans were never equipped to maintain at scale, holding the thread so humans can focus on what the thread is made of.

Decisions become predictable without becoming controlled

When actors across the organisation share a coherent picture of reality — what matters, what’s true, what was decided and why — decisions made at the edge become predictable to people at the centre. Not because they’re approved or constrained, but because they’re made from the same relational foundation.

When the architecture is right, coherence stops being a demand you place on people and becomes a property the system provides. Trust is no longer a cultural aspiration. It becomes structurally possible — built into the way meaning moves, not into the way people are asked to behave.

That is what Relational Intelligence makes real.

Topic 5

Designing the Space Between

Four principles for making meaning survive movement by design rather than heroism

The diagnosis is complete. This is the prescription.

Not another initiative. Not a values workshop that produces laminated posters and two days of goodwill before everyone returns to the same calendar. Something more fundamental: a different standard of continuity — one that makes meaning survive movement because the architecture carries it, not because someone heroic remembers to.

Relational System is an operating approach for making companies behave like a single, continuous system. It doesn’t redesign teams, tools, or org charts. It designs the space between them — the transfers, dependencies, and translations where reality breaks and meaning disappears. It treats that space not as a coordination problem but as the real product.

What follows is how that space gets designed.

The Space Between Is Where Work Actually Dies

The architecture of the typical company is designed entirely from the inside out: teams, tickets, tools, OKRs. The inside of each unit gets refined, optimised, measured. The transitions between them are left to chance and personality.

You’ve seen the handoff that looked clean and wasn’t. The transition meeting happened. The documentation was thorough. Everyone said they were aligned. And three months later the person who inherited the work was making decisions the previous owner would have recognised as wrong — not because they were careless, but because the crossing had transferred the outputs without the judgment that made those outputs intelligent. Nobody designed the space between. They designed the parts on either side of it and assumed the middle would take care of itself.

It doesn’t. It never does.

The cost lives in the structural reality that every transfer is a translation, and translation degrades. The context obvious to the person passing the work is invisible to the person receiving it. The intent clear at the origin arrives as instruction at the edge. The state honest when the decision was made has quietly become convenient by the time it reaches execution.

A Relational System doesn’t ask people to communicate better. It designs the crossings so meaning survives them.

Four Principles

Make intent durable

Intent is the first thing lost because it isn’t required for execution — work gets done without it, decisions get enforced long after anyone can explain them. When intent is durable, strategy stops being a speech and becomes a carried signal. The rationale survives the transfer. Teams don’t need to guess what matters. And when multiple teams can see each other’s intent clearly, contradictions surface before they become expensive.

Make context hard to strip

Context is what turns information into understanding. Most systems hollow it out automatically: tools compress reality into fields, urgency flattens nuance, transfers reduce history to “FYI.” A coherent company preserves the relational story around the work — what happened, what changed, what’s at stake, what must not drift. Not as documentation. As architecture. Context that must be manually preserved will be dropped the moment pressure is applied. Context that is structurally preserved survives.

Make state legible

State is where work actually is — what’s true right now, not what the roadmap claims. When it’s unclear, teams restart. They re-check, re-approve, re-ask. Legible state is coherence’s most practical feature: fewer status meetings, fewer defensive messages, fewer “just checking in” exchanges that are actually “I don’t trust what I was told.” It also makes cross-initiative incoherence visible before contradictions accumulate.

Give contradictions an owner

In incoherent systems, contradictions are free-range — wandering through the company, generating rework and quiet resentment while everyone waits for someone else to claim them. A coherent system makes contradictions assignable. Not as blame — as stewardship. Because ownership exists, candour becomes survivable. Disagreement can alter decisions rather than just being registered. Reality has somewhere to enter. That is how you reduce politics without motivational speeches — not by asking people to trust each other more, but by removing ambiguity as a shelter.

Where AI Fits, and Where It Stops

The division established in the previous section applies here with full force. The four principles above describe what must be maintained. The question is who maintains what.

AI is the steward of continuity. Not the author of meaning. It holds the threads between layers, surfacing when the translation from intent to execution has drifted. It identifies cross-initiative incoherence as it emerges, before it becomes globally destructive.

What it cannot do: decide what matters. Determine which trade-offs are acceptable. Judge which contradictions are worth preserving and which must be resolved. The moment AI starts making those calls, the coherence field stops being negotiated and starts being automated. An automated coherence field isn’t coherence. It’s insulation.

The posture isn’t control. It’s stewardship. You don’t deploy AI into a Relational System. You design a relationship with it — one where the machine carries continuity so the human can focus entirely on what the continuity is for.

The Promise a Coherent Company Makes

A coherent company makes one promise, even if it never says it aloud: that meaning will not be lost in transit — not between today and tomorrow, not between decision and outcome, not between what was intended and what actually lands.

This promise is not made in a values document. It is made in the design of the crossings — in the ownership of contradictions, in the legibility of state, in the structures that carry intent forward as a living signal rather than a historical record.

The organisations that make this promise and keep it don’t announce it. You feel it in how they move — the absence of the drag, the compounding of effort rather than the reset of it. They are not faster. They are cumulative. And cumulative, over time, is the only kind of advantage that compounds rather than erodes.

The principles exist. The thread is now in your hands — and what you do with it, inside your own organisation, is where this either becomes real or remains a conversation.

Conclusion

What You Cannot Unsee

The Only Question That Remains

You already know something is breaking.

Not loudly. Not all at once. In the accumulation of frictions that shouldn’t exist. The meetings that shouldn’t be necessary. The explanations that shouldn’t need repeating. The feeling that your smartest people are spending their intelligence compensating for a system that can’t hold its own logic.

You’ve built workarounds. Shadow docs. Redundant check-ins. Extra layers of review. You’ve hired for cultural fit and communication skills when what you actually needed was architecture.

There is something almost admirable about it — the sheer creative energy your organisation has spent working around a problem it never decided to solve. Shadow docs are a form of artistry. Redundant check-ins are a kind of folk wisdom. “Just to confirm” is the lingua franca of a system that doesn’t trust itself. None of it is laziness. All of it is intelligence applied to the wrong problem.

The question is not whether your people are capable of coherence. They clearly are. The question is why the architecture keeps making coherence their personal responsibility.

AI Is Already Making a Choice on Your Behalf

AI is scaling whatever system you currently have.

If that system is coherent — if meaning survives movement, if the continuity bundle travels intact, if the coherence field is real and permeable — AI becomes a genuine accelerant. It stewards memory. It surfaces drift before it accumulates. It frees humans to focus on judgment instead of reconstruction.

If that system is fragmented, AI becomes something else entirely: automated fragmentation. The same leaks, the same contradictions, the same quiet drift — now operating at machine speed with machine confidence.

You don’t get to opt out of this. You’re making this choice right now — in every structural decision, in every norm you reinforce or ignore, in every transfer you examine or leave unexamined.

Solving fragmentation with better communication, stronger values, and more alignment sessions is mopping faster while the pipe leaks. The real work is designing the space between — and seeing your architecture through one question: does this make meaning easier to preserve, or easier to lose?

One Question. One Standard.

You are building systems right now that will either compound intelligence or accumulate confusion. You are architecting relationships right now — between people, tools, teams, and machines — that will either hold a real coherence field together or produce an increasingly expensive simulation of one.

So here is the question, as plainly as it can be asked:

Does your system hold together, or does it merely move?

If you don’t know the answer, you’re already paying for it. If you know the answer and it’s wrong, you’re making a choice. And if you know the answer is right, then you understand what comes next: making that coherence intentional, durable, and real enough to be tested by the reality you’re actually operating in.

That’s the work that matters. That’s Relational Intelligence.

The thread continues — in The Coherence Standard: Book One — The Architecture.

What runs along it, in the end, is not just meaning. It is dignity.

About the Authors

Vecchietti & Loeffler

Elton Vecchietti and Sascha Loeffler have spent two decades working across organisational systems, experience design, product strategy, and human-AI collaboration — from startups to global enterprises in technology, mobility, healthcare, and service ecosystems.

Their shared focus has never been on tools or trends, but on a single question: why does meaning keep getting lost in transit, and what does it actually take to build a system that holds?

Elton Vecchietti

Product Design · Systems Thinking · Organisational Architecture

Eighteen years at the intersection of design, technology, and organisational systems — across three continents, from startups to global enterprises. He is also a musician: saxophone and voice. He lives between the beach and the city in Portugal.

Sascha Loeffler

Architecture · Service Design · User Experience Research

Trained as an architect before moving into organisational design. At Audi Business Innovation GmbH since 2018, focused on user experience research and service design. His architectural sensibility — designing for the use that follows — runs beneath every page of this manifesto. He gardens obsessively.

CoHere is a practice dedicated to helping organisations design for coherence — leaving behind not a solution but the ability to keep solving as conditions change. This manifesto is the first work in the Relational Intelligence Studies series, and the distillation of what they’ve learned about why that ability is so rare, and what it actually takes to build it.

This manifesto is a prelude to The Coherence Standard The full architecture. The cases. The toolkit. The standard.

The thread continues.

References

Sources & Acknowledgements

The concept of Coherence Debt is original to this work — the accumulated cost of failing to build the architecture that carries meaning. Its structural logic borrows the debt metaphor from Ward Cunningham, who coined Technical Debt in 1992 to describe the long-term cost of expedient software decisions. Cunningham, W. (1992). The WyCash Portfolio Management System. OOPSLA '92 Experience Report.

The Amazon recruiting tool case: Dastin, J. (2018, October 10). Amazon scraps secret AI recruiting tool that showed bias against women. Reuters. The authors' interpretation — that the lesson concerns coherence degradation at scale rather than bias specifically — is original.

The distinction between stated and practiced organisational culture echoes Chris Argyris's concept of espoused theory versus theory-in-use — the gap between what organisations claim to value and what their structures actually reward. Argyris, C., & Schön, D. A. (1978). Organizational Learning: A Theory of Action Perspective. Addison-Wesley.

The observation that organisational intelligence has migrated across successive substrates — people, systems, data, and now relationships — builds on a tradition of knowledge management scholarship including Jarche, H. (2014). The Seek > Sense > Share Framework; and Boyd, S. (2012). The Networked Organisation. The authors' extension of this migration into the relational layer is original to this work.

The CoherenceStandard