OCE.7 - Coordinate Product-or-Service and Organization Architecture Decisions
Type: Method pattern Status: Eternal alpha
Primary working result: coordinated but separately governed product-or-service and organization architecture decisions that name both holons and selected structures, correspondence pressure, four candidate forms, expected gains and losses, authority, evolution window, realization returns, and observations that reopen either decision.
OCE.7:0 - Use This When
Use this pattern when a product or service architecture and an organization design constrain each other strongly enough that deciding one side alone would create avoidable coordination, evidence, provider, safety, or evolution burden. Enter when “the organization should mirror the product”, “teams own services end to end”, or “change the platform to fit the organization” is being used as the decision.
Begin with one bounded contribution and at least one organization-side structure or candidate from OCE.3 or OCE.4. Recover the product-or-service focus, concept, architecture claims, and qualified specialist contributions from Systems Engineering or another owning practice when available. State missing inputs and what they block.
The first useful result can be a bounded mismatch: two separately governed decisions that say which structures will align, which will remain deliberately non-isomorphic, what burden is accepted, and what observation will reopen that choice.
Use C.32.CONWAY directly when only correspondence candidate synthesis is needed. Use C.32.PAD or the owning domain pattern for each architecture decision. Use OCE.4 when only organization contribution structure is changing, Systems Engineering when only the engineered product architecture is current, and Operations when the question concerns managing continuing Work rather than changing the organization.
OCE.7:0.1 - Working Distinctions
| Name used here | Meaning |
|---|---|
| organization-side architecture content | Actual or modal C.30 content about the named organization holon and selected contribution, Work, decision, information, access, service, coordination, legal, or other structures. |
| product-or-service-side architecture content | Actual or modal architecture content about the named product, service System, offering System, platform, or other exact product-or-service-side holon. For each stated directional pressure, identify whether this content describes the influence-source side or the side being changed. |
| correspondence pressure | A bounded claim that one side’s selected structures influence the feasibility or burden of candidates on the other side through a named relation. It is not a universal law or automatic decision. |
| correspondence frame | The C.32.CONWAY synthesis frame used while either side is modal or the direct influence relation is unresolved. |
| exact correspondence row | A reusable C.32.CONWAY row about one obtaining direct influence relation whose two participants are obtaining C.30 architecture-relation occurrences, with each holon and selected structure recoverable. |
| evolution window | The period and expected change range over which the correspondence decision is intended to guide Work. |
| coordinated decision set | Two or more separately governed decisions linked by shared assumptions, constraints, and reopen conditions. |
| bounded mismatch | A deliberate choice to keep selected structures non-isomorphic for the stated window while accepting and managing the resulting burden. |
OCE.7:1 - Problem Frame
Products and services are produced, operated, supported, assured, and changed through organizations. Communication, deployment, test, approval, provider, evidence, and capability-home structures can make some product or service architectures easier to sustain. Technical dependencies can in turn create coordination and specialization pressure in the organization.
Correspondence can help without becoming one-to-one mirroring. Independent safety acceptance, scarce specialist homes, legal entities, platform services, regional operations, provider contracts, and continuing service can justify deliberate non-isomorphism. The task is to decide both sides with their gains, losses, authorities, and evolution windows visible.
OCE.7:2 - Problem
One-sided decisions externalize burden. A modular product can be assigned to nominally independent teams that still share one test rig, safety decision, data source, or specialist. A reorganized stream can inherit a tightly coupled product that requires constant cross-stream integration. A service boundary can be redrawn without the provider authority, observability, or recovery conditions needed to operate it.
Mirroring language hides these facts when it treats similarity as adequacy or inevitability. Practitioners can also mistake an influence on the design, expressed in a chart, architecture description, or decision record, for the Work that realizes it. They then cannot tell what relation created the pressure, which System performed the Work, or which side should change.
OCE.7:3 - Forces
| Force | Tension |
|---|---|
| Local autonomy | Aligned boundaries can reduce coordination, while shared safety, evidence, platform, and capability conditions can require cross-boundary relations. |
| Technical integrity | Product or service cohesion matters, while organization migration and provider arrangements constrain feasible change. |
| Specialist depth | Stable capability homes improve difficult Work, while contributors may face queues and extra handoffs across those boundaries. |
| Independent authority | Separate acceptance or governance can protect a characteristic, while it prevents full end-to-end ownership. |
| Evolution | Current alignment can be useful, while products, services, people, providers, and regulation change at different rates. |
| Evidence | Correspondence studies reveal contingent patterns, while a local decision still needs direct structures, relations, and consequences. |
OCE.7:4 - Solution
Frame one exact organization/product-or-service architecture pair and generate four candidate forms: change the organization side, change the product-or-service side, change both, or keep a bounded mismatch. Compare complete candidates across the declared evolution window. Make the organization and product-or-service decisions under their own authorities, then connect them through explicit constraints, accepted burdens, realization returns, and shared reopen conditions.
Recognition is cheap: one architecture choice whose feasibility depends on an unlike structure on the other side is enough to enter. Assurance is stronger: actual correspondence claims require the exact holons, selected structures, direct influence predicate and occurrence, conditions, evidence, and window. Modal material remains in the synthesis frame.
OCE.7:4.0 - One first paired decision
An instrument maker’s engineering organization and inspection product form one bounded pair: two contribution groups serve two product modules that share a test setup. For the next two releases, reuse their current architecture accounts, participant corrections and qualified test, safety and service constraints. The proposed coordination pressure stays in a synthesis frame unless its direct influence relation is established.
The two authorized decision-makers can work from one short paired note:
The organization-only candidate would merge the groups and disrupt an existing specialist-service commitment. The product-only and joint candidates require redesign that cannot fit the two-release window under the current engineering assessment. We therefore choose a bounded mismatch. The product decision retains the module and test boundaries. The organization decision retains the two groups and specifies one integration contribution and exception return per release. The gain is low migration burden; the accepted cost is shared-test coordination and no claim of independent release by each group. Reopen both decisions if the shared-test burden defeats the protected service commitment, or when the two-release window ends.
Each decision-maker records the decision for their own subject and authority. Send the needed assignment, access, test-time and service-protection requests to their owners for action before the dependent releases. Keep the existing evidence with the note. Use the fuller questions below for an unresolved claim or consequence that could change the pair, rather than rebuilding settled architecture accounts.
OCE.7:4.1 - Pattern-Use Unfolding
- Bind the paired question. Name the intended contribution, organization holon, product-or-service holon, decision subjects, authorities, current Work, horizon, evolution window, protected characteristics, and first users of both decisions.
- Recover both architecture sides. Separate actual
ArchitectureRelationoccurrences from candidate, required, desired, or expectedArchitectureClaimcontent. Name each selected structure, description source, currentness, and known loss. - Recover each directional pressure relation. State which side supplies the influence source and which side contains the changed architecture referent for this candidate; the direction can reverse between pressures. State how a communication, Work, test, deployment, approval, evidence, provider, capability-home, legal, service, or other source structure constrains the transformed-side candidate. Use the applicable direct predicate or keep a
C.32.CONWAYframe with a missing governor. A reciprocal claim requires its own reversed frame or relation occurrence. - Select decision characteristics. Name the few characteristics and burdens that can reverse the choice. Ask about the intended contribution, coordination, latency, changeability, safety, evidence, resilience, provider dependence, capability and service continuity, migration, effects on affected Systems, or another exact concern.
- Prepare all four candidate forms. Change the organization side while retaining product/service content; change the product/service side while retaining organization content; change both; or keep a bounded mismatch with explicit cost and return. A form can be rejected quickly when a non-negotiable condition fails.
- Obtain qualified domain inputs. Use available
SYSE.1,SYSE.2, andSYSE.9results for engineering focus, linked use/System concepts, and professional contributions when compatible. Use qualified direct sources or return the missing result when a sibling body is unavailable. - Compare whole consequences. Include current and transition Work, provider and platform arrangements, independent authority, evidence paths, scarce capability, legal and service boundaries, affected Systems, reversibility, and the burden of preserving deliberate non-isomorphism.
- Challenge the preferred pair. Ask which omitted dependency, exception, configuration, operating episode, or later evolution would reverse the choice. Use prototypes, simulations, participant criticism, sampled Work, or specialist results only within their evidence limits.
- Make separate decisions. Use
C.32.PAD,C.11, or the owning pattern for each decision. State selected structures, fixed constraints, open refinements, accepted losses, authority, effectivity, retained alternatives, and relation to the paired decision. - Specify realization and coexistence returns. Name the organization-change Work, product/service realization Work, continuing-service conditions, assignment/access needs, and observations each owner must return. Keep the realization Work and its evidence separate from the decisions selecting it.
- Reopen locally or jointly. Reconsider only the affected decision when one side changes without altering the correspondence choice. Reopen both when the pressure relation, accepted mismatch, protected characteristic, or evolution window changes materially.
- Stop at coordinated sufficiency. Return when both owners know what is selected, what remains open, why structures align or differ, which burden is accepted, and what evidence can reopen the pair.
OCE.7:4.2 - Record the Result
| Result position | Required content |
|---|---|
| paired boundary | Contribution, two holons, decision subjects, authorities, current Work, horizon, evolution window, protected characteristics, and first users. |
| architecture sides | Actual relation or modal claim status, selected structures, descriptions, sources, currentness, and known losses for each side. |
| correspondence | Direction of each pressure, direct influence predicate/occurrence or synthesis frame, source and transformed architecture content, affected characteristic, evidence, and missing governor. |
| candidate forms | Organization-side change, product/service-side change, joint change, bounded mismatch, expected gain, known loss, migration burden, and rejected non-negotiables. |
| comparison | Contribution, coordination, safety, evidence, resilience, provider, capability, service, affected-System, reversibility, and uncertainty consequences that change this decision. |
| separate decisions | Selected option and structure effects, authority, fixed/open boundary, accepted losses, retained alternatives, effectivity, and cross-reference to the paired decision. |
| realization returns | Work, assignments, access, provider, coexistence, specialist, observation, and evidence results required from each owner. |
| continuation | Local and joint reopen observations, source-return conditions, and end of the evolution window. |
OCE.7:4.3 - What Changes in Practice
Practitioners stop treating product and organization architecture as either independent or forced copies. They can change the cheaper or more valuable side, change both, or accept a visible mismatch. Independent safety, evidence, platform, provider, capability-home, and service boundaries become design constraints rather than embarrassing exceptions to a topology slogan.
OCE.7:5 - Archetypal Grounding – PumpWorks Product and Organization Decisions
PumpWorks is deciding how weekly AI-inspection releases should relate to the product’s module/evidence structure. The organization-side input is the OCE.4 contribution-architecture design. The product-side input identifies field-module boundaries, model artifacts, electrical compatibility evidence, shared platform services, safety evidence, and release configuration. Both sides remain possible-future where their direct relations do not yet obtain.
The current directional pressure uses the product-side architecture as influence source and the organization-side architecture as transformed side: module and evidence dependencies influence how independently release contributions can be prepared, tested, accepted, and deployed. The synthesis stays in a C.32.CONWAY frame until direct influence and both actual architecture relations are available. If an organization-side structure is later claimed to constrain a product-architecture candidate, PumpWorks records a second frame or occurrence with the direction reversed; reciprocity is not inferred from the first pressure.
| Candidate form | Proposed change | Main gain | Known loss or burden |
|---|---|---|---|
| organization-side change | Create one stream-aligned release configuration around the current product/evidence couplings | Shorter recurring coordination path | Cross-stream shared rig, Safety, platform, and scarce specialists remain bottlenecks |
| product/service-side change | Refactor module and evidence-package boundaries while retaining functional contribution homes | More independent test and evidence preparation | Product migration and assurance cost; organization coordination still spans releases |
| joint change | Align selected release-evidence packages and stream contributions while keeping shared platform and independent Safety relations explicit | Reduces some repeated crossings without hiding protected independence | Requires coordinated product refactoring, new assignments/access, and transition Work |
| bounded mismatch | Retain current product and functional organization for the window; add named integration and evidence-return relations | Lowest migration burden | Continuing coordination load and slower learning are accepted and measured |
PumpWorks selects the joint candidate for a bounded release family. The product architecture decision selects the alignment of module/evidence-package boundaries. The organization decision selects stream contribution boundaries and identifies the release-evidence integration need. Safety acceptance remains independent, platform and scarce capability homes remain shared, and provider support does not mirror a product module.
The two decisions cite the same assumptions and evolution window but retain separate authorities and realization Work. OCE.6 supplies assignments and access; product realization remains with Systems Engineering; OCE.9 later tests organization relations and capability; Operations supplies continuing-release and service observations. A changed safety regime, provider boundary, platform coupling, product family, or observed coordination burden reopens the affected decision pair.
OCE.7:5.1 - Transfer Probes
| Setting | Reusable move | Required return or changed content |
|---|---|---|
| public-hospital emergency flow | Pair the hospital organization structures with the emergency-service architecture: triage, diagnostics, treatment, bed flow, escalation, and information continuity | There may be no product modularity question; statutory clinical authority, privacy, labor, safety, facility, and continuing Operations can justify deliberate non-isomorphism |
| distributed standards association | Pair volunteer/editorial organization structures with the standard-development and publication-service architecture | Bylaw decisions, ballots, employer resources, volunteer availability, repositories, publication services, and language communities evolve at different rates; no single firm boundary or executive authority is assumed |
OCE.7:6 - Bias-Annotation
| Recurring bias | Likely drift | Repair |
|---|---|---|
| mirroring determinism | Structural similarity is treated as a law or quality. | Recover the exact pressure, candidate forms, gains, losses, and local decision. |
| organization-only repair | Teams move while product/service dependencies remain unchanged. | Include the product/service-side and joint candidates. |
| product-only repair | Technical modularity is expected to remove authority, evidence, or provider coordination. | Preserve organization-side structures and enabling relations. |
| autonomy prestige | End-to-end ownership hides shared safety, platform, capability, and service relations. | State protected independent/shared contributions and their burdens. |
| diagram causality | Architecture descriptions are said to create the result. | Name Systems, Work, direct influence relations, decisions, and realization separately. |
| window blindness | Current alignment is treated as permanent. | State the evolution window and asymmetric change rates. |
OCE.7:7 - Conformance Checklist
- The contribution, two exact holons, decisions, authorities, current Work, and evolution window are explicit.
- Actual architecture relations and modal claims remain distinguishable on both sides.
-
The correspondence uses a direct predicate/occurrence or an explicitly provisional
C.32.CONWAYframe. - Organization-side, product/service-side, joint, and bounded-mismatch candidates are considered.
- Comparison includes the characteristics, independent/shared contributions, migration, affected Systems, and uncertainty that can change this decision.
- Systems Engineering or other domain results are consumed only when available and compatible.
- The paired decisions retain separate authorities, subjects, fixed/open boundaries, and realization Work.
- Local and joint reopen observations are stated.
OCE.7:8 - Common Anti-Patterns and How to Avoid Them
| Anti-pattern | Repair |
|---|---|
| “The organization must mirror the product architecture.” | State the exact structures and pressure relation, then compare all four candidate forms and accepted losses. |
| “Every service has one autonomous team.” | Recover shared platform, data, safety, evidence, provider, capability, Operations, and governance relations before deciding autonomy. |
| “Refactor the monolith and coordination will disappear.” | Name the organization Work and relations that the technical change is expected to alter, and observe them after realization. |
| “Reorganize around customer journeys.” | Identify the service and contribution structures, decision authority, shared capabilities, provider boundaries, and journeys’ representation losses. |
| “The Conway workshop approved the target.” | Treat workshop output as candidate and participant evidence; make the separate architecture decisions under current authority. |
OCE.7:9 - Consequences
Product-or-service and organization changes become one coordinated design question without losing their distinct subjects and authorities. Deliberate non-isomorphism becomes a controllable decision with visible cost.
The cost is broader evidence and coordination. Two decision owners may need different sources, and the attractive joint candidate can remain blocked by migration, safety, provider, capability, or continuing-service conditions.
OCE.7:10 - Rationale
Mirroring is useful as a contingent hypothesis about coordination and cognitive burden. Current reviews and cases also show co-evolution, changing causal direction, value- and regulation-sensitive exceptions, and situations where deliberate non-correspondence supports search or contribution. A four-form comparison turns that evidence into constructive alternatives instead of a slogan.
Separate decisions preserve accountability and distinguish selected changes from realized ones. An organization change can be selected while product realization is still future, or a product architecture can change while the organization remains stable. Their correspondence matters only through named relations and consequences.
OCE.7:11 - SoTA-Echoing
OCE.7:11.1 - Current-Line Selection for Coordinated Architectures
| Comparison position | Selected result for practice |
|---|---|
| Current question | Which organization-side, product-or-service-side, joint, or bounded-mismatch candidate best serves the contribution across the declared evolution window? |
| Selected current line | Treat architecture correspondence as contingent, directional, and co-evolving. Compare all four candidate forms because value capture, regulation, authority, evidence, search, migration, provider, and capability conditions can make correspondence or deliberate non-correspondence preferable for a bounded use. |
| Serious alternative | Apply universal mirroring or fixed end-to-end ownership from product modularity, a communication pattern, or a team-topology slogan. |
| When the alternative is sufficient | Similar boundaries can be selected when direct local evidence shows that they reduce the decision-bearing burden without defeating protected shared or independent relations. |
| When the selected line changes action | If value, search, regulation, authority, shared platform, independent assurance, provider, or changing interorganization boundaries alter the consequences, retain the unlike structures and compare joint change and bounded mismatch explicitly. |
| Reopen | Reconsider when a recurring case needs another candidate form, current evidence changes the correspondence answer, or observed realization exposes a missed pressure or burden. |
OCE.7:11.2 - Source Contributions and Boundaries
| Source line | Retained contribution | Use boundary |
|---|---|---|
Current FPF C.30, C.32, C.32.CONWAY, C.32.PAD, A.22, and A.6.REL | Exact holons, actual and modal architecture, candidate synthesis, four correspondence forms, decisions, structures, and relation obtaining remain distinct. | Apply these distinctions while the authorized owners make the organization and product/service decisions separately. |
| Joseph and Sengul, current organization-design review | Contemporary organization design is multi-approach and contingent; coordination and structure cannot be reduced to one representation or feature. | Select the local product/organization pair and test the relevant consequences before deciding. |
| Zani, Denicol, and Broyd, megaproject organization-design review | Temporary and interorganizational boundaries evolve, and integration of product and organization architectures is a distinct design question. | Qualify transfer from megaprojects before using the evidence to choose a lifecycle, correspondence strategy, or product change in another setting. |
| Brusoni et al., twenty-year modularity synthesis | Product, organization, and industry architectures co-evolve; causal direction and the value of fit depend on context, evolution, and the question being asked. | Recover local authority and any claimed direct influence, then compare candidates for the declared window. |
| Burton and Galvin, value and mirroring exceptions | Regulation and value capture can make deliberate non-correspondence and stronger supplier ties preferable despite product modularity. | Treat the result as evidence from one industry case; test whether its value and regulatory conditions apply to PumpWorks. |
| Conway, historical communication/architecture anchor | Communication arrangements can shape designed-system structure. | Use this historical anchor to ask about communication pressure; qualify the direct local relation and compare mirroring with the other candidate forms. |
| MacCormack, Rusnak, and Baldwin, historical product/organization architecture study | Comparable products developed under different organization forms provide evidence of architecture correspondence. | Test correspondence in the local pair and compare redesign consequences before selecting a change. |
| Colfer and Baldwin, mirroring evidence and exceptions | Mirroring is prevalent but not universal; exceptions and organizational forms matter. | Assess local adequacy separately; recover decision authority and, when realization is claimed, the actual performers and Work. |
| DORA, loosely coupled teams | Team independence, testing, deployment, and coordination load provide practitioner recognition and possible pressure variables. | Use this software-heavy practitioner guidance for recognition, then test local relations and ownership against the broader research and the organization’s conditions. |
Reopen when a recurring organization/product-or-service case needs another decision-changing candidate form, a current source changes the correspondence answer, or observed realization shows that the selected pressure and burden variables miss the actual constraint.
OCE.7:12 - Relations
OCE.3supplies organization concepts and correspondence assumptions.OCE.4supplies the organization-side contribution-architecture design.OCE.6can supply effective assignments and enabling relations needed by realization.C.32.CONWAYsupplies correspondence framing and exact rows;C.32.PADsupplies project architecture decisions;C.30andA.22govern architecture content and selected structures.- Available compatible
SYSE.1,SYSE.2, andSYSE.9results may supply engineering focus, linked use/System concepts, and qualified contributions. Systems Engineering retains product architecture and realization. OCE.9consumes organization design constraints and later returns actual relation/capability evidence.OCE.11can coordinate organization-change and continuing-service Work. Operations returns operating observations.- Governance, Administration, legal, labor, safety, privacy, finance, providers, HCD, and other practices keep their decisions, predicates, evidence, and authority.
OCE.7:End