Haute Lumière · The reading

Most designs assume one kind of mind, and quietly serve only the people who have it

The people who leave are not confused; they are meeting something built for someone who makes meaning differently.

She is reading the part of the report that disagrees with her, which is the part she asked for.

THE BRILLIANT FAILURE

Three scenes, and the same wound in each. A team of unusually capable designers ships a meditation app. The interface is minimal and responsive, the onboarding is elegant, the sessions are recorded by a teacher people already trust, and the whole thing is backed by extensive user research and serious investment. Six months later the retention figure arrives: ninety-two percent of the people who download it are gone within three weeks. Nothing in the design review accounts for it, because there is nothing wrong with the design.

A progressive school district adopts a new social-emotional learning curriculum. It is research-based, carefully made, and piloted with teachers who wanted it. It asks students to reflect on their own values, question social norms, and construct personal ethical frameworks. In implementation something strange happens. The students who were already thriving go on thriving, and the students the curriculum was especially meant to reach disengage — some visibly frustrated, a few acting out in ways their teachers have never seen from them.

A chief executive redesigns a company around self-managing teams, distributed authority and a purpose-driven culture. The leadership reads the literature, attends the conference, and returns convinced. Within a year the organisation is in disarray. Not because the model is wrong — in many ways it is deeply right — but because a large part of the workforce experiences the new structure as bewildering, unsafe, and subtly contemptuous of the way they have always worked.

Three good designs. Three expensive failures. In each case the root cause is the same, and it is not one that surfaces in a usability test: a developmental mismatch between what the design assumed about the people using it and where those people actually are in the way they make meaning. The designers were not careless or unkind. They were designing, without noticing, for themselves.

This is not an exotic diagnosis. It accounts for a great deal of work that is well made and still does not land — the training everyone attended and nobody used, the policy that worked in one community and collapsed in the next with near-identical demographics, the feature half a company adopted with delight while the other half treated it as a threat. The pattern is consistent enough to be designed for. Once it has been named it is difficult to stop seeing.

Each of those teams tested their work against the one person almost nobody in the audience was: someone who already thought the way the designers thought.

WHAT ACTUALLY BROKE

A developmental mismatch occurs when a product, programme, policy or environment is designed for people at one level of meaning-making complexity and encountered by people at a different one. It is not a matter of intelligence. It is not education, age, income or cultural background, though all of those interact with it in complicated ways. It is about something more basic: how a person organises their experience of reality. Two people can hold identical facts, want the identical outcome, and still be running those facts through structures with almost nothing in common.

Jean Piaget established the principle with children, and it still does most of the work. A five-year-old who says the taller glass holds more juice than the wider one is not making an error born of missing information. That child is operating from a structure that cannot yet coordinate height and width simultaneously, and no amount of further explanation will produce understanding until the structure itself shifts. Children do not think like adults who happen to know less. They think in a qualitatively different way.

The larger claim of developmental psychology is that this does not stop at adolescence. Human beings do not only accumulate knowledge and skill across a life; they periodically reorganise the structure through which they make meaning at all. Horizontal growth adds content inside an existing structure — more facts, more skills, more experience poured into a container whose shape stays the same. Vertical growth changes the container. The person does not know more so much as know differently, and can now hold things that were previously not merely difficult but invisible.

Return to the meditation app with that in hand. Its onboarding asked users to observe their thoughts without attachment, a sentence that quietly requires a specific capacity: the ability to take one's own thinking as an object of awareness rather than being wholly inside it. For someone who has not yet developed that capacity, the instruction is not difficult. It is empty. They are being addressed in a language they have not learned to hear, and the app offers them no evidence that the failure belongs to anybody but themselves.

This is why the term names the design rather than the user. When a building's doorways are too narrow for a wheelchair, nobody concludes that the person failed. That conclusion was reached about buildings some decades ago and it changed architecture permanently. Stage-sensitive design applies the identical logic to cognitive and meaning-making access. The mismatch is a property of what was built, which means it belongs to the people who built it, which means it can be repaired.

THE MISSING AXIS

Contemporary design practice understands users better than it ever has. Personas, empathy interviews, ethnographic research, maps of the path a person takes through a product, accessibility standards, inclusive design principles — these are genuine advances, and nothing in this argument asks anyone to give them up. They have made a great deal of software and a great many services measurably more humane.

They also share a blind spot. Almost all of contemporary user research measures horizontal dimensions of difference: demographics, behaviours, preferences, pain points, contexts of use. All of it is important and all of it lives on a single plane — the plane of what people want, what they do, and what they experience. A persona template can grow indefinitely along that plane without ever leaving it.

Stage-sensitive design adds a vertical dimension: how people make meaning. This is not another row to add to the template. It is a different axis, one that cuts across every horizontal variable and reshapes how each of them is experienced. Two users can share a row for row match on the template and still be living in different worlds.

Consider two thirty-five-year-old professionals with similar incomes, similar education and similar stated preferences on a survey. If one makes meaning primarily through the expectations and norms of the groups they belong to, and the other through self-authored values and independent analysis, they will experience the same product, the same training programme and the same organisational structure in profoundly different ways. The first will look for clear guidance, social proof, and alignment with a trusted authority. The second will look for customisation, autonomy, and evidence-based reasoning. Neither is wrong, and a design that serves one beautifully may alienate the other.

The axis also explains a pattern most teams have already collected and filed under noise. User feedback that appears flatly contradictory — this is too rigid, this is too vague; tell me what to do, stop telling me what to do — is frequently not contradictory at all. It is two altitudes describing the same artefact accurately from where each of them stands. Read horizontally it looks like an unresolvable tie to be settled by whoever argues longest. Read vertically it is a specification, and it names both of the doors the design is missing.

This is not a niche concern at the edge of a user base. Research by Robert Kegan and others suggests that the majority of the adult population is in transition between or within the socialised and self-authoring orders of meaning-making, with a significant minority operating from earlier or later positions. Any product, programme or policy designed for adults is in fact designed for a developmentally diverse audience, whether or not the designer knows it. The only variable is whether the diversity was planned for.

The plant was here before the desk was. It sets the pace of the room, and the room has kept it.

THREE MINDS

Kegan extended developmental thinking into adulthood with a model describing five orders of mind — five qualitatively different ways a person organises their experience of self, other and world. For design work, three of them carry almost all the weight, and learning to recognise them is the single highest-return investment available to a practitioner.

The socialised mind is constituted by its relationships. A person here does not merely have relationships; in a real sense they are their relationships. Values, beliefs and identity are shaped by the groups they belong to, and what they seek from a system is alignment, harmony, and the approval of trusted authority. This is not dependence or weakness. It is a coherent architecture that does something no other architecture does as well: it holds a community together and carries forward what works.

The self-authoring mind has developed an internal seat of authority — a self-generated system of values and principles that can evaluate and choose among competing external expectations. This person can step back from their relationships and ask what they themselves actually think. What they seek is coherence, autonomy, and personal integrity, and what they contribute is the willingness to challenge a consensus that has stopped serving.

The self-transforming mind can step back from its own self-authored system and see it as one system among many, partial and historically situated. It is comfortable with paradox, holds several frameworks at once, and is often more interested in the space between perspectives than in defending any of them. Its gift is the capacity to hold tension and bridge difference without collapsing either side.

Now put one feature in front of all three. An open feedback tool is experienced as threatening by the socialised mind, whose sense of self is constituted by how others see them; as useful by the self-authoring mind, who can take feedback into a system they already own; and as interesting by the self-transforming mind, who welcomes what unsettles their framework. One feature, three experiences, and only one of them was designed for.

A human resources technology company learned this at full price. Their performance feedback tool let anyone give anyone anonymous real-time feedback, and it was adopted enthusiastically by about thirty percent of the workforce — the people already comfortable with self-evaluation and the ambiguity of unfiltered opinion. The remaining seventy percent found it threatening, or used it to reinforce social hierarchies rather than flatten them, or simply ignored a premise they found baffling. The redesign that eventually worked offered structured guided feedback, mentor-mediated feedback for people whose meaning-making runs through trusted relationships, and the original open format for those who thrived in it. The company arrived there after a year of declining engagement and a near-revolt from engineering.

THE OTHER MAPS

Kegan gives the architecture. The ego development line, mapped by Jane Loevinger and extended with considerable precision by Susanne Cook-Greuter, gives the texture — granular detail about how people at each stage relate to rules, authority, conflict, ambiguity and complexity. That level of detail is what you need when designing a specific feature rather than a general stance.

The model distinguishes, among others, a conformist stage that identifies strongly with its in-group and experiences deviation as threatening; a conscientious stage that has internalised self-evaluated standards and thinks in terms of long-term goals and consequences; and an autonomous stage that acknowledges inner conflict, tolerates ambiguity, and respects other people's need for autonomy. Put a mandatory compliance module in front of all three. The conformist participant finds the structure reassuring. The conscientious participant wants to understand the reasoning behind the rules. The autonomous participant may find the mandatory format condescending while still appreciating why a system needs one.

Spiral Dynamics, developed by Don Beck and Christopher Cowan from the work of Clare Graves, maps a third thing: value systems. Rather than tracking cognitive structure or ego development it describes the evolution of worldviews, colour-coded for reference — beige survival, purple tribal, red power, blue order, orange achievement, green communitarian, yellow integrative, turquoise holistic. Its particular usefulness is that it maps cultures and organisations as well as individuals. A design that works beautifully in an orange-dominant technology startup may fail in a blue-dominant government agency, not because it is bad but because it speaks to different values.

That model carries a qualification the book states plainly. Spiral Dynamics is more contested within academic psychology than the others, its empirical base is less robust, and it has been criticised for oversimplification, cultural bias and misuse in corporate settings. It earns its place on practical grounds for organisational and cultural work, and it should be held lightly. The colours are pointers, not people.

The honesty extends to the whole field, and a designer who skips this part will misuse everything above it. Developmental psychology does not establish that development is purely linear; the actual experience is spiral, recursive and messy, and people oscillate. It does not establish that later stages are better in any moral sense — history is well supplied with sophisticated thinkers who used their sophistication for cruelty. It does not establish that these sequences are universal across cultures, since the models were built and validated largely in Western, educated, industrialised contexts. And it does not establish that development can be reliably accelerated. Pushing someone toward a later stage before they are ready can do genuine psychological harm.

These are the best maps currently available, and they are maps. Hold them the way a navigator holds a chart: with respect for what it shows and readiness to adjust when the water disagrees.

CENTRE AND RANGE

The practical synthesis does not require choosing a favourite model. It requires three working concepts, and they are the whole toolkit: altitude, centre of gravity, and range.

Altitude is a general measure of the complexity of a person's or a group's meaning-making — a meta-descriptor rather than any single model's label. The word was chosen deliberately over level or tier, because at different altitudes you see different landscapes. Someone at the base of a mountain sees the meadow in intimate detail: the wildflowers, the insects, the texture of the soil. Someone at the summit sees the whole valley and the ranges beyond it, and has lost the soil entirely. Neither view is wrong, and a design that serves only high altitude — all abstraction, all systems thinking, all paradox — is exactly as insensitive as one that serves only low.

Centre of gravity is the altitude at which a person or group most characteristically operates. It is a home base rather than a fixed point; people move around it depending on context, stress and domain. For design purposes it is the most useful single data point you can hold, because it tells you where to build the core experience, from which scaffolding then runs in both directions.

Range is the spread of altitudes present in a given audience, and it is the variable that determines how much work is ahead of you. A doctoral seminar in developmental psychology is a narrow-range audience and needs very little adjustment. A public school parent community, or the workforce of a large organisation, is a wide-range audience where the distance between the lowest and highest altitudes in the room may be enormous. Assessing range is the first move, before any design decision is taken, because it sets the scope of everything that follows.

Three questions, in this order, cover the whole assessment. What is the likely centre of gravity of my primary audience, so I know where to build the core experience? What is the range, so I know how far the scaffolding has to reach? And what altitude am I designing from, so I can catch my own assumptions before they become the product?

The method that turns those questions into work is a range map, and it takes an afternoon. List the groups who actually encounter your design, including the informal ones — the people who love it, the people who struggle, the people who quietly leave. Estimate each group's centre of gravity in plain language rather than in stage labels: this group needs clear structure and guidance from a trusted source; this group wants autonomy and evidence. Draw a vertical line and mark them on it. Mark the zone your current design serves best. Everything on that line outside the marked zone is an underserved zone, and each one gets a single written hypothesis about what a person there might need. Those hypotheses are the design work. They are also, usually, the growth.

Half an hour in, she has stopped rehearsing what she thinks and started noticing how she is thinking it.

FIVE PRINCIPLES

The first principle is to design for the centre of gravity and scaffold toward the edge. Build the core experience so that it fits where most of your audience actually lives, then extend support in both directions rather than splitting the product into tiers. A design with no centre serves nobody well, and a design with a centre and no scaffolding serves only its centre. The two moves are one move.

The second is multiple entry points for multiple altitudes. The same underlying content can be entered by more than one door: a structured, guided, concrete pathway and an open, self-directed, contemplative one, both leading into the same material. This is not dumbing down on one side or gatekeeping on the other. It is the recognition that people arrive from different places and the front door cannot be in all of them at once, so you build more than one.

The third is respect without condescension, and it is the hardest of the five to hold under deadline. Simplifying for them is condescension wearing a helpful expression. Designing something that meets people where they are and serves them beautifully is respect, and the difference is visible in the artefact: condescension removes capability, respect changes its framing. If your accommodation for an earlier altitude would embarrass you to show the people it was built for, it is the wrong accommodation.

The fourth is invitation over imposition. Stage-sensitive design does not attempt to develop anyone. It creates environments that meet people where they are while offering gentle scaffolding toward the next edge — scaffolding that can be declined without penalty, without a locked feature, and without the interface implying that declining was a failure. Growth is always an invitation.

The fifth is that context shapes everything, including the body. Developmental position is not a fixed attribute a person carries unchanged through every room of their life. Stress, grief, unfamiliarity and threat all reliably shift meaning-making toward earlier and more concrete modes, and that shift is adaptive rather than pathological. The consequence for design is concrete: build for range within the individual, not only range across individuals. The same person who engages brilliantly with your autonomous features on a calm Tuesday may need the guided, concrete, relationally supported version of the same thing on a bad Friday, and should be able to reach it without asking anyone.

A stage-sensitive product is not a simpler product. It is a product with more than one way in and no penalty attached to any of them.

WHERE IT GOES WRONG

Knowing about developmental stages can make a designer worse, and the failure modes are predictable enough to name in advance. The first and most dangerous is developmental elitism. The designer learns the models and immediately begins sorting people, placing themselves at or near the top with remarkable consistency. Disagreement stops being disagreement and becomes evidence of a lower altitude. The theory turns into a sophisticated vocabulary for dismissal, and the resulting designs are quietly contemptuous of the people they claim to serve. The check is a question asked honestly and often: am I using this to understand people or to rank them?

The second is the myth of linear progress. Stage models are sequential, which invites the picture of a steady irreversible climb, and that picture is empirically false. Under stress, threat, trauma or unfamiliarity people reliably move toward earlier and more concrete modes of meaning-making. A self-authoring leader who receives devastating news may operate for a while from a socialised place, seeking belonging and the guidance of trusted others. That is health responding to overwhelm, not regression in any clinical sense, and a design that cannot accommodate it will lose people at exactly the moment they most needed it.

The third is stage labelling. Once the vocabulary is available the temptation to apply it to individuals is strong and it should be resisted. A person's centre of gravity is one dimension of an irreducibly complex human being, and reducing them to a stage label is like reducing a symphony to its key signature — accurate in one narrow respect and deeply misleading about everything else. The useful question is never what stage is this person. It is always how do I build this so that it serves people across the range.

The fourth is forced development. Some practitioners, newly excited, begin to see their role as developing their users — nudging, scaffolding, eventually pushing people toward greater complexity. A product that covertly pushes its users in that direction is not empowering them. It is manipulating them, and it is doing so in a domain where the person has not consented and the conditions of safety and support are unknown.

The fifth is ignoring cultural context. These models were built predominantly inside Western frames, and applying them without care risks mistaking cultural difference for developmental difference, or treating one culture's developmental endpoints as universal human ideals. The remedy is unglamorous: seek input from the communities you are designing for, and be genuinely willing to discover that your developmental assumptions do not apply and the design has to change.

Underneath all five sits one commitment that prevents them. Complexity is a capacity, not a rank. The person who makes meaning through loyalty, tradition and trusted authority contributes stability, continuity and the preservation of what works, without which institutions fall apart. The person who makes meaning through self-authored principle contributes innovation and the nerve to challenge what no longer serves. The person who holds paradox contributes the ability to bridge divides. A team composed entirely of self-transforming minds would be superb at seeing paradox and dreadful at shipping a project plan. The health of any system depends on all of them.

NOT LESS, DIFFERENT

A consulting firm built a three-hundred-and-sixty-degree leadership assessment that produced complex multi-variable reports: nuanced feedback, self-authored goal setting, systems-level organisational analysis. It was genuinely excellent work, and it was excellent for leaders operating at a self-authoring or self-transforming centre of gravity.

For leaders operating at a socialised centre of gravity — which, on Kegan's research, describes a substantial share of managers in most organisations — the same report was overwhelming. What those leaders needed was clearer evaluative framing, explicit guidance on what good looked like, and structured action steps tied to the expectations of their role. The sentence that should be printed above every stage-sensitive redesign is the one the firm eventually reached: they did not need less. They needed different.

The redesign created three lenses onto the same assessment data. A guided lens: structured, evaluative, action-oriented. A self-authored lens: reflective, self-directed, principle-based. An integrative lens: systemic, comfortable with paradox, multi-perspective. Each user chose a lens during onboarding and could change it at any time without explanation or penalty. One assessment, one data set, three ways in, and no one was told which one they belonged in.

That move generalises, and it is cheaper than it looks. The expensive thing in most products is the content and the data model, and neither changed here. What changed was the framing at the point of entry, which is a presentation layer decision, and the right to switch, which is a preference. A great many stage-sensitive redesigns turn out to be that shape: the same substance, entered from more than one direction, with movement between directions made free.

There is a discipline in how the choice gets offered, and it is easy to lose. A lens chosen at onboarding must never be described as a level, never carry a label implying more or less, and never hold anything behind it. The moment a chooser reads as a ranking it stops collecting honest answers and begins collecting aspiration, and the design has lost the single piece of information it asked for. Name the lenses by what they do rather than by who you imagine they are for.

The discipline that keeps it honest is testing across stages rather than testing across demographics. Recruit participants whose centres of gravity differ, not merely whose ages and job titles do. Watch specifically for the person who goes quiet, because disengagement rarely announces itself as a complaint and almost never arrives as a support ticket. Then build the feedback loop that lets the map be revised, because the first range map you draw will be an estimate and should be treated as one.

Disengagement is the most expensive signal in design, because it never files a ticket and never appears in the reasons people give for leaving.

DEVELOPMENTAL HOSPITALITY

Hospitality, in its deepest sense, is the art of making a stranger feel at home — not by pretending the stranger is the same as you, but by creating conditions in which their difference is welcomed and provided for. A good host does not serve only their own favourite food. A good host asks what would nourish you, and then makes room for you as you actually are.

Developmental hospitality is that stance extended to the meaning-making dimension of human difference. It asks designers to move past the unconscious assumption that their own way of making sense is the universal way, and to build products, programmes and environments that make room for the full range. It is an ethical commitment before it is a method, and the method follows from it rather than the other way round.

It is also, on the evidence, better work. A design that serves one altitude finds its own kind and loses the rest, and the loss is invisible in the metrics that most teams watch, because the people who leave rarely say why. A design built across a range keeps people who would otherwise have concluded, silently, that the thing was not made for them — and that conclusion, once reached, is almost never revisited.

Learning to see this does not make design easier. You can no longer assume your own perspective is universal, and you can no longer work comfortably inside the bubble of your own altitude. It changes how you sit in meetings, how you read user feedback, how you hear disagreement, and how you interpret resistance, which stops looking like obstruction and starts looking like information. It makes the work harder and it makes it better, which is the usual trade and usually worth taking.

The whole argument, worked through at length with the models, the principles, the pitfalls and the assessment methods, is Stage-Sensitive Design™, on the shelf at Haute Lumière. The reading is free and needs no account, and it is read aloud in full for anyone who would rather listen. Buying a volume is for keeping it — the file, the press edition, the copy that stays yours. Nothing here is being held back.

Free to read

Free to read, and free to hear. Every chapter of every book in this house, and every narration of it, is open to anybody. No account, no card, nothing to cancel.

What is in it

• Forewordopen this in the house search
• Preface: How to Read This Bookopen this in the house search
• Introduction: The Design Problem No One Talks Aboutopen this in the house search
• Chapter One: The Developmental Landscape — A Primer for Designersopen this in the house search

Incomprehension is not always a failure of explanation. Sometimes it is a difference in architecture.

Complexity is a capacity, not a virtue. It confers range, never goodness.

Growth offered is an invitation. Growth required is a toll.

A design built from one altitude finds its own kind and loses the rest without ever being told.

Keep looking

Every phrase on this page opens into the house search. The shelf holds The Developmental Canon and six other shelves, and the reading is free.