Frame a decision-led BA engagement.
Move from an ambiguous request to a bounded question, targeted discovery, and a decision-ready review.
Cookbook · Recipes
Use a recipe for the method, a work product for the artifact, or the Northstar case study to see the pieces connected. Recipes begin with a real organizational problem and end with a stop condition so the architecture does not become work for its own sake.
Do not run every recipe and do not produce every artifact. Find the decision you need to improve, choose the closest problem pattern, and tailor the moves to the situation.
The recipes assume that capabilities, value streams, stakeholders, information, organization, products, strategies, initiatives, policies, and metrics form a linked body of knowledge. They are separated here only to make the work manageable.
The Cookbook contains 45 complete, practical recipes. Each explains when to use the method, what evidence to bring, the moves to make, the work products it can support, and when to stop.
Common recipe progressions
These are decision paths, not a required methodology. Core steps usually preserve the logic of that path. Conditional steps apply only when the stated condition exists. Enter at the first unresolved question, reuse what the enterprise already governs, and stop when another recipe will not improve the decision.
Discover and structure
Start here when the decision is unclear, the connected architecture is missing, or the practice itself needs a workable mandate and service model.
Move from an ambiguous request to a bounded question, targeted discovery, and a decision-ready review.
Connect capabilities, value, information, and organization around the decisions the baseline must support.
Clarify the mandate, operationalize a small service, and enter the forums where enterprise choices are already made.
Analyze and decide
Enter at the unresolved question. Build only the missing evidence needed to compare options, expose consequences, and support the accountable decision.
Clarify outcomes before testing portfolio coverage, capability needs, concentration, and transition order.
Separate the observed condition from assumed causes, then determine what analysis and change are actually justified.
Distinguish prioritization from diagnosis, then define the operating change and transition needed to improve the outcome.
Scan the portfolio first, then investigate only the pairs, gaps, or stalled commitments that require a disposition.
Change and realize
Use these paths when the change must become operable, traceable, adopted, and measurable—not merely approved or delivered.
Preserve the intended outcome while defining scope, operating change, traceability, readiness, and post-release value evidence.
Separate interpretation from architecture impact, then carry the required change through readiness, evidence, and realized outcome.
Govern and maintain
Use these paths when existing architecture has uncertain structure, authority, consistency, lifecycle, or continuing usefulness.
Choose the defect you actually have, preserve useful institutional knowledge, and validate only the repaired or reconciled baseline.
Keep knowledge authoritative through publication, reuse, initiative closure, supersession, review, and retirement.
Find the work
Search the full cookbook or narrow it by problem family, architecture domain, engagement stage, and practice maturity. For field notes, public resources, CBA guidance, and SCALE, use site-wide search.
Showing all 45 recipes.
The request arrives as a proposed solution, a vague transformation, or a demand for “a capability map,” but the accountable decision, evidence threshold, authority, and boundary have not been established.
Stop condition: Stop framing when the accountable owner can state the decision, authority, evidence required, scope, participants, review point, and completion test. Do not begin modeling merely because an artifact was requested or because the proposed solution is politically advanced.
Frame the decision before selecting architecture artifacts or beginning modeling.
Boundary: Does not replace a project charter, investment approval, requirements package, or the accountable executive’s decision.
Provide just enough structure to frame an engagement, establish a small baseline, assess initiative impact, and conduct a bounded review.
Boundary: These are starting structures, not a substitute for local governance, evidence, accountable decisions, or the deeper work products required by complex change.
The enterprise needs a stable view of what it must be able to do, but the first draft mirrors departments, applications, products, process steps, or a borrowed reference model instead of enduring business abilities.
Stop condition: Stop when the map provides a stable, governed vocabulary at sufficient depth for the stated decision and the next level would not change assessment, investment, accountability, dependency, sequence, scope, or risk. Do not complete every branch merely for visual symmetry.
Build a governable capability baseline that is sufficiently detailed for a stated decision.
Boundary: Does not provide an authoritative industry reference model or prove that a capability belongs at a particular level without enterprise context and governance.
Provide just enough structure to frame an engagement, establish a small baseline, assess initiative impact, and conduct a bounded review.
Boundary: These are starting structures, not a substitute for local governance, evidence, accountable decisions, or the deeper work products required by complex change.
The “value stream” becomes a swimlane full of activities, systems, functions, and handoffs, obscuring the progression through which a stakeholder actually receives value.
Stop condition: Stop when the stakeholder, trigger, value proposition, stages, criteria, enabling capabilities, and material variations are sufficient for the decision. If a stage must be explained as a task sequence, move that detail to a process model rather than weakening the value-stream view.
Connect stakeholder value progression to the stable abilities that enable each stage.
Boundary: Does not replace a process, journey, operating procedure, service blueprint, or control design when activity and handoff detail is required.
Provide just enough structure to frame an engagement, establish a small baseline, assess initiative impact, and conduct a bounded review.
Boundary: These are starting structures, not a substitute for local governance, evidence, accountable decisions, or the deeper work products required by complex change.
Initiative scope is expressed as features, applications, workstreams, or a project charter with no defensible explanation of which business outcomes, value stages, capabilities, information, organization, policies, or products must actually change.
Stop condition: Stop when every material scope item traces to a business outcome and affected architecture relationship, dependencies and operating obligations are visible, and decision makers can select a bounded option. Untraceable scope must be removed, deferred, or retained through an explicit exception.
Define the smallest release that can produce measurable business value or decision-quality learning.
Boundary: Does not replace funding approval, delivery estimation, solution design, backlog management, or benefits-realization ownership.
Provide just enough structure to frame an engagement, establish a small baseline, assess initiative impact, and conduct a bounded review.
Boundary: These are starting structures, not a substitute for local governance, evidence, accountable decisions, or the deeper work products required by complex change.
The strategy has been translated into explicit outcomes, but leaders still cannot show whether funded initiatives collectively cover those outcomes, which investments lack strategic justification, or where measures and ownership break the execution logic.
Stop condition: Stop when portfolio authorities can see which investments credibly implement each strategic outcome, which outcomes remain uncovered, which initiatives lack justification, and what decision or evidence is required. Do not reinterpret unresolved strategy or infer contribution from a linkage alone.
Make the logic from strategic intent to capability change and investment testable.
Boundary: Does not create strategy, authorize investment, or establish causation between an initiative and a business outcome merely because they are linked.
Architecture review becomes a document checkpoint or presentation ritual rather than a forum for resolving a consequential choice with explicit evidence, authority, alternatives, and follow-through.
Stop condition: Do not bring the review forward until the authorized forum can approve, reject, conditionally approve, defer, or return a clearly stated decision using sufficient evidence. “For awareness” belongs in a different communication, and the review brief does not replace the authoritative decision record.
Frame the decision before selecting architecture artifacts or beginning modeling.
Boundary: Does not replace a project charter, investment approval, requirements package, or the accountable executive’s decision.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
Provide just enough structure to frame an engagement, establish a small baseline, assess initiative impact, and conduct a bounded review.
Boundary: These are starting structures, not a substitute for local governance, evidence, accountable decisions, or the deeper work products required by complex change.
A charter or executive intention exists, but the practice still lacks a workable service model, intake, capacity rules, knowledge governance, forum integration, and evidence that it improves decisions.
Stop condition: Stop initial operationalization when stakeholders can engage the practice, practitioners can govern demand within capacity, defined forums consume the evidence, and the first services have measurable decision value. Do not scale the repository, service catalog, or team faster than the organization can govern and use them.
A workshop, engagement, or initiative produces useful architecture, but the whiteboard dies in a deck, proposals become indistinguishable from approved enterprise truth, and later teams must rediscover the same knowledge.
Stop condition: Stop when a consumer can determine what is authoritative, proposed, disputed, historical, or retired; who owns it; what evidence supports it; when it applies; what changed; and when it must be reviewed. Do not govern transient material merely because it was produced in an architecture workshop.
The practice begins with enthusiasm but no shared purpose, boundaries, decision rights, engagement model, or definition of success.
Stop condition: Stop when the sponsor and core participants can explain why the practice exists, what it will do first, who owns its decisions, how teams engage it, and what evidence will justify expansion. Do not wait for a complete operating model.
Download the BA practice charter
Make a proposed BA practice mandate and initial operating assumptions explicit enough for sponsor agreement.
Boundary: The template does not create authority by itself; its mandate, services, funding, and decision rights still require accountable approval and operational follow-through.
The information model has become a long, inconsistent list that mixes genuine business objects with synonyms, attributes, documents, events, roles, applications, and local terminology.
Stop condition: Stop when the model supports the stated decisions and scenarios, every retained object has a distinct business meaning or lifecycle, and adding another object would not change ownership, rules, scope, or linkage. Unresolved terms may remain as governed issues.
Create a shared business vocabulary and lifecycle view that can support architecture, policy, process, and automation work.
Boundary: Does not replace a logical or physical data model, master-data design, records-retention schedule, API contract, or legal definition.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
Business information is documented, but the organization cannot show where it enters the business, which capabilities change it, how it moves through value delivery, or who governs its lifecycle.
Stop condition: Stop when every material lifecycle change in scope has a business reason, a value-stream location, a responsible capability, and an accountable steward, and the resulting relationships are sufficient for the stated decision. Do not map every object to every capability that merely views it.
Create a shared business vocabulary and lifecycle view that can support architecture, policy, process, and automation work.
Boundary: Does not replace a logical or physical data model, master-data design, records-retention schedule, API contract, or legal definition.
The capability map is either too shallow to support a decision or decomposed into uneven detail that nobody can consistently assess, own, or use.
Stop condition: Stop when the next level would not change a decision, expose a material difference, or support separate governance. Do not decompose merely to make every branch visually symmetrical.
Build a governable capability baseline that is sufficiently detailed for a stated decision.
Boundary: Does not provide an authoritative industry reference model or prove that a capability belongs at a particular level without enterprise context and governance.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
An existing capability map is dominated by departments, roles, processes, applications, products, and duplicated local abilities, but stakeholders already depend on it and replacing it wholesale would destroy useful knowledge and credibility.
Stop condition: Stop when the map remains intelligible through plausible organization and technology changes, duplicated abilities have a disposition, and existing consumers can trace old terms to the governed baseline. Do not discard useful relationships merely to produce a visually clean replacement.
Build a governable capability baseline that is sufficiently detailed for a stated decision.
Boundary: Does not provide an authoritative industry reference model or prove that a capability belongs at a particular level without enterprise context and governance.
A broad statement such as “improve customer experience” or “modernize operations” creates activity, but it does not identify the outcomes, business changes, evidence, or choices needed for execution.
Stop condition: Stop when the strategy owner can choose priorities using a traceable view of the outcomes, capability changes, dependencies, measures, and options. Do not invent implementation detail to compensate for unresolved strategic choices.
Make the logic from strategic intent to capability change and investment testable.
Boundary: Does not create strategy, authorize investment, or establish causation between an initiative and a business outcome merely because they are linked.
A portfolio scan, capability heat map, or governance discussion has flagged two or more initiatives as potentially duplicative, conflicting, complementary, or dependent, but shared mappings alone do not establish the correct disposition.
Stop condition: Stop when the flagged initiative relationship has an evidence-backed classification, the portfolio authority can choose a disposition, and every resulting condition or dependency has an owner. Do not label initiatives redundant solely because they map to the same capability, value stage, system, or strategic objective.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
A discovery meeting is scheduled before the decision, participants, evidence, working structure, and expected outputs are clear, so the session produces broad conversation instead of usable architecture knowledge.
Stop condition: Stop when the right participants understand the decision, facts are separated from assumptions and disagreements, the minimum architecture outputs exist, and every unresolved item has an owner and date. Full consensus and a complete enterprise model are not required.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
A capability model needs business validation, but the workshop risks becoming an unstructured debate about department names, process steps, systems, terminology, and personal ownership claims.
Stop condition: Stop when the model is sufficiently validated for its stated decision, accepted changes and dissent are recorded, and unresolved issues have owners and governance paths. Do not prolong the workshop to force unanimity or perfect every branch outside the agreed scope.
Build a governable capability baseline that is sufficiently detailed for a stated decision.
Boundary: Does not provide an authoritative industry reference model or prove that a capability belongs at a particular level without enterprise context and governance.
Two groups use different terms, definitions, or boundaries and each insists that its version represents a separate capability, leaving the enterprise with duplication, false standardization, or unresolved ownership conflict.
Stop condition: Stop when the competing terms have a documented semantic disposition that works across representative scenarios and supports the required decision. Escalate unresolved decision rights separately rather than multiplying capabilities to satisfy organizational claims.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
A capability map has been colored red, yellow, and green without a clear decision, consistent dimensions, defensible evidence, or any indication of confidence, creating visual certainty without analytical discipline.
Stop condition: Stop when each emphasized capability has a defensible score, visible evidence and confidence, and a clear connection to the decision. Do not color unassessed capabilities or use a composite score that decision makers cannot explain.
Support two distinct decisions: Recipe 19 prioritizes capability attention, while Recipe 41 diagnoses what prevents a selected capability from producing the required outcome.
Boundary: A heat-map color prioritizes attention; it does not diagnose cause. The improvement assessment structures a diagnosis; it does not prove causation or authorize investment.
The first release is either overloaded with desirable scope or reduced to a technical component that cannot produce usable business value on its own.
Stop condition: Stop when the release can produce a safe, usable, measurable business outcome under accountable operational ownership and additional scope adds less value than cost or delay. Do not call an isolated technical component a useful release unless it independently resolves the stated business need.
Define the smallest release that can produce measurable business value or decision-quality learning.
Boundary: Does not replace funding approval, delivery estimation, solution design, backlog management, or benefits-realization ownership.
An initiative remains nominally active but repeatedly misses decisions, funding, dependencies, or mobilization, while status reporting describes delay without identifying the constraint that must actually change.
Stop condition: Stop when the governing authority can choose among explicit dispositions using evidence about the real constraint, consequences, and owner. If no actor has both authority and incentive to proceed, classify the initiative as paused or stopped rather than architecturally active.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
A broad transformation is divided by department, application, funding source, or delivery team, producing work packages that cannot independently explain their value, dependencies, or contribution to the intended business change.
Stop condition: Stop when each package has a distinguishable contribution, accountable owner, coherent boundary, and explicit dependencies, and the set collectively covers the intended outcomes. Do not turn the decomposition into task-level delivery planning.
Create manageable investment and transition decisions while preserving end-to-end business intent.
Boundary: Does not replace a project schedule, resource plan, delivery backlog, committed date, or portfolio authority.
A capability roadmap either remains a timeless aspiration or becomes a delivery schedule full of projects, tasks, and dates that obscures the business abilities and outcomes being changed.
Stop condition: Stop when leaders can decide what capability changes should occur, in what dependency order, for what outcome, and under whose accountability. Do not add delivery detail that belongs in a program or project plan.
Create manageable investment and transition decisions while preserving end-to-end business intent.
Boundary: Does not replace a project schedule, resource plan, delivery backlog, committed date, or portfolio authority.
A requirement is documented and prioritized without a durable explanation of which stakeholder outcome, value-stream stage, capability change, information need, policy, or initiative decision justifies it.
Stop condition: Stop when a reviewer can explain why the requirement exists, what business change and evidence justify it, and what would be affected if it changed or disappeared. Do not trace every requirement to every nearby architecture element or duplicate delivery-tool detail.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
A platform, application, or emerging technology is presented as the solution before the desired outcome, capability gap, value-stream impact, information need, operating-model change, and policy implications have been established.
Stop condition: Stop when decision makers can see whether technology is necessary, what non-technology changes are also required, who will own the resulting capability, and how value will be measured. Return the request for reframing if the product remains the only defined outcome.
Support focused analysis when a lightweight, portable structure is more useful than a governed workbook.
Boundary: The pack structures analysis but does not supply enterprise evidence, adjudicate ownership, or convert a working conclusion into an approved decision.
Business architecture is expected to create enterprise value without the mandate, access, funding, or decision rights normally supplied by an executive sponsor.
Stop condition: Stop when the bounded decision has been improved and its owner accepts the result, or when missing authority, evidence, or access makes further work misleading. Do not quietly build an enterprise repository or imply governance that the organization has not granted.
Business architecture sits inside an organization whose governance, funding, language, and demand are centered on technology, creating pressure to become an upstream solution-design service instead of a business decision discipline.
Stop condition: Stop when BA has a recognized role in defined business decisions, business owners govern the relevant content, technical teams can consume traceable business context, and the practice can decline work that is only disguised solution design. Organizational relocation is not required for the discipline to remain business led.
Multiple business units claim to own one capability because ownership is being used to mean content stewardship, investment authority, performance accountability, policy control, or operational execution at the same time.
Stop condition: Stop when every material capability decision has an accountable authority, contributors know their role, and enterprise consistency and local discretion have explicit boundaries. Do not force a single undifferentiated owner onto a capability whose governance is legitimately distributed.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
Business architecture is added as a separate review, template, or approval gate, increasing cycle time while portfolio decisions continue to rely on the same incomplete strategic and business evidence.
Stop condition: Stop when existing portfolio authorities receive the minimum architecture evidence at the point of decision, engagement is proportionate to risk and value, and no separate BA approval is needed. Do not scale the integration if it adds effort without measurably improving decisions or reducing downstream rework.
The EA repository is becoming either an indiscriminate document store or an incomplete diagram catalog, while practitioners duplicate authoritative data and struggle to decide which knowledge deserves governed relationships and lifecycle control.
Stop condition: Stop when each material content type has an authoritative home, relationship and lifecycle needs are met, and the repository can answer priority questions without duplicating specialist systems. Do not put information in the EA tool merely because it can be modeled there.
A proposed strategy, policy, business model, or emerging concept needs architectural analysis before implementation exists, creating pressure either to invent a target solution or to postpone useful modeling until decisions have already hardened.
Stop condition: Stop when decision makers can compare the concept and alternatives, see what would have to be true, and identify the next evidence or authorization required. Do not fill implementation gaps with invented certainty or publish proposed elements as current enterprise truth.
When an initiative is cancelled, merged, defunded, or abandoned, validated enterprise knowledge disappears with its project files while obsolete scope, target states, and assumptions remain discoverable without clear status.
Stop condition: Stop when consumers can distinguish current enterprise knowledge from historical initiative intent, successor links are accurate, and retained material has an owner or retention basis. Do not preserve obsolete project scope as an active target simply because it once received funding.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
Architecture from a prior initiative is copied into a new context as an assumed standard, carrying forward local choices, stale assessments, ownership, and constraints that may not represent durable enterprise knowledge.
Stop condition: Stop when every reused element has a valid target-context rationale and local differences have been either justified or removed. Do not copy scores, owners, target states, or implementation choices merely because the source initiative was successful.
Teams either model process by habit when capability and value views would answer the decision, or avoid process detail when sequence, handoffs, controls, timing, and exception behavior are essential.
Stop condition: Stop when the process detail is sufficient to resolve the sequence, handoff, control, timing, or execution decision. Do not decompose activity further when the next level only documents routine procedure without changing requirements, risk, accountability, or improvement action.
Architecture continues because the enterprise can always be modeled more completely, or it ends when a deadline arrives without confirming that the decision, governance, and downstream consumers have what they need.
Stop condition: Stop when the decision can be made responsibly, material knowledge is governed, unresolved issues have explicit dispositions, and another modeling cycle would not change scope, priority, accountability, risk, or action enough to justify its cost. The enterprise architecture may continue to evolve after the engagement is finished.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
The portfolio shows projects, spend, and dates, but leaders cannot see where change is concentrated across business capabilities, which important abilities are neglected, or where multiple initiatives depend on the same enterprise capacity.
Stop condition: Stop when portfolio leaders can see where change is concentrated, where important capability needs lack credible investment, and which overlaps or dependencies require a decision. Do not infer value, redundancy, or adequate coverage from color, project count, or spend alone.
Define the smallest release that can produce measurable business value or decision-quality learning.
Boundary: Does not replace funding approval, delivery estimation, solution design, backlog management, or benefits-realization ownership.
Support two distinct decisions: Recipe 19 prioritizes capability attention, while Recipe 41 diagnoses what prevents a selected capability from producing the required outcome.
Boundary: A heat-map color prioritizes attention; it does not diagnose cause. The improvement assessment structures a diagnosis; it does not prove causation or authorize investment.
Create manageable investment and transition decisions while preserving end-to-end business intent.
Boundary: Does not replace a project schedule, resource plan, delivery backlog, committed date, or portfolio authority.
A visible symptom such as customer abandonment, rising cost-to-serve, delay, rework, or control failure is being assigned to a familiar function or proposed solution before the enterprise relationships and plausible causes are understood.
Stop condition: Stop when the decision owner can distinguish the observed condition from its hypothesized causes, see the material enterprise relationships, and authorize the next test or intervention. Do not map the whole enterprise or claim that architectural proximity proves causation.
Frame the decision before selecting architecture artifacts or beginning modeling.
Boundary: Does not replace a project charter, investment approval, requirements package, or the accountable executive’s decision.
Connect stakeholder value progression to the stable abilities that enable each stage.
Boundary: Does not replace a process, journey, operating procedure, service blueprint, or control design when activity and handoff detail is required.
Two capability maps use different names, boundaries, levels, identifiers, and organizing assumptions, and each carries useful relationships or institutional authority that would be lost by simply declaring the other map correct.
Stop condition: Stop when every material source element has a disposition, consumers can trace old identifiers and terms to the governed baseline, and unresolved conflicts have owners and decision paths. Do not force semantic uniformity where evidence supports a legitimate business variation.
Build a governable capability baseline that is sufficiently detailed for a stated decision.
Boundary: Does not provide an authoritative industry reference model or prove that a capability belongs at a particular level without enterprise context and governance.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
A consequential architecture decision is buried in meeting notes, presentation slides, email, or memory, leaving later teams unable to tell what was decided, by whom, from which evidence, under what conditions, or when the decision should be reconsidered.
Stop condition: Stop when a future reader can identify the authoritative decision, authority, rationale, evidence, conditions, affected architecture, required actions, and reconsideration trigger without reconstructing the meeting. Do not use the record as a substitute for required corporate minutes, legal approval, delegated authority, or delivery governance.
Preserve the institutional memory required to govern architecture through review, dispute, handoff, and completion.
Boundary: Does not replace formal corporate minutes, legal approvals, records-management policy, delegated-authority schedules, or delivery issue systems.
The organization chart shows reporting relationships, but architecture decisions still cannot identify which internal and external organizations participate in value delivery, perform or govern capabilities, control resources, or hold material decision rights.
Stop condition: Stop when the organizations and relationships material to the decision can be traced to capability performance, value delivery, information, policy, resources, and authority, and unresolved accountability has a decision path. Do not reproduce an HR org chart, person-level directory, or universal RACI in the architecture repository.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.
A capability has been rated important or underperforming, but a maturity score or heat-map color does not explain what prevents it from producing the required outcome, which constraints matter most, or what change is justified.
Stop condition: Stop when decision makers can identify the material constraints preventing the required capability outcome, compare credible intervention options, and authorize the next change or evidence test. Do not convert the assessment into an enterprise-wide maturity exercise or assume that a numerical score proves causation.
Support two distinct decisions: Recipe 19 prioritizes capability attention, while Recipe 41 diagnoses what prevents a selected capability from producing the required outcome.
Boundary: A heat-map color prioritizes attention; it does not diagnose cause. The improvement assessment structures a diagnosis; it does not prove causation or authorize investment.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.
An initiative describes features, processes, systems, or work packages but does not define how the affected capabilities must operate after delivery, who will own the outcome, or which organizational, information, policy, capacity, and support changes must endure.
Stop condition: Stop when delivery and operating authorities can explain how the changed capabilities will function, who will own and support them, what transition obligations are required, and how performance will be measured. Do not turn the operating-model view into a detailed organization design, solution architecture, workforce plan, or project schedule.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.
Deployment, feature completion, or project closure is being treated as benefit realization even though the organization cannot show the adoption, capability performance, value-stream change, and stakeholder outcome through which value would actually appear.
Stop condition: Stop when the accountable authority can determine from observed evidence whether the change is being adopted, improving capability performance, and producing the intended stakeholder outcome, and can choose the next disposition. Do not claim realized value from delivery completion, adoption activity, or architecture linkage alone.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.
A release is technically complete but the enterprise has not established whether people, capacity, information, controls, policies, support, ownership, partners, and operating measures are ready to sustain the intended business outcome.
Stop condition: Stop when the authorized owner can make a bounded release disposition using evidence about the enterprise’s ability to operate, support, control, measure, and recover the change. Do not use this recipe as a substitute for technical production readiness, legal approval, safety certification, security authorization, or delivery governance.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.
A new regulation, policy, contract, standard, or control obligation is assigned to a familiar function or technology team before the enterprise can trace what behavior, information, accountability, value delivery, and evidence must actually change.
Stop condition: Stop when authorized leaders can see where the obligation applies, what business behavior and evidence must change, who holds each decision right, and how implementation and readiness will be governed before the effective date. Do not substitute business architecture interpretation for legal advice or formal regulatory authority.
Keep the business side of change connected from architecture diagnosis through adoption and measurable outcome realization.
Boundary: The pack does not replace organizational design, delivery plans, change-management plans, control testing, legal interpretation, or benefits ownership; it connects those disciplines around explicit architecture evidence and decisions.