The shape of a career

A career is a worker, inside a system, inside a market. All three produce the shape, and most identification mistakes come from mapping only one: knowing your mode but not your system, knowing your system but not how the market reshapes it, knowing your title but not your actual depth.

Five dimensions describe the worker. Five describe the system. Five describe the market.

Visual companion: View the diagram → (best on a wide screen)

The three modes

Three modes produce impact: Execution, Strategy, and Backing.

Execution produces deliverables. The output is the thing itself: the code, the model, the brief, the design, the operation. An Execution-primary week resolves into deliverables.

Strategy produces position decisions: what shape the work should take, what direction gets pursued or abandoned at the frame level. The output is a position that governs what other decisions operate inside. A Strategy-primary week resolves into resolved questions.

Backing produces resource commitments: which specific people, projects, or bets receive capital, headcount, or mandate. The output is weighted deployment within a frame already set. A Backing-primary week resolves into committed resources.

These aren’t a ladder. An engineer deepens Execution. A policy analyst deepens Strategy. A fund partner deepens Backing. Careers run for decades through deepening in one mode without switching primary. The mode you operate in is set by your domain and your role within it, not by hierarchy.

Most calendars mix all three. The question isn’t whether any week is pure but which texture dominates over a month or a quarter. Strategy and Backing especially overlap: Strategy sets positions, Backing deploys specifics within them, and in practice the two often happen in one decision. The frame holds when the week’s dominant mechanism is the one being named, not when the week is free of the others.

The five worker dimensions

The worker dimensions describe what the person does and accumulates. Two are positional: where the worker is. Three are accumulated: what the worker has built up by being there.

The positional dimensions:

Mode is the mechanism. Execution, Strategy, or Backing as the primary, with the other two stacked around it in varying proportions.

Mode also has two readings that can diverge. The assigned mode is what the role formally requires. The natural mode is where the worker’s sharpest judgment lives. When they agree, the role is using the worker well. When they diverge, the gap is structural information about whether the role is fitting the worker.

Domain is the field the mode is applied in: engineering, medicine, policy, venture, operations, law, education, finance, science. Domain determines what specific knowledge accumulates and what counts as expertise. Two people sharing a mode in different domains aren’t doing comparable work. Their judgment, networks, and reputations live in different spaces and don’t transfer cleanly.

The accumulated dimensions, collectively depth, are what builds inside a mode-domain-system-market combination over time. Depth grows with engagement and time, sharpens through feedback loops, and produces the difference between experienced practitioners and new ones. It resolves into three components that build separately and can diverge:

Insight is pattern recognition and judgment built through reps. It comes from feedback loops that have been tight enough, and stakes that have been real enough, to actually teach the worker. Insight often builds invisibly from the inside; it sharpens before any external signal shifts. Time-served and insight correlate but are not equivalent.

Network is who will act for the worker because they know them. It is activated relationship: bilateral, reciprocal, requiring prior contact. A network connection takes the call, shares the information, routes the opportunity, vouches when asked. Network builds through use and decays through neglect. A worker can have deep insight and a thin network, in which case real judgment exists but opportunity doesn’t flow to it.

Network has two facets that can move independently: strength, how activated and valuable the relationships are, and reach, how geographically distributed they are. Reach is determined by the structural source of the relationships. A relationship is place-anchored when it was formed through presence in a specific market, domain-anchored when it was formed through a field that spans markets, and institution-anchored when it was formed through a shared institution that produces an activated tie between people who passed through it. The same institution can do both at once, seeding an activated relationship and carrying a prior about the worker, and the two effects often overlap in practice even though they run through different dimensions. Any real network is a mix of the three sources, and the mix determines whether the network is concentrated or distributed.

Reputation is what people assume about the worker whether or not they know them. It is ambient belief: often unilateral, propagating through people who have never made contact, decaying through being disproven or superseded rather than through lack of maintenance. A respected institutional affiliation is one of the clearest carriers of it: the market holds a prior about someone from a recognized firm, program, or university before any contact and independent of any shared tie. This is the prestige effect, distinct from the activated institutional tie that seeds network. Reputation can run ahead of insight (the worker whose markers outpace their judgment) or behind it (the worker whose excellence is invisible to those who evaluate from outside).

The distinction between network and reputation is activated relationship versus ambient belief: who will act for you because they know you, against what people assume about you whether or not they do. The two correlate, and at the highest levels of accumulated depth they reinforce each other. The distinction is sharpest in the cases where they diverge: the well-connected operator without a reputation for judgment, the widely-respected practitioner with few activated relationships.

The three accumulated components usually move together over time. Strong insight tends to build network and reputation. They fail independently. The clearest diagnoses come from the cases where they diverge: insight without network, insight without reputation, network or reputation without insight.

The five worker dimensions are roughly orthogonal. Two people with the same mode can be in different domains. Two with the same mode and domain can have very different insight, network, and reputation, and very different balances among the three.

The worker dimensions travel with the worker. When the worker moves institutions or markets, mode, domain, and the accumulated dimensions come along, accumulating across the moves rather than resetting. They also evolve over time even when the role is stable: mode mix shifts as new modes get stacked onto the primary or as institutional growth changes what the role demands, and insight, network, and reputation accumulate with use. The worker-side location is a snapshot of dimensions that themselves keep moving.

The five system dimensions

The system dimensions describe the institutional structure the work happens inside. They are properties of the system, not the worker, and they shape the work as much as the worker dimensions do.

Scale is the size and structure of the vehicle. A seed fund and a buyout fund are both finance institutions, but the work is different enough that scale earns a separate axis. The same logic runs inside operating companies: a corp dev team at a megacap doing billion-dollar acquisitions is a vehicle of its own scale, separate from the parent company’s primary scale. Each scale shapes a different kind of depth: larger scale builds reach and institutional pattern recognition, smaller scale builds end-to-end judgment and proximity. Scale combines with governance, capital structure, and headcount to determine how much consequence the institution produces per person.

A civil servant works inside a tenure measured in administrations; a hedge-fund analyst works inside one measured in quarters. Time horizon is the cycle the work, and the worker’s tenure, operate in. It has two facets that usually move together but can diverge: the cadence of the work itself and the security of tenure. A consultant has fast work cadence with long firm tenure; a political appointee has moderate work cadence with tenure tied to an electoral clock. Time horizon shapes which decisions are reversible, what feedback arrives when, and what cycles drive resource flows.

Capital structure is how the work is financed and the worker compensated: fixed pay at a public agency or salaried firm, ownership stake at a startup or partnership, project funding at a research institution, retainer fees at a professional services firm, performance compensation at a fund or a sales-driven role. It sets who bears risk, who captures upside, and what incentive shape the work operates under.

Governance is how decisions get made: founder authority at an early-stage company, board oversight at a public corporation, partnership consensus at a firm, hierarchical bureaucracy at a government agency, peer review in academia, political accountability in elected office. It shapes which decisions are possible, how quickly, and with what autonomy.

Success metric is what the work is judged by: profit at a for-profit company, mission impact at a nonprofit, expert judgment and citation count in academia, electoral outcome in politics, capital returns at a fund, billings and reputation at a partnership firm, audience reach for a creative practitioner. It shapes what counts as visible progress and which contributions actually get rewarded. The official success metric and the operational one often diverge.

The five system dimensions are not fully independent. Capital structure and governance often travel together when capital sources carry governance rights. Capital structure and success metric are linked when compensation is metric-tied. Time horizon and capital structure correlate. Scale shapes what kind of governance and capital structure are even possible. The dimensions still earn their separate names. A fund and a partnership share governance structure but differ on time horizon; a civil servant and an academic share salaried capital structure but differ on success metric.

System dimensions also evolve over time. A startup has different system dimensions at seed than at Series C, with the same institution shifting scale, time horizon, capital structure, governance, and success metric across stages. Long-tenure workers in stable institutions experience similar drift as the institution itself changes around them: agencies shift across administrations, firms shift as partnerships grow, research institutions shift as funding sources change. The system the worker is inside today isn’t necessarily the system they entered.

The five market dimensions

The market dimensions describe the environment the worker-system pair operates inside. They are properties of neither the worker nor the specific institution, but of the broader ecosystem both sit within. The same worker in the same kind of system produces a different career in a different market.

Work density is how much of the relevant work exists in the market. Deal flow for an investor; case volume and complexity for a surgeon; the volume and type of disputes for a litigator; the density of fundable questions and collaborators for a researcher; the concentration of hard problems worth solving for an engineer; the number of live policy windows for a policy professional. Thin markets and deep markets compound differently for the same worker-system pair, regardless of profession.

Resource environment is what the work depends on, how abundant it is, and what the source expects in return. Capital from institutional investors versus development finance versus family offices, with different return and impact expectations; research funding from government versus foundations versus industry, with different strings; support for a creative from patrons versus institutions versus market sales. The funder shapes the work. A market dominated by one kind of resource source produces different dynamics than one dominated by another, regardless of how any single institution in it is structured.

Talent pool is how many people with relevant depth are in the market and how correlated their judgment is. A deep pool means stronger competition and stronger possible teams; a thin pool means the opposite. The correlation matters as much as the depth: a pool drawn from a narrow set of backgrounds tends to see the same things and miss the same things.

Conversion paths are how work converts into the outcomes the success metric measures, and how much each conversion can carry. How positions become returns for an investor; how research becomes recognition for an academic; how analysis becomes implemented policy for a policy professional; how work becomes audience and income for a creative; how built things become deployed and adopted for an engineer. Markets vary on both the channels and the per-unit ceiling: a market can have rich channels with high per-unit ceiling (capital markets, federal policy, mass media), rich channels with bounded per-unit ceiling (saturated metropolitan services), thin channels with high per-unit ceiling (federal regulatory work, appellate decisions), or thin channels with bounded ceiling. Conversion paths cap what compounding can turn into regardless of work quality, on both axes. Work density and conversion paths fail independently: a market can be dense with work but thin on conversion, which is the busy-but-trapped environment where a worker accumulates insight that never converts into measured success.

Institutional environment is what the rules and infrastructure permit: contract enforcement, regulatory clarity, institutional reliability, the infrastructure the work depends on. Weak institutions cap what builds for an investor, a builder, a clinician, and a researcher in the same structural way, by limiting what work is even possible to do well.

Market dimensions vary by geography and by ecosystem, and they shift over time the way system dimensions do. Capital availability evolves as new sources enter or exit, regulatory environments shift across political cycles, talent pools deepen or thin as the field attracts or repels people, and conversion paths change as the infrastructure of outcomes is rebuilt. A given institutional type carries different market dimensions in different places: the same fund structure, the same hospital model, the same research institute operates inside a different work density, resource environment, talent pool, conversion environment, and institutional environment depending on where it sits. Career models built in one market often assume that market’s dimensions and transfer poorly to another without adjustment.

Depth across markets

Depth is specific to the market it was built in, the way it is specific to mode, domain, and the system the worker operated inside. A worker moving markets is in the same structural position as a worker porting domains or shifting systems: part of the depth transfers and part resets, and the three accumulated components transfer at different rates.

Insight transfers most. The codifiable, structural pattern recognition (how to think about a business model, how to evaluate a team, the analytical frameworks, the mode-level judgment) carries across markets. A worker with deep insight in one market is not starting from zero in another.

Network transfers in proportion to how little of it was place-anchored. The place-anchored portion resets: the relationships that existed because the worker was present in the old market mostly do not reach into the new one, and they rebuild slowly through contact that has to actually happen. The domain-anchored and institution-anchored portions largely survive, because they were never tied to the old market’s geography in the first place. A worker whose network is mostly the people they knew because they were in a particular market loses most of it on a move; a worker whose network is the global community of their domain, or the cross-market reach of a shared institution, keeps much of it and can enter the new market through it.

Reputation transfers partially and unevenly. Because it is ambient belief rather than activated relationship, it can precede the worker into a market where they have no network at all. Travels through channels the new market can see. Reaches further into markets that give credit to the old one, less far into markets that do not. The same gradient runs at the system level: receiving institutions give credit to prior affiliations at varying rates, independent of the market they sit in. A reputation from one firm can land differently at two firms in the same market. Domain- and institution-anchored networks are also the channels reputation propagates through, which is why a worker with a globally distributed network often has reputation that arrives ahead of them: the network carries the belief into the new market. Reputation can arrive in a market the worker has never operated in. A place-anchored network cannot.

An institution is the clearest worked example of why the accumulated dimensions transfer at different rates, because one institution touches a career through three distinct channels. It shapes insight: what the worker actually learned and the judgment they built there, which is theirs and travels with them. It seeds network: the people they came through it with who will act for them because of the shared tie, which travels only where those people are present in the new market and the tie still activates. And it carries reputation: what the market assumes because of the affiliation, independent of anyone the worker knows or anything they specifically learned, which travels wherever the institution is recognized. The same institution, three channels, three transfer rates. A worker leaving a respected firm for a new market keeps the bulk of the insight, keeps the shared-institution network only where former colleagues are present, and keeps the institutional reputation wherever the new market gives credit to the firm, the way credentials transfer across institutions only where they are recognized. This is often the widest-reaching of the three, and the one requiring mostly that the affiliation be visible.

The talent-pool correlation has a structural consequence here. A worker whose insight was built in a different market brings pattern recognition that is uncorrelated with the local pool’s. That is an edge where the local consensus is systematically wrong and a liability where the local consensus encodes real local knowledge the outsider hasn’t built.

Borrowed and owned depth

The three accumulated dimensions can be held in two forms. In borrowed form, an institution is the conduit: it lends reputation, seeds network, and shapes insight, and the worker draws on these while inside it. The same arrangement viewed from the worker’s side is borrowed form: depth that is available but not the worker’s to keep. In owned form, the worker holds a platform: a vehicle they control that reaches a market or audience directly. A named fund, a firm they founded, a body of public work, a following. The same insight, network, and reputation exist in both forms; what differs is whether they route through an institution or through something the worker owns.

The difference shows on departure. Borrowed depth is borrowed and degrades when the worker leaves: the institutional reputation fades over years, the relationships that were the institution’s stay with it, the insight remains but loses the mandate that made it actionable. Owned depth persists across moves because the channel is the worker’s. A platform is the mechanism by which borrowed depth gets converted into owned depth, and it is often how a network becomes domain-anchored rather than place-anchored, since a platform reaches across markets where a seat does not.

What a platform uniquely adds, beyond self-carried reputation and a distributed network, is an independent conversion channel: a way to turn insight into action and outcome without institutional permission. This is the worker-side analogue of the market’s conversion paths. A worker can have strong owned reputation and a wide network and still lack a vehicle that lets them act independently; the platform is that vehicle.

Ownership relocates constraint rather than removing it. A borrowed platform constrains through the institution: its mandate, its compliance function, its brand-risk tolerance, its governance determine what the worker may do. An owned platform removes the institution as gatekeeper but exposes the worker directly to what the institution previously absorbed: regulation lands on the worker, reputational consequences are not diffused by an institutional byline, fiduciary and disclosure burdens are borne personally. Every platform, owned or borrowed, has constraints. The owned platform trades institutional permission-seeking for direct exposure to the system and market dimensions the institution used to absorb.

What gets exposed isn’t only regulatory and reputational. Owned form also exposes the worker to the requirement to articulate what problem the work solves and what impact it produces, because there is no institution above doing that articulation on the worker’s behalf. Owned form compounds when the owner can name the problem precisely, often more precisely than the customer does, and the impact in operational terms: what changes for whom, by how much, on what timescale. Owned form fails when the articulation isn’t there, because there is no institutional layer to absorb the gap. The articulation work is part of what the institution was doing in borrowed form, made directly load-bearing in owned form.

Neither form is better. Owned depth is portable, persists across moves, and makes insight independently actionable, but it is slower to build, offers no institutional reputation to draw on while building it, and carries regulatory and reputational exposure directly. Borrowed depth is available immediately, carries the institution’s reputation and conversion channel from the first day, and offloads compliance and risk onto the institution, but it is borrowed, bounded by the institution’s mandate, and not the worker’s to keep. Which trade fits depends on the rest of the structural picture.

Form is also a question of vantage. From the worker’s own side, an owned platform is owned form: a vehicle they control, a channel that reaches market directly. From the side of customers, audience, or anyone who relates to the work through the platform, the same platform is borrowed form: they experience an institution they’re contracting with, not the worker directly. The owner is operating as both at once. This is part of why Reputation reads differently from inside and outside owned form. From inside, reputation is built through accumulated work and visible relationships. From outside, reputation appears as a property of the platform itself, partly detached from the person who built it. The two readings can diverge, and the divergence is part of what the owner has to read accurately to operate.

Problem and impact

Every job of work has two structural properties: a problem it operationally solves and an impact it produces. These exist whether or not the worker has named them. A payment-processing engineer is solving the operational problem of making transactions complete reliably under load; the impact is the volume and value of transactions that go through without error. These structural facts are true even if the engineer would describe their job as “writing code for the payments team.”

Workers vary in how clearly they have surfaced these properties to themselves. Some can name precisely what problem the work addresses and what changes because of it. Others know their tasks and metrics but not what the tasks are operationally for. The variation is largely independent of capability and seniority: highly capable workers can compound for years inside an institution that does the problem-and-impact articulation on their behalf without ever having to surface it personally.

The articulation matters because it is upstream of using the framework. A worker who has named the problem reads domain more accurately. One who has named the impact can tell whether the success metric tracks the operational job. A worker who can name both can better identify whether a move would carry the same orientation into a different system or land them somewhere structurally different despite a familiar-looking title.

Worker dependence

Outcomes vary in how much they depend on the specific worker producing them and the dependence has two distinct aspects.

The first is operational continuity: whether the system can continue producing outcomes without the specific person. A worker can be load-bearing for present-tense outcomes (the system collapses or degrades without them) or not (a substitute would produce comparable results). This depends on whether the work is codifiable, whether judgment-heavy decisions have been pushed into standing systems, and whether the institution around the work is designed to absorb the worker’s absence.

The second is unique contribution: whether the outputs that exist carry a signature that no substitute would have produced. Some work product is structurally over-determined: different workers would have produced comparable outputs. Other work product carries judgment, taste, or relationship-shaped decisions that wouldn’t have emerged from a substitute. A worker can be present-tense substitutable (the system runs without them now) while having shaped what the system became in ways no substitute would have shaped it.

The two aspects move independently. A solo consultant whose work depends entirely on their daily presence is high on operational continuity. A SaaS founder whose product runs without them is low on operational continuity but may be very high on unique contribution: the product exists in its specific form because of choices only they would have made. A senior engineer doing competent but substitutable work in a well-structured team is low on both. A founder who built systems that now run without them but whose original architectural choices shape every future outcome is the temporally interesting case: high unique contribution measured backwards, low operational continuity measured forwards.

Both aspects are largely independent of role, form, and seniority. Building a system that doesn’t depend on you is often the right structural move, especially in owned form, and reducing operational continuity is frequently the explicit goal. Unique contribution is harder to engineer: it tracks whether the work hinges on something specific to the person that wouldn’t have emerged otherwise.

Operational infrastructure

Operational infrastructure is what the work runs on: the codified tools, systems, playbooks, frameworks, data, and rules that make a given piece of work possible to do at all, and faster to do well. It exists at all three levels of the framework, and a worker’s actual capacity to produce outcomes is shaped by what infrastructure they sit on top of as much as by the depth they bring.

On the worker side, operational infrastructure is what the worker has personally built and carries across moves: codified playbooks, templates, scripts, personal automation, the artifacts that turn pattern recognition into a tool that runs without the worker having to think it through each time. A researcher’s analysis pipelines, a consultant’s library of templates, a clinician’s decision aids built from their own pattern recognition all sit here. This is the codified form of insight. A worker who has done the work to externalize their judgment into reusable artifacts moves through new problems faster than a worker with comparable insight that has never been codified.

On the system side, operational infrastructure is what the institution has accumulated and provides to the workers inside it: proprietary internal systems, frameworks built up over years, internal documentation, access to bespoke data sources, the tooling that exists because someone built it for the institution and the institution kept it. Bloomberg terminals inside a bank, the case-management systems a hospital runs on, internal compliance frameworks at a firm, the bespoke datasets a research lab has accumulated, the proprietary tooling a tech company has built. Institutions accumulate this over time, and it is one of the clearest ways the institution shapes what the work even is. The same worker inside an institution with rich operational infrastructure produces work that the same worker outside, or inside a thinner one, would struggle to produce.

On the market side, operational infrastructure is the public layer the work depends on: open standards, shared data resources, regulatory frameworks, open-source ecosystems, public infrastructure of all kinds. GAAP for accountants, FAA regulations for aviation, open-source toolchains for developers, public health surveillance for epidemiologists, building codes for architects. A market with strong public infrastructure permits work that a thin-infrastructure market does not. The same worker-system pair operates inside a different feasibility set depending on what the market provides at the public layer.

Operational infrastructure is often the multiplier that determines what work is possible at all. A worker with insight but no codified infrastructure has to do the work fresh each time. A worker inside an institution with rich operational infrastructure can move much faster than the same worker outside. The infrastructure compounds with depth: insight built without codification stays in the worker’s head and travels person by person, while insight that has been codified into operational infrastructure travels through the artifact itself.

Operational infrastructure also raises a capture question that is structurally distinct from the borrowed and owned distinction. Much of what workers build inside institutions becomes institutional property by default. Worker-built tooling, frameworks, and codified playbooks typically become firm IP, even when the codified artifact is the worker’s own pattern recognition. The worker keeps the insight that produced it and the meta-skill of producing it, but the specific artifact stays. This is a third structural state, distinct from purely borrowed form (the institution provides the infrastructure) and purely owned form (the worker built and keeps the infrastructure): worker-built capability that is institutionally captured. Owned form is the form that escapes capture. The implication for compounding is that workers can spend significant time building durable artifacts that compound for the institution but not for them personally, and recognizing this is part of reading what is actually accumulating across the career.

Leverage across the three sides

Leverage is the per-unit consequence ratio of the work, and it operates at three levels at once. Each side of the framework contributes its own facet through the dimensions already named.

On the worker side, leverage is what Insight, Network, and Reputation produce in combination. On the system side, leverage is what Scale, Governance, Capital structure, and Success metric structure in combination. On the market side, leverage is what Conversion paths and Institutional environment permit in combination.

The three are not additive but compositional. A worker’s actual leverage is bounded by the lowest of the three: a worker with high accumulated depth in a structurally diffuse institution exerts less consequence than the worker side alone would suggest, and a high-consequence institution in a low-leverage market hits a ceiling the system would otherwise clear. The binding side is the structural diagnosis.

The borrowed and owned distinction applies to leverage as well as depth: a worker can route consequence through institutional and market position, or originate consequence from their own accumulated depth.

Each side of leverage has its own portability. Worker-side leverage is built and travels with the person: insight, network, and reputation come along on a move. System-side leverage is position-granted and stays with the seat: scale, governance, capital structure, success metric structure, and the operational infrastructure that the institution provides. Market-side leverage is geography-granted and stays with the ecosystem: conversion paths and institutional environment that the worker-system pair was operating inside.

How worker, system, and market combine

The shape of a career is not the worker dimensions alone, nor the worker inside a system alone. It is the worker, inside a system, inside a market. The same worker profile produces fundamentally different careers depending on the system and the market around it.

A Backing-primary worker in finance, with deep insight, a strong network, and solid reputation, inside a partnership system at fund scale, in a deep market with abundant work and high leverage, is a fund partner in a mature capital market. The same worker, inside the same system type, in a thin market (sparse work density, a resource environment dominated by a single source with mixed expectations, a small and correlated talent pool, few conversion paths, and lower leverage per deployment) does work that is recognizably Backing but operates inside a different career: different sourcing, longer holds, different exits, different compounding. Same worker, same system type, different market, different career.

A Strategy-primary worker in policy, deep, inside a civil service system at government scale, operates differently in a market with strong institutional infrastructure and many implementation pathways than in one where institutions are weak and analysis rarely converts into enacted policy. The mode and domain are the same; the market changes what the work can compound into.

Careers with the same worker dimensions can look very different across systems and across markets. Neither system nor market is background.

The interaction also runs in the other direction. Some workers fit some systems and markets; others don’t. A Backing-primary worker who needs short feedback loops compounds in a deep, fast market and stalls in a thin, slow one. Worker-system fit and worker-market fit are both real variables, and a mismatch on either can produce friction even when the worker, the system, and the market are each individually strong.

Each side has its own mobility profile. Worker dimensions travel with the person: when the worker moves institutions or markets, mode, domain, insight, network, and reputation come along. System dimensions are properties of the institution: they remain with the institution when the worker leaves, and they shift over time as the institution evolves, though workers with sufficient leverage can shift system dimensions from inside. Market dimensions are properties of the ecosystem: they remain with the geography when the worker-system pair leaves, and they shift over time as the market evolves, though institutions with sufficient market position can shift market dimensions from inside.

Drift since entry is its own fact worth surfacing. The system the worker entered isn’t necessarily the one they’re inside today. The market the worker-system pair entered isn’t necessarily the one they’re operating in now. Re-reading the dimensions against current conditions is a separate identification from the original locating, and the gap between the original and the current is itself information about what has compounded and what has eroded.

How careers take shape across the dimensions

Three patterns describe how careers take shape, beyond what the dimensions are individually.

Concentration and diversification. Concentration and diversification aren’t a binary choice. They operate at different levels of resolution, and most compounding careers are concentrated at one level and diversified at another. A cardiologist concentrated on heart disease is still diversified across many patients, conditions, and interventions within that specialty. A venture generalist concentrated on early-stage investing is still diversified across sectors. An engineer concentrated in one technical stack is still diversified across problem shapes within it.

Concentration builds depth: specific knowledge, pattern recognition within a bounded space, mastery of craft. Diversification builds breadth: cross-context pattern recognition, meta-level patterns that transfer, connections that specialists miss. Both are real forms of compounding. The structural question is which level is concentrated and which is diversified, not whether to do one or the other. Careers concentrating at every level become brittle; careers diversifying at every level accumulate breadth without depth. Network reach is one instance of this: a place-anchored network is concentrated in one market and a domain- or institution-anchored one is diversified across many, the latter slower to build but more resistant to a market move.

Clock. What you do has its own natural rhythm, and what you do it inside slows it down or speeds it up. Each mode has its own timeline: Execution returns arrive within a quarter, Strategy rarely shows full returns inside two years, Backing depends on the horizon of the backed (hires need months to reach full context, portfolio companies years, institutional capital across cycles).

The mode clock is filtered through the system time horizon and the market. A quarter inside a public company means something different than a quarter inside a foundation. A two-year Strategy bet is short inside a research institution and long inside a quarterly-reporting public company. The mode’s natural timeline interacts with the system’s cycle and the market’s pace to determine what feedback actually arrives, and when.

Apparent and actual depth. Titles, brand of platform, and speed of promotion are the visible markers of a career. They track reputation more reliably than they track insight. Reputation, network, and insight correlate but not perfectly, and the visible markers see reputation most clearly, network partially, and insight least of all. The gap is usually visible to people who do the work and invisible to people who evaluate it from outside. Deep insight at small scale, a strong network without visible reputation, and accumulated practical knowledge that doesn’t show up cleanly on a resume are all forms of depth the markers miss. Identifying actual depth, rather than visible signals of it, is the harder half of locating accurately.

Locating yourself

The framework is most useful when you apply it to a specific career.

For mode, the calendar is the fastest test. Look at last month. What did each week end in? Deliverables (Execution), position decisions (Strategy), or resource commitments (Backing)? Most calendars mix; the question is which texture dominates.

A second reading sharpens the answer. Calendar texture is what you actually did. Sharpest judgment is what you do best. If both agree, your primary mode is clear. If they diverge (calendar texture is one mode but the sharpest judgments come from another), the gap is structural information. The divergence often indicates an assigned mode that doesn’t match the natural one.

For domain, identify the field your mode operates in and at what level of specificity. A surgeon is in medicine and more specifically in cardiac surgery; a fund partner is in finance and more specifically in early-stage software; a policy analyst is in public policy and more specifically in healthcare regulation. Domain has levels, and locating across them tells you where specific knowledge actually accumulates.

For the accumulated dimensions, separate the three. For insight, the structural test is whether you’ve crossed from rule-following to pattern recognition: at low insight, you apply principles; at higher insight, you see patterns and your intuitions about specific cases go beyond what the principles would predict. For network, the test is who would act for you because they know you (take the call, share the information, route the opportunity, vouch when asked) and whether those relationships are place-anchored or would survive a move to another market. For reputation, the test is what people who have never worked with you assume about you before any contact. The three can be at very different levels; identifying which is ahead and which is behind is more informative than a single judgment of depth.

For the system dimensions, ask:

• What’s the scale of the vehicle? Scale is the size of the unit, function, or business inside the organization, not necessarily the size of the parent. A corp dev leader at a large company is in a Backing vehicle at acquisition scale, not at the parent’s full operational scale.
• What time horizon do the work and your tenure operate in, and do work cadence and tenure security agree or diverge?
• What’s the capital structure? Salary, equity, partnership, grants, performance compensation, fees?
• How are decisions made? Single authority, board, partnership consensus, hierarchy, peer review, political process?
• What’s the work judged by? Profit, returns, mission impact, peer recognition, electoral outcome, billings, reach?

For the market dimensions, ask:

• How much relevant work exists in this market? Deep or thin?
• What does the work depend on, how abundant is it, and what does the source expect?
• How large and how correlated is the talent pool?
• How does work convert into outcomes here, how many paths exist, and how much does each conversion carry?
• What do the rules and infrastructure permit or prevent?

The result is a structural picture of position and what’s compounding.

Why you’re where you are

Identification includes how you arrived at the location, because the path determines whether the location was designed or inherited.

Five patterns produce careers, and most careers contain more than one.

Designed. You examined alternatives and selected the mode, domain, system, and market because they suited you. You have continued to choose the location actively as conditions have changed. A designed career has been tested against alternatives, and the location is the chosen answer.

Optionality-seeking. You actively selected a path that preserved future options without committing to a specific direction. Some paths are structured to do exactly this, rewarding deferral itself with status and compensation that build while you keep multiple destinations available. The pattern resolves either into deliberate commitment to one of the destinations the optionality preserved, into the holding pattern becoming the destination by default, or into the optionality eroding before you commit.

Inherited. You arrived at the location through some mechanism that wasn’t active selection: a path-dependent start that branched without examination, social or external pressure that channeled you into a high-prestige or expected direction, or accidental compounding where you stumbled into something and discovered you were good at it. Inherited careers can land in locations that suit you, but the suitability is unconfirmed because alternatives were never tested against.

Constrained. You selected the path that best satisfied financial, geographic, family, or immigration constraints at the time. The career is a constrained optimization. The path that was the best available solution under those constraints may persist after the constraints lift, at which point the question of whether the location still fits becomes a separate identification.

Held past relevance. You entered a location that fit at one phase and stayed past when the fit shifted. Drift, sunk cost, and inertia all produce this pattern. The location was right once; whether it still is requires a separate read against current conditions.

These patterns aren’t mutually exclusive. Most careers contain several: an inherited start, a constrained middle phase, a designed re-commitment after the constraints lifted, and held-past-relevance fragments in dimensions that haven’t been re-examined. Designed names both an arrival and an ongoing posture, which is why it can overlay any of the other four: the arrival pattern doesn’t lock the posture.

The path that produced the location doesn’t determine whether the location is right. Designed paths can produce mismatch. Optionality-seeking can resolve into the right destination by accident. Inherited paths can produce alignment. Constrained paths can outlast their constraints into genuine fit. Locations held past relevance can still be redeemable if examination is done. Whether the location is right is a separate identification from how you got there.

What the framework reveals

The framework names fifteen dimensions across three sides: five worker, five system, five market. Five cross-cutting concepts run through them: borrowed and owned form, problem and impact, worker dependence, operational infrastructure, and leverage. What it surfaces is the interactions between them: across sides, within sides, and over time. The shape lives in the interactions. The dimensions exist to make the interactions visible.

For a visual map of the fifteen dimensions and how they interact, see the diagram.

Reference card

(Mode and domain are positional. Insight, network, and reputation are accumulated, collectively depth.)