Plan-Execute Gap¶

The pathology in which a coarse-grained planning artefact cannot be enacted because the execution layer faces binding fine-grained constraints the planning model never represented — so the fix is constraint propagation and feedback, never a better forecast.

Core Idea¶

The plan-execute gap is the operations-management pathology in which a prescriptive planning artefact — a master production schedule, a sales-and-operations plan, a project Gantt, a logistics dispatch — cannot be enacted by the execution layer because the constraints binding at execution time were abstracted away when the plan was built. The plan was constructed at a coarser granularity (weekly or monthly buckets, aggregate capacity, idealised lead times) against a constraint model that omits the binding fine-grained realities the execution layer faces: hour-by-hour labour availability, dock-door scheduling windows, allergen-changeover sequencing rules, equipment fault states, instrument-tray readiness, picking-route congestion. Because the planning model does not represent those constraints, it generates plans that are infeasible under them.

The structural mechanism is a layered information asymmetry: the execution layer holds detailed, real-time constraint information that the planning layer does not model, and the planning layer holds coordination authority (the schedule is the directive) that the execution layer cannot override without improvising. When the execution layer cannot enact the plan as written, it reconciles in-place — resequencing, substituting, absorbing delays — without a channel to feed that reconciliation back into the planning model. The plan and operational reality diverge further over time, the plan loses authority as a control signal, and execution improvisation becomes structural rather than exceptional. The corrective is not better demand forecasting; it is constraint propagation (moving execution-layer constraints into the planning model), finer replanning cadence (daily rolling horizons rather than monthly master schedules), and plan-versus-actual telemetry that routes execution-layer improvisation back as learning into the next plan generation.

Structural Signature¶

Sig role-phrases:

the prescriptive plan — a master schedule, S&OP plan, Gantt, or dispatch intended to direct execution and holding coordination authority
the planning abstraction — the coarse-granularity model (weekly/monthly buckets, aggregate capacity, idealised lead times) the plan was generated from
the execution constraint set — the finer binding realities the execution layer faces (changeover sequencing, dock-door windows, tray readiness, fault states, route congestion) that the planning model omits
the layered information asymmetry — execution holds the detailed real-time constraint knowledge; planning holds the directive authority but cannot model those constraints
the infeasibility surface — the subset of the plan that cannot be enacted as written under the execution constraints
the improvisation response — the execution layer reconciling in place (resequencing, substituting, absorbing delays) to make the plan run
the feedback gap — the absence of a channel routing improvisation back into the planning model
the compounding divergence — plan and reality drifting further each cycle, the plan losing control authority, improvisation hardening from exceptional to structural

What It Is Not¶

Not a forecasting failure. The plan-execute gap is the case where the plan was a reasonable abstraction of the external world but the execution layer cannot enact it because constraints binding at execution time — changeover sequencing, tray readiness, dock-door windows, fault states — were never represented in the model. The binding realities are not misforecast demand or capacity; sharpening the forecast leaves the planning model just as blind. The discriminating test is whether a better forecast would have changed the outcome — if not, this is constraint-blindness, not plan-failure.
Not shop-floor indiscipline. Resequencing and substitution on the floor are not execution failing to follow the plan; they are a signal that the planning artefact has lost some control authority and that the execution layer sees binding constraints the plan does not. The volume and recurrence of improvisation measure how wide the gap is — information to route back, not behaviour to suppress.
Not generic model-vs-reality or map-vs-territory mismatch. Those cover descriptive representations diverging from their referents. The plan-execute gap is the specific case of a prescriptive artefact — a directive meant to be enacted — that cannot be carried out under execution constraints. The failure is in enactability, not in descriptive accuracy.
Not a Goodhart problem. It arises with no optimisation pressure on a metric at all — purely from abstraction mismatch between a coarse plan and a finer execution constraint set. Nothing is being gamed; the plan is simply blind by construction to realities it never modelled.
Not a one-off miss that self-corrects. Without a feedback channel routing improvisation back into the next plan, the divergence recurs by construction every cycle: the next plan is generated against the same blind model and reproduces the same gaps in new positions. The pathology is structural and compounding — improvisation hardens from exceptional to routine — not a single exceptional event.

Scope of Application¶

The plan-execute gap lives across the operations-management and prescriptive-planning subfields that share the plan/execution layering — a coarse directive handed to an execution layer holding finer binding constraints; the term genuinely travels between these, while on fundamentally different substrates the pattern dissolves into its parents (abstraction + local_versus_global_information + inter-layer feedback failure), which carry the cross-domain weight instead.

Manufacturing S&OP — a monthly sales-and-operations plan the production-control layer cannot sequence against changeover rules, raw-material arrival jitter, and preventive-maintenance windows.
Healthcare operations — surgical master schedules broken on the day by instrument-tray readiness, anaesthesia staffing, prior-case overrun, and surgeon-specific duration drift.
Construction project planning — a master schedule the field cannot honour against trade-sequencing and material-arrival realities, silently re-sequenced on site.
Software delivery — sprint plans broken by production incidents, dependency blocks, rework, and integration-environment outages the velocity assumption never modelled.
Emergency logistics — pre-disaster contingency plans overtaken by post-disaster road, fuel, and personnel realities, replaced wholesale by improvisation.

Clarity¶

Naming the plan-execute gap separates two failure modes that look identical in the finished outcome — the plan was not met — but demand opposite correctives. Plan failure means the plan was wrong about the external world: demand was misforecast, capacity overstated, lead times misjudged; the fix is better forecasting and a more accurate constraint model. The plan-execute gap is the other case: the plan was a reasonable abstraction of the external world, but the execution layer cannot enact it because the constraints that bind at execution time — changeover sequencing, instrument-tray readiness, dock-door windows, equipment fault state — were never represented in the planning model. Here, sharpening the demand forecast does nothing; the binding realities are not forecasting errors but constraints the plan was blind to by construction. The label tells the practitioner to stop tuning the forecast and instead push execution-layer constraint information up into the plan, shorten the replanning cadence, or reclassify the artefact from a hard schedule to a direction.

The concept also reframes what shop-floor improvisation means. Without the label, resequencing and substitution on the floor read as indiscipline — execution failing to follow the plan. Named, the same improvisation reads as a signal: the planning artefact has lost some of its authority as a control signal, and the gap it is papering over is information that needs to flow back into the next plan rather than be absorbed silently. The sharper question stops being "why didn't they follow the schedule?" and becomes "which binding constraints does the execution layer see that the planning model does not — and is there a channel carrying that knowledge back, or is the divergence compounding unobserved?"

Manages Complexity¶

The execution layer's binding constraints are an endless, substrate-specific list — changeover sequencing, dock-door windows, instrument-tray readiness, equipment fault states, picking-route congestion, anaesthesia staffing, integration-environment outages — and a planner who tried to reason about each as its own failure source would face a different catalogue in every facility and never finish. The plan-execute gap compresses that variety by asserting that what matters is not the identity of any particular constraint but its location relative to two layers: the constraint binds at execution time and was abstracted away when the plan was built. Every item on the list collapses to a single structural fact — a constraint visible to execution but absent from the planning model — and the diagnostic question becomes uniform across all of them: which binding realities does the execution layer see that the planning model does not, and is there a channel carrying that knowledge back? The planner stops enumerating constraint types and starts reasoning about the information geometry between the two layers.

What that planner then tracks reduces to a few quantities: the granularity gap between plan and execution (monthly buckets versus hour-by-hour reality), the replanning cadence relative to the rate at which execution constraints change, and the presence or absence of a feedback channel routing improvisation back into the next plan. From these the qualitative trajectory reads off without modeling any individual case. A coarse plan, a slow cadence, and no feedback channel guarantee compounding divergence: the execution layer reconciles in place, the plan loses authority as a control signal, and improvisation hardens from exceptional to structural. The branch structure that organizes the corrective is sharp and cuts against the instinctive fix. A missed plan splits first on cause — was the plan wrong about the external world (a forecasting error, fixed by a better demand model) or was it a reasonable abstraction the execution layer could not enact (a constraint-blindness error, where sharpening the forecast does nothing)? Only the second branch is the plan-execute gap, and within it the lever set is fixed: propagate execution constraints up into the planning model, shorten the replanning horizon, or reclassify the artefact from hard schedule to direction. Shop-floor improvisation, read through this frame, stops being indiscipline to suppress and becomes a measurable signal of how wide the gap is and where the missing constraint information lives. A bewildering, facility-by-facility sprawl of why-the-plan-failed collapses to a two-layer information asymmetry with a small parameter set, telling the practitioner which corrective family applies and that tuning the forecast — the reflexive move — is precisely the wrong one.

Abstract Reasoning¶

The plan-execute gap licenses reasoning moves that all turn on the information geometry between two layers — an execution layer that sees binding fine-grained constraints and a planning layer that holds coordination authority but cannot model them.

Diagnostic (infer the hidden cause from a surface signature): the defining move is to read shop-floor improvisation as evidence rather than indiscipline. When the execution layer resequences, substitutes, or absorbs delays to make a written plan run, the analyst infers from that improvisation to the existence of constraints the planning model did not represent — and reads the volume and recurrence of improvisation as a measure of how wide the gap is and where the missing constraint information lives. A second, prior diagnostic splits the failure at its root: a missed plan is read against its cause — was the plan wrong about the external world (demand misforecast, capacity overstated: a forecasting error) or was it a reasonable abstraction the execution layer could not enact (a constraint-blindness error)? Only the second is the plan-execute gap, and the discriminating test is sharp: if sharpening the demand forecast would not change the outcome because the binding realities are sequencing rules, fault states, and readiness windows rather than demand, the failure is constraint-blindness, not plan-failure. The analyst reasons from "which fix would have helped" back to which failure mode obtains.

Interventionist (name the change and its predicted effect): the corrective family is fixed once the gap is identified, and the predictions are specific. Propagating execution-layer constraints up into the planning model (finite-capacity scheduling, equipment-state visibility, actor-specific durations) is predicted to shrink the infeasibility surface at generation time — the plan stops prescribing what cannot be sequenced. Shortening the replanning cadence (daily rolling horizons rather than monthly master schedules) is predicted to keep the plan's constraint picture fresh relative to the rate at which execution constraints change, so divergence has less time to accumulate before the next plan absorbs it. Reclassifying the artefact from hard schedule to direction is predicted to convert improvisation from a control failure into sanctioned local reconciliation. The frame also makes a strong negative prediction that overrides the reflexive move: tuning the demand forecast will move nothing, because the binding constraints are not forecasting errors — they are constraints the plan was blind to by construction, and a better forecast leaves the planning model just as blind.

Boundary-drawing (when the move applies, which regime): the concept marks the regime in which divergence is guaranteed to compound — a coarse plan, a slow replanning cadence, and no feedback channel routing improvisation back into the next plan generation. In that regime the analyst predicts a one-way trajectory: the execution layer reconciles in place, the plan loses authority as a control signal, and improvisation hardens from exceptional to structural. The boundary also fixes where the concept does not apply: where the planning model already represents the execution-time constraints (or where execution faces no constraints finer than the plan's granularity), there is no gap — the artefact is enactable as written, and a miss must be a genuine plan-failure instead. The pathology requires a granularity gap and an unmodeled binding constraint to exist at all.

Order-of-events / feedback: the framing predicts the compounding sequence and locates the missing link in it. The plan is built coarse and early; execution meets the unmodeled constraint and improvises; the improvisation is absorbed silently with no channel back; the next plan is generated against the same blind model and reproduces the same gaps in new positions. Reasoning along that loop, the analyst predicts that without plan-versus-actual telemetry the divergence is not self-correcting — it recurs by construction every cycle — and that the leverage point is the return channel that turns each round of improvisation into a constraint the next plan can see.

Knowledge Transfer¶

Within operations management and supply-chain practice the plan-execute gap transfers as mechanism across its sub-fields, and the transfer is direct because those substrates share the load-bearing arrangement: a prescriptive planning artefact built at coarse granularity, handed to an execution layer that holds finer-grained binding constraints. The same diagnostic (split a missed plan into plan-failure versus constraint-blindness; read shop-floor improvisation as a measure of the gap), the same corrective family (propagate execution constraints up into the planning model, shorten the replanning cadence, reclassify schedule-to-direction), and the same toolkit (finite-capacity scheduling, rolling-horizon replanning, plan-versus-actual telemetry, digital-twin simulation) carry intact from manufacturing S&OP (a monthly plan that the production-control layer cannot sequence against changeovers, material jitter, and maintenance windows) to healthcare operations (surgical master schedules broken by instrument-tray readiness, anaesthesia staffing, and prior-case overrun) to construction project planning (a master schedule the field cannot honour and silently re-sequences) to software delivery (sprint plans broken by production incidents, dependency blocks, and integration outages) to emergency logistics (contingency plans overtaken by post-disaster road, fuel, and personnel realities). The negative prediction — tuning the demand forecast moves nothing, because the binding realities are sequencing rules and readiness states, not demand — ports with equal force. These are operations substrates differing in domain but not in the plan/execution layering, so the term genuinely travels between them.

Beyond operations the honest reading is shared abstract mechanism via a conjunction of parents (case B), and the boundary is worth stating precisely: the named concept does not transfer to fundamentally different substrates without dissolving into the primes it is built from. Stripped of operations vocabulary, the plan-execute gap is a recognisable conjunction — the abstraction mismatch is abstraction applied to a prescriptive artefact (not merely a descriptive model); the binding-constraint-only-visible-at-execution piece is local_versus_global_information (the execution layer holds constraint information the planning layer does not model); and the divergence-that-recurs-by-construction is feedback failure between layers (no channel routes improvisation back into the next plan). Wherever a plan-versus-execution divergence appears outside operations, it is recognisable precisely because those three substrate-neutral patterns recur there as genuine co-instances, not because "plan-execute gap" travels. So the cross-domain lesson should carry those parents — abstraction + local-versus-global information + inter-layer feedback failure — which transfer literally to any prescriptive-planning situation. What the operations term adds, and keeps home-bound, is the planning-artefact arrangement itself: its institutional role, its cadence, its formal hand-off from a planning function to an execution function, and the operations-calibrated interventions (finite-capacity scheduling, rolling horizons, digital twins) that are useful precisely because they are tuned to that arrangement. The cleanest disposition is the seed's: keep plan-execute gap as the operations-management instance, and let abstraction, local_versus_global_information, and feedback carry the weight when the pattern is needed on a different substrate.

Examples¶

Canonical¶

The defining instance in operations management is the classic Material Requirements Planning (MRP) schedule that the shop floor cannot run. Traditional MRP explodes a master production schedule into timed work orders using two idealising assumptions: fixed, constant lead times and effectively infinite capacity. Neither holds at execution time. A real work centre has finite machine hours, sequence-dependent changeovers (an allergen or colour change forbids certain orderings), tooling and maintenance windows, and material that arrives jittery. So MRP routinely releases a schedule requiring more capacity in a week than exists, or an order sequence the changeover rules forbid. Production controllers then expedite, resequence, and split lots by hand to make it run — and because the plan is regenerated next period against the same infinite-capacity model, the same infeasibilities reappear. The industry response, finite-capacity scheduling and advanced planning systems, exists precisely to push those constraints into the plan.

Mapped back: The MRP-generated work orders are the prescriptive plan, built on the planning abstraction of fixed lead times and infinite capacity. Finite machine hours and changeover rules are the execution constraint set the model omits, producing the infeasibility surface. The controllers' hand expediting is the improvisation response, and regeneration against the same blind model each period is the feedback gap driving the compounding divergence.

Applied / In Practice¶

In construction, the Last Planner System (developed by Glenn Ballard and Greg Howell and promoted through the Lean Construction Institute) is a named practice built to close this gap. Traditional construction runs off a master CPM schedule set months out, which the field silently re-sequences as trade dependencies, material deliveries, and site conditions bite. Last Planner inverts the flow: instead of pushing the master schedule down, the "last planners" — the foremen who will actually do the work — commit only to tasks whose constraints are verified ready, track the percent of commitments actually completed, and log the reasons for every miss. Those reason codes feed back to make-ready planning, so recurring execution constraints get surfaced and removed before they break the next week's plan. The master schedule is demoted from a hard directive to a direction, with feasible commitments made close to execution.

Mapped back: Last Planner attacks the layered information asymmetry by letting the foremen who hold the execution constraint set set the near-term commitments, and it closes the feedback gap directly through reason-coded miss tracking that routes constraint knowledge back into make-ready planning. Demoting the CPM master schedule to a direction is the reclassify-schedule-to-direction corrective, converting the improvisation response from silent drift into sanctioned, recorded reconciliation.

Structural Tensions¶

T1: Plan-failure versus constraint-blindness (one surface, opposite fixes, and a counterfactual hard to run). The concept's founding move splits a missed plan by cause: a plan wrong about the external world (misforecast demand, overstated capacity) demands a better forecast, while a reasonable abstraction the execution layer could not enact demands constraint propagation — and the discriminating test is whether sharpening the forecast would have changed the outcome. The split is load-bearing because the two branches call for opposite correctives, and confusing them wastes effort tuning a forecast that was never the problem. But the test is a counterfactual that is hard to run cleanly, and real misses are often mixed — a genuinely over-optimistic forecast and an unmodeled changeover rule both contributing — so attributing the whole miss to one branch can misdirect the corrective. Diagnostic: Would a sharper demand forecast have changed this outcome (plan-failure), or are the binding realities sequencing rules and readiness states no forecast touches (constraint-blindness) — and is the miss purely one branch or a blend?

T2: Coarse abstraction as coordination device versus fine detail as enactable (the price of propagating constraints up). The plan is coarse by design: weekly buckets, aggregate capacity, and idealised lead times are what let it be built early and coordinate across the whole system at once. The concept's headline corrective — push execution-layer constraints up into the planning model (finite-capacity scheduling, actor-specific durations, equipment-state visibility) — shrinks the infeasibility surface, but it does so by trading the abstraction's economy for enactability: a plan that represents every hour-by-hour reality grows heavy, slow to regenerate, brittle to change, and at the limit as complex as the reality it was meant to compress. Not every execution constraint belongs in the plan; some are better absorbed by local reconciliation than modeled centrally. The corrective is not free — it spends the coordination value coarseness bought. Diagnostic: Does pushing this particular constraint into the planning model improve feasibility enough to justify the added planning complexity and lost abstraction, or should it be left to execution-layer reconciliation?

T3: Improvisation as signal versus improvisation as drift (the reframe that does not settle what to do). The concept's valuable reframe reads shop-floor resequencing and substitution not as indiscipline but as a signal measuring the gap and locating the missing constraint information. That is right and it rescues execution from unearned blame. But improvisation is genuinely double-edged in a way the reframe alone does not resolve: the same act that keeps the operation running today — adaptive, often necessary — is also what silently erodes the plan's authority as a control signal and compounds divergence when it is absorbed without a feedback channel. Sanctioning improvisation (reclassify the schedule to a direction) risks legitimizing drift; capturing every instance as constraint learning risks normalizing a plan that should have held. The reframe tells you improvisation is informative without telling you, in a given case, whether to institutionalize or eliminate it. Diagnostic: Is this improvisation adaptive reconciliation to be captured and fed back as constraint learning, or drift quietly dissolving the plan's control authority that should instead trigger a replan?

T4: Replanning freshness versus commitment stability (the cadence lever's whipsaw). Shortening the replanning cadence — daily rolling horizons rather than monthly master schedules — keeps the plan's constraint picture fresh relative to how fast execution constraints change, so divergence has less time to accumulate before the next plan absorbs it. But a plan's coordinating power depends partly on holding still long enough for downstream actors, suppliers, and adjacent teams to commit to it; replanning too fast induces the classic nervousness of constant resequencing, whipsawing commitments and imposing coordination cost on everyone who acted on the prior plan. The cadence that minimizes divergence is not the cadence that maximizes committable stability, and the two pull against each other. Diagnostic: Is the replanning cadence keeping the plan feasible against changing constraints, or churning it fast enough that downstream actors can no longer commit to a stable directive?

T5: Autonomy versus reduction (the operations pathology versus its abstraction-and-feedback parents). Within operations management the plan-execute gap transfers as full mechanism across manufacturing S&OP, healthcare scheduling, construction, software delivery, and emergency logistics, because those substrates share the load-bearing arrangement — a coarse prescriptive artefact handed to an execution layer holding finer binding constraints — and the diagnostic, corrective family, and toolkit (finite-capacity scheduling, rolling horizons, plan-versus-actual telemetry) carry intact. But stripped of operations vocabulary the concept dissolves into a conjunction of parents: abstraction applied to a prescriptive artefact, local_versus_global_information (execution holds constraint knowledge planning does not model), and inter-layer feedback failure (no channel routes improvisation back). Wherever a plan-versus-execution divergence recurs outside operations, it is those three substrate-neutral patterns recurring as co-instances, not "plan-execute gap" travelling. Diagnostic: Resolve toward the parents (abstraction + local-versus-global information + inter-layer feedback failure) when the pattern is needed on a non-operations substrate; toward "plan-execute gap" when the operations planning-artefact arrangement — its cadence, its formal planning-to-execution hand-off, and the operations-calibrated interventions — is actually in play.

Structural–Framed Character¶

Plan-execute gap sits on the framed side of the spectrum — best read as framed-leaning: a normatively-charged operations pathology constituted by an institutional planning practice, though resting on a genuinely structural information-geometry skeleton. Three criteria point framed. Its evaluative weight is real: it is a named pathology, a failure mode — to call something a plan-execute gap is to diagnose it as having gone wrong, and the concept's whole payload is prescriptive (stop tuning the forecast; propagate constraints, shorten cadence, reclassify schedule-to-direction). It is human-practice-bound: the pattern exists only where a planning function hands a directive to an execution function — a planning/execution layering that is an institutional arrangement, not a fact of nature — so remove the organization and its plan-as-directive and there is nothing for the gap to be a gap in. Its institutional origin is likewise pronounced: master schedules, S&OP plans, Gantts, dispatch, the formal hand-off, the replanning cadence are all operations-institutional furniture, and vocab_travels fails for that layer (finite-capacity scheduling, rolling horizons, plan-versus-actual telemetry are home-bound). The structural-pulling mark is import_vs_recognize: outside operations the pattern is recognized as a genuine co-instance of its parents (not a mere analogy) wherever a prescriptive plan meets a finer execution constraint set — the abstraction-mismatch is real substrate-neutral structure — even though the operations term imports only by dissolving into those parents.

The portable structural skeleton is a single compound the entry names precisely: an information-geometry between two layers built from three substrate-neutral parents — abstraction applied to a prescriptive artefact, local_versus_global_information (execution holds constraint knowledge the planning layer does not model), and inter-layer feedback failure (no channel routes improvisation back). That conjunction is exactly what plan-execute gap instantiates from its parents, not what makes the operations term travel: the cross-domain reach belongs to abstraction + local-versus-global information + inter-layer feedback failure, while the planning-artefact arrangement, its cadence, and the operations-calibrated interventions stay home. Its character: an evaluatively-charged operations pathology, constituted by an institutional planning-to-execution hand-off and clad in operations-planning machinery, structural only in the abstraction-plus-information-asymmetry-plus-feedback-failure skeleton it composes from its parents — framed-leaning, not a prime.

Structural Core vs. Domain Accent¶

This section decides why the plan-execute gap is a domain-specific abstraction and not a prime — separating the portable information-geometry skeleton from the operations planning machinery that composes it.

What is skeletal (could lift toward a cross-domain prime). Strip the operations vocabulary and a thin relational structure survives, a conjunction of three substrate-neutral pieces: a directive is built from a coarse-grained model that abstracts away realities a lower layer actually faces, that lower layer holds binding information the upper layer does not represent, and there is no channel routing the lower layer's improvisation back — so the divergence recurs by construction. Each piece is a genuine prime the entry names: abstraction applied to a prescriptive artefact (a directive, not merely a descriptive model); local_versus_global_information (the execution layer holds constraint knowledge the planning layer does not model); and inter-layer feedback failure (no return channel, so the next plan is generated against the same blind model). This conjunction is genuinely substrate-portable — wherever a plan meets a finer execution reality outside operations, those three patterns recur as real co-instances — which is exactly why the entry treats them as the parents. That portable core is what the plan-execute gap shares, not what makes it the plan-execute gap.

What is domain-bound. Almost everything that makes the construct this pathology is operations-management furniture and none of it survives extraction: the master production schedule, S&OP plan, Gantt, and dispatch as concrete artefacts; the formal institutional hand-off from a planning function to an execution function; the replanning cadence (monthly buckets versus daily rolling horizons); the substrate-specific execution constraints (changeover sequencing, dock-door windows, instrument-tray readiness, finite machine hours); and the operations-calibrated interventions (finite-capacity scheduling, advanced planning systems, plan-versus-actual telemetry, digital twins, Last Planner reason-coding) that are useful precisely because they are tuned to that arrangement. These are the worked vocabulary, the instruments, and the empirical cases the discipline studies. The decisive test: remove the organizational planning-to-execution hand-off and its plan-as-directive and there is no gap for anything to be a gap in — what remains is the bare fact that a coarse abstraction can miss constraints a lower layer sees, which is the looser conjunction of parents, not this named pathology.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. The plan-execute gap's transfer is bimodal. Within operations it travels intact — the diagnostic (plan-failure versus constraint-blindness), the corrective family (propagate constraints up, shorten cadence, reclassify schedule-to-direction), and the toolkit carry from manufacturing S&OP to healthcare scheduling to construction to software delivery to emergency logistics, genuine recognition of one mechanism across substrates that share the plan/execution layering. Beyond operations the term does not travel: the pattern is recognizable on a new substrate only because its three parents recur there as co-instances, not because "plan-execute gap" crossed over — invoking the operations term off-substrate imports the planning-artefact-and-cadence machinery by analogy. And when the bare structural lesson is needed cross-domain — a prescriptive abstraction blind to a lower layer's binding constraints, with no feedback channel — it is already carried, in more general form, by abstraction + local_versus_global_information + inter-layer feedback failure. The cross-domain reach belongs to that conjunction of parents; "plan-execute gap," as named, carries the operations planning-artefact arrangement and its calibrated interventions, which should stay home.

Relationships to Other Abstractions¶

Current abstraction Plan-Execute Gap Domain-specific

Parents (2) — more general patterns this builds on

Plan-Execute Gap is a kind of Procedure-Work Mismatch Prime

Plan-execute gap is procedure-work mismatch specialized to a planning directive whose coarse model cannot be enacted under the fine constraints of actual operations.
Plan-Execute Gap presupposes Information Locality Prime

Plan-execute gap presupposes that execution holds criterion-relevant constraint information the planning level lacks.

Hierarchy paths (4) — routes to 4 parentless roots

Plan-Execute Gap → Procedure-Work Mismatch → Formal vs. Informal Structures → Formalization → Representation → Abstraction

Show alternative paths (3)

Not to Be Confused With¶

Plan-failure (forecasting error). The sibling failure mode where the plan was wrong about the external world — demand misforecast, capacity overstated, lead times misjudged — and the fix is a better demand model. The plan-execute gap is the opposite branch: the plan was a reasonable abstraction the execution layer could not enact because binding constraints were never modelled, and sharpening the forecast changes nothing. Both surface identically ("the plan was not met"), which is exactly why they are confused. Tell: would a sharper demand forecast have changed the outcome (plan-failure), or are the binding realities sequencing rules and readiness states no forecast touches (plan-execute gap)?
Knowing-doing gap (Pfeffer & Sutton). The organizational pathology in which a firm knows what to do but fails to act on that knowledge — a gap between insight and implementation driven by fear, inertia, or talk substituting for action. The plan-execute gap is not a failure of will or knowledge transfer but of representation: the planning model is blind by construction to constraints the execution layer sees, and execution does act (it improvises). Tell: is the organization failing to act on knowledge it holds (knowing-doing gap), or acting fine while the plan simply cannot be enacted as written (plan-execute gap)?
Principal-agent problem. The economics of a directive-holding principal and a better-informed agent whose incentives diverge, so the agent may shirk or misreport. Both share a layered information asymmetry, but the principal-agent problem is about misaligned interests; the plan-execute gap has no incentive conflict — the execution layer wants to enact the plan and cannot, and its improvisation is cooperative reconciliation, not self-interested deviation. Tell: does the lower layer deviate because its interests differ (principal-agent), or because the plan is infeasible under constraints the model omits despite aligned goals (plan-execute gap)?
Goodhart's law. The degradation that sets in when a measure becomes a target and optimization pressure games the metric. The plan-execute gap arises with no optimization pressure on any metric — purely from abstraction mismatch between a coarse plan and a finer constraint set; nothing is being gamed, the plan is simply blind. Tell: is a metric being gamed under target pressure (Goodhart), or is a directive infeasible because it never modelled the binding constraints (plan-execute gap)?
Map-territory / model-reality mismatch. The generic divergence of a descriptive representation from what it represents. The plan-execute gap is the specific case of a prescriptive artefact — a directive meant to be enacted — that cannot be carried out; the failure is in enactability, not descriptive accuracy. Part-versus-whole with a twist: it is abstraction applied to a directive rather than a description. Tell: is the artefact a description that is merely inaccurate (map-territory), or a directive that cannot be executed under the constraints it omits (plan-execute gap)?
The parent conjunction (abstraction + local-vs-global information + inter-layer feedback failure). The substrate-neutral trio the plan-execute gap composes and that actually travels beyond operations. The named pathology is their composition on the operations planning-artefact substrate; off that substrate the pattern is recognizable only because those three primes recur as co-instances. Treated more fully in the Knowledge Transfer and Structural Core vs. Domain Accent sections. Tell: strip the master schedule, cadence, and planning-to-execution hand-off and what remains — a coarse abstraction blind to a lower layer's constraints, with no return channel — is the parent conjunction, not the plan-execute gap.

Neighborhood in Abstraction Space¶

Plan-Execute Gap sits in a moderately populated region (48^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Unclustered & Miscellaneous (309 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12