The Transformation Scope Decision: Improve, Automate, Modernise or Transform

Before commissioning a transformation programme, agree what actually needs to change. A workflow defect, an automation opportunity and a system replacement need different responses. Choose the scope the evidence supports rather than asking a technology renewal to solve unclear ownership or incentives.

This operating guide helps a CEO, COO and CIO classify the smallest coherent intervention before commissioning work. Its controlled disposition is IMPROVE | AUTOMATE | MODERNISE | TRANSFORM | VALIDATE | DEFER | STOP. A decision may contain subordinate components, but it must still have one primary scope class, one accountable outcome and one next evidence gate.

1. The Market Connection Context

The recognisable condition is a leadership paper that asks for transformation funding while different participants mean different things. Operations expects fewer handoffs, technology expects platform renewal, finance expects a cost change, and the sponsor expects a different customer outcome. None can state which operating conditions must change together or which cheaper intervention has been rejected with evidence.

The primary value path is:

Strategy and funded decisions
→ correctly bounded intervention
→ changed work, technology or operating model
→ accepted customer or service outcome
→ measured operating evidence
→ continue, change, stop or scale decision

A scope error affects the bottom line indirectly but materially. Overscoping can consume scarce capacity, increase transition risk and postpone a bounded repair. Underscoping can automate a broken rule, modernise technology while preserving a failing workflow, or improve one team while a cross-functional constraint remains. Revenue, margin, cost, capacity, retention, risk, investment efficiency and service value effects are REQUIRES VALIDATION until the organisation accepts a population, period, baseline, source and causal chain.

The seven controlled dispositions

“Hybrid” is not a disposition. Most meaningful interventions contain several component treatments. Record those components beneath one primary class so leadership can see what governs the whole decision. For example, a transformation may contain automation and modernisation components; those components do not make every affected activity a transformation.

The working classification is an Abrasim operating methodology, PROPOSED until the organisation accepts it for the decision. It is not an industry taxonomy and does not prove value.

The decision to make

Complete this statement:

By [decision date], [named decision owner] will classify [bounded outcome and intervention] as [IMPROVE, AUTOMATE, MODERNISE, TRANSFORM, VALIDATE, DEFER or STOP], because retained evidence shows that [named work, technology and operating-model conditions] must change together, subject to [customer, employee, service, financial, legal, privacy, security and operational guardrails].

The decision authorises the next gate, not an entire implementation. Funding, capacity and procurement remain separate approvals.

Appropriate use, exclusions and dependencies

Use this method when several teams are using incompatible scope language; a proposed solution precedes a settled problem; a programme appears too large or too narrow; technology renewal is being confused with operating change; or leadership needs a common classification before an investment case, roadmap, architecture or procurement.

Do not use the full method for an urgent incident requiring immediate qualified containment, a small defect with a trusted cause and authorised owner, or a decision already governed by a more specific assurance process. Stop if leaders will not allow alternatives, evidence access is unlawful or unsafe, or no owner can accept changed operation.

The scope includes one outcome, affected population, value-path boundary, performed workflow, relevant policy and controls, technology/data constraints, operating-model conditions, capability, dependencies, component treatments and decision gate. It excludes detailed root-cause diagnosis under The Operating Model and Workflow Root-Cause Audit: Finding the Constraint Beneath the Programme, portfolio economics under The Transformation Portfolio and Investment Case: Funding the Smallest Coherent Set of Changes, roadmap sequencing under The Fundable Transformation Roadmap: Outcomes, Dependencies, Capacity and Decision Gates, governance redesign under The Transformation Governance and Decision-Rights Standard: Sponsorship, Ownership and Escalation, adoption delivery under The Adoption and Capability Transfer Playbook: Making Change Hold in Daily Work, platform-option analysis under The Legacy Modernisation Decision: Retain, Rehost, Replatform, Refactor, Replace or Retire, partner appointment under The Transformation Partner Appointment and Commercial-Control Playbook and programme recovery under The Stalled Transformation Recovery and Value-Realisation Standard: Recover, Stop or Scale. Automation architecture and build/buy/no-build choices remain The Automation Architecture Decision: Rules, Platforms, Agents, Custom Software or No Build.

Dependencies are an authorised sponsor, an operating owner, evidence access, affected-person participation, finance involvement where economic claims matter, technology and data owners, qualified privacy/security/legal/employment authorities, and capacity to act on the result.

Expected results are a scope charter, evidence register, classification matrix, component map, option comparison, authorised disposition, next-gate brief, measures and review record. They are not a guaranteed saving, roadmap, business case, solution design or delivery promise.

Official GOV.UK guidance advises teams to understand and reframe the problem before committing to a build (GOV.UK Service Manual: discovery, VERIFIED official source at the source cut-off). Its service context does not make it a universal corporate mandate; this guide uses the narrower decision discipline. GOV.UK user-research guidance also calls for understanding users, their current behaviour and end-to-end context before planning a service (GOV.UK Service Manual: user research in discovery, VERIFIED official source at the source cut-off).

2. The Leadership vs. Delegation Boundary

The seven scope paths cross functional budgets, evidence and ownership; classification therefore needs cross-functional authority, controlled access and independent challenge, whether led internally or externally.

When an internal team can lead

A capable internal team can lead when it can challenge programme labels and preferred solutions, observe work across functions, reconcile technology and operating evidence, involve affected people safely, and place a cross-functional decision before an authorised forum. The team needs an independent challenger who is not rewarded for protecting the proposed scope.

When an independent senior partner is justified

Independent leadership is proportionate when several budgets or suppliers are implicated, sponsors use incompatible definitions, sunk cost shapes the recommendation, the proposed intervention crosses operating and technology boundaries, or no internal function can challenge all source owners. Independence does not transfer business authority.

When a narrower specialist is sufficient

Use a process specialist when one bounded workflow is accepted and the question is how to improve it. Use an automation specialist after the rule, exception and control boundary is stable. Use a technology or architecture specialist when the business outcome and target operation are settled and the remaining decision is technical. Use qualified security, privacy, legal, employment or financial specialists for conclusions within their mandates. A simpler repair is preferable when it passes the same outcome and guardrail gates.

Stop and escalation routes

Pause for unauthorised evidence, unsafe participation, uncontained material risk, unresolved conflicts of interest, or an unavailable operating owner. Escalate outcome and capital conflicts to the economic buyer; workflow authority to the COO; technology and data conditions to accountable owners; control conclusions to qualified authorities; and unresolved classification or supplier interest to the sponsor and independent challenger. Stop when evidence rejects materiality or no feasible operating owner exists.

Questions for evaluating a provider

1. How will you distinguish a local improvement from a cross-functional operating-model change?
2. What evidence would make you recommend no automation or no replacement?
3. How will you observe performed work and exceptions rather than rely on process documents?
4. How will you keep component treatments beneath one primary scope class?
5. Which claims require our finance, legal, privacy, security, employment or assurance owners?
6. How will you compare do-nothing, repair, improve, automate, modernise and transform options?
7. Which stable IDs, evidence records, decision rules and working files remain with us?
8. How will supplier incentives and sunk-cost interests be disclosed?
9. Where does this scope decision stop before architecture, business case or roadmap work begins?
10. What evidence would cause change, defer or stop?

Reject a provider who treats programme size, technology novelty or executive ambition as proof that transformation is required.

3. The Step-by-Step Execution Standard

Use the Scope Charter and Authority Register, Condition and Evidence Register, Scope Classification and Component Matrix, and Executive Scope Decision and Review Register. Store them in an approved controlled location.

Step 01: Charter the scope decision

Purpose and decision served Replace a broad programme label with one outcome, one affected population and one dated classification decision.

Accountable owner Executive sponsor; the economic buyer accepts the capital and scope boundary.

Delegated lead and contributors Scope lead, operating owner, customer/service owner, finance, technology, data, people and risk contributors.

Preconditions, inputs and access Decision forum, proposed initiative, known commitments, affected operation and authority to challenge the current label.

Actions

1. State the outcome without naming a solution.
2. Name the affected population, value-path break and operating consequence.
3. Set the decision date, evidence close, boundary and exclusions.
4. Record the seven available dispositions and explicit stop conditions.

Evidence to retain Approved charter, original proposal, decision reference, scope boundary and conflict declarations.

Decision or scoring rule Proceed only when one outcome and one classification decision are actionable; change an oversized question, defer missing authority or stop an ownerless question.

Output and template Charter rows in the Scope Charter and Authority Register using stable A4S-DEC-* and A4S-CHAR-* IDs.

Acceptance gate PROCEED only if sponsor, owner, population, outcome, dates, evidence threshold and stop route are accepted; otherwise CHANGE, DEFER or STOP.

Leadership checkpoint The sponsor confirms that no preferred platform, supplier or programme class is protected from challenge.

Human control point Only the authorised sponsor may approve the charter and evidence burden.

Failure modes and safeguards Solution-led wording and enterprise-wide scope are controlled by outcome-first language and explicit boundaries.

Change, stop or escalation condition Escalate incompatible sponsor and operating-owner outcomes; stop if no forum can decide.

Handoff to the next step Pass the accepted charter and conflicts to Step 02.

Step 02: Establish authority and evidence controls

Purpose and decision served Make access, decisions, controls and dissent explicit before investigation.

Accountable owner Executive sponsor, with information and control owners retaining their mandates.

Delegated lead and contributors Scope lead, evidence custodian, operating owner, privacy, security, legal, employment, finance and procurement representatives as applicable.

Preconditions, inputs and access Accepted charter, source list, participant groups, supplier interests and approved storage.

Actions

1. Map decision, evidence, access, risk, funding and operating authorities.
2. Record conflicts, recusals, source restrictions and escalation recipients.
3. Approve minimisation, retention, redaction and participant-protection rules.
4. Assign a challenger able to dispute scope and evidence.

Evidence to retain Authority matrix, approvals, access log, conflict register and control referrals.

Decision or scoring rule No evidence is collected without a lawful authorised purpose and custodian.

Output and template Authority rows in the Scope Charter and Authority Register.

Acceptance gate PROCEED when every material authority and escalation has an accepted owner; otherwise VALIDATE, DEFER or STOP.

Leadership checkpoint The sponsor accepts that control owners may narrow or halt the work.

Human control point Qualified authorities decide privacy, security, legal, employment and regulatory questions.

Failure modes and safeguards Informal exports and coerced participation are prevented through approved access and protected routes.

Change, stop or escalation condition Pause for unsafe participation, unlawful access or an unmanaged supplier conflict.

Handoff to the next step Pass authorised sources and handling rules to Step 03.

Step 03: Define the classification tests

Purpose and decision served Agree what evidence would distinguish each disposition before seeing preferred results.

Accountable owner Decision owner.

Delegated lead and contributors Scope lead, operating, technology, data, finance, customer, people and control contributors.

Preconditions, inputs and access Charter, authority map, existing definitions and adjacent-guide boundaries.

Actions

1. Convert each disposition into necessary evidence conditions.
2. Define work, rule stability, technical constraint and operating-model change tests.
3. Define materiality, feasibility, reversibility, dependency and guardrail tests.
4. Record disconfirming evidence and tie-breaking authority.

Evidence to retain Versioned classification rules, definitions, dissent and approval.

Decision or scoring rule No class wins by score alone; all necessary conditions for the selected class must pass and contradictions remain visible.

Output and template Test rows in the Scope Classification and Component Matrix.

Acceptance gate PROCEED when classes are mutually understandable, component-aware and falsifiable; otherwise CHANGE.

Leadership checkpoint Leaders accept that the smallest coherent class governs, not the largest ambition.

Human control point The decision owner approves definitions; an algorithm cannot assign scope.

Failure modes and safeguards Biased scoring is controlled by necessary-condition tests, raw evidence links and recorded dissent.

Change, stop or escalation condition Escalate if participants cannot agree what transformation or modernisation means for this decision.

Handoff to the next step Pass approved tests to Step 04.

Step 04: Build the evidence plan

Purpose and decision served Collect only evidence that can change the scope decision.

Accountable owner Decision owner and evidence custodian.

Delegated lead and contributors Evidence lead, source owners, researcher, workflow observer, analyst and control owners.

Preconditions, inputs and access Approved tests, available sources, known gaps and burden limits.

Actions

1. Map each test to source, population, period, method and owner.
2. Include customer/service evidence, performed work, exceptions, technology/data health, roles, incentives and decision rights.
3. Separate independent corroboration from duplicated reports.
4. Set sufficiency, quality, expiry and stop rules.

Evidence to retain Evidence request, source lineage, access approval and gap register.

Decision or scoring rule Collect evidence only when its possible results lead to different scope dispositions or guardrail decisions.

Output and template Planned rows in the Condition and Evidence Register.

Acceptance gate PROCEED when every material test has a proportionate source or visible gap; otherwise VALIDATE, CHANGE or STOP.

Leadership checkpoint The sponsor accepts the evidence burden and protects access.

Human control point Source and control owners approve access and interpretation boundaries.

Failure modes and safeguards Dashboard convenience and interview dominance are controlled by source diversity and performed-work observation.

Change, stop or escalation condition Narrow scope when evidence burden exceeds the decision consequence.

Handoff to the next step Pass the source plan and gap log to Step 05.

Step 05: Observe the market connection and performed work

Purpose and decision served Establish what affected people try to achieve and how work actually produces or obstructs that outcome.

Accountable owner Operating owner.

Delegated lead and contributors Researcher, workflow observer, frontline participants, customer/service owner and accessibility or inclusion contributors.

Preconditions, inputs and access Participant protection, approved research plan, workflow boundary and source controls.

Actions

1. Observe representative work, handoffs, queues, exceptions, rework and workarounds.
2. Reconcile observations with customer, service and operating evidence.
3. Identify where policy, ownership, information, capability or tools constrain the outcome.
4. Record variation and excluded populations without generalising beyond evidence.

Evidence to retain De-identified observations, journey/workflow trace, source records, limitations and evidence gaps.

Decision or scoring rule A performed-work claim requires direct evidence or explicit INFERRED status and validation.

Output and template Observation and condition rows in the Condition and Evidence Register.

Acceptance gate PROCEED when material work and affected-person conditions are evidenced within bounds; otherwise VALIDATE or CHANGE.

Leadership checkpoint The operating owner accepts observed exceptions, including evidence that challenges policy or reported process.

Human control point Researchers and participant authorities protect consent, confidentiality, accessibility and non-retaliation.

Failure modes and safeguards Workshop-only maps and convenient-user bias are controlled by observation, varied participants and limits.

Change, stop or escalation condition Stop unsafe research; escalate systemic harm or material control breaches through qualified routes.

Handoff to the next step Pass accepted work conditions to Step 06.

Step 06: Test rule and automation fitness

Purpose and decision served Determine whether a stable bounded rule can be automated without preserving a broken process.

Accountable owner Operating owner, with the accountable control owner.

Delegated lead and contributors Process analyst, automation specialist, frontline users, data, technology, security and privacy contributors.

Preconditions, inputs and access Observed workflow, exception evidence, rule owner, input/output definitions and control constraints.

Actions

1. Test rule stability, frequency, volume, variation and exception ownership.
2. Identify judgement points and mandatory human decisions.
3. Test input quality, auditability, fallback, reversibility and support.
4. Compare automation with policy removal, simplification, conventional improvement and no action.

Evidence to retain Rule catalogue, exception sample, input/output lineage, control decision and alternative comparison.

Decision or scoring rule Classify AUTOMATE only if the rule and operating context are accepted; otherwise improve, validate, modernise, transform, defer or stop.

Output and template Automation-test and component rows in the Scope Classification and Component Matrix.

Acceptance gate PASS when the rule, exceptions, controls, owner and fallback are explicit; otherwise FAIL or VALIDATE.

Leadership checkpoint The sponsor confirms automation is not being used to avoid policy or ownership decisions.

Human control point Named humans retain consequential exception and risk decisions.

Failure modes and safeguards Automating waste and hidden exception labour are controlled by alternatives and full-work observation.

Change, stop or escalation condition Escalate material security, privacy, employment or customer harm; stop unsafe automation.

Handoff to the next step Pass rule fitness and gaps to Step 07.

Step 07: Test technology modernisation necessity

Purpose and decision served Determine whether technology, data, integration or supportability is a material constraint rather than a convenient explanation.

Accountable owner CIO with the operating owner jointly accepting business consequence.

Delegated lead and contributors Architecture, engineering, service management, data, security, finance, procurement and vendor contributors.

Preconditions, inputs and access Accepted operating outcome, service evidence, asset and dependency records, incidents, costs and contractual limits.

Actions

1. Link each technical condition to an observable service or operating consequence.
2. Test reliability, maintainability, interoperability, data quality, security, support and exitability.
3. Compare containment, repair, reconfiguration, rehost, replatform, refactor, replace and retire as later The Legacy Modernisation Decision: Retain, Rehost, Replatform, Refactor, Replace or Retire work where material.
4. Separate technical renewal from operating-model change.

Evidence to retain Condition records, service evidence, dependencies, control findings, cost definitions and uncertainty.

Decision or scoring rule Classify MODERNISE only when a material technical constraint is evidenced and the target operation is substantially known.

Output and template Technology-condition and component rows in the two evidence and classification tools.

Acceptance gate PASS when consequence, constraint, owner and next technical decision are traceable; otherwise VALIDATE or reject modernisation as primary class.

Leadership checkpoint The CIO and COO agree whether technology is cause, dependency or component.

Human control point Qualified technology and control owners accept technical interpretations; vendor claims are not self-validating.

Failure modes and safeguards Age-based replacement and vendor preference are controlled by consequence evidence and alternatives.

Change, stop or escalation condition Escalate material resilience or security exposure immediately; do not wait for the scope exercise.

Handoff to the next step Pass technical conditions to Step 08.

Step 08: Test operating-model transformation necessity

Purpose and decision served Determine whether the outcome requires coordinated changes across functions and operating conditions.

Accountable owner CEO or COO as authorised decision owner.

Delegated lead and contributors Scope lead, operating-model specialist, customer/service, workflow, finance, people, data, technology and governance owners.

Preconditions, inputs and access Work, rule and technology findings; role, incentive, forum, capacity and dependency evidence.

Actions

1. Test whether workflow, decision rights, structure, measures, incentives, capability, data and technology must change together.
2. Identify cross-functional dependencies and local optimisations that would preserve the failure.
3. Test smaller improve, automate and modernise routes against the same outcome and guardrails.
4. Record transformation components without treating size as evidence.

Evidence to retain Operating-condition map, dependency links, rejected smaller options, dissent and uncertainty.

Decision or scoring rule Classify TRANSFORM only when no smaller coherent class can achieve the accepted outcome within guardrails.

Output and template Operating-model tests and component relationships in the Scope Classification and Component Matrix.

Acceptance gate PASS when multiple necessary operating changes and dependencies are evidenced; otherwise select a smaller class or validate.

Leadership checkpoint The economic buyer accepts the coordination and capacity implications without yet approving a portfolio or roadmap.

Human control point Only leadership can accept cross-functional authority, risk and resource consequences.

Failure modes and safeguards Programme prestige and comprehensive wish lists are controlled by necessary conditions and smallest-coherent-scope logic.

Change, stop or escalation condition Defer when required authority or capacity is unavailable; stop when no operating owner can hold the target state.

Handoff to the next step Pass all classification evidence to Step 09.

Step 09: Reconcile claims and uncertainty

Purpose and decision served Create one traceable evidence view without flattening contradiction into a score.

Accountable owner Decision owner.

Delegated lead and contributors Evidence lead, analyst, source owners, independent challenger and qualified reviewers.

Preconditions, inputs and access Condition records, classifications, source lineage, gaps, conflicts and limitations.

Actions

1. Assign VERIFIED, INFERRED, PROPOSED or REQUIRES VALIDATION to each decision claim.
2. Separate observed fact, interpretation, statement, assumption, gap and constraint.
3. Test source independence, population, period and alternative explanations.
4. Record unresolved dissent and the evidence needed to settle it.

Evidence to retain Claim-evidence links, proof status, reconciliation notes, dissent and validation tasks.

Decision or scoring rule No material scope condition may be treated as verified without fit evidence; unresolved critical claims produce VALIDATE.

Output and template Reconciled rows in the Condition and Evidence Register.

Acceptance gate PROCEED when the decision can be made within stated uncertainty; otherwise VALIDATE, DEFER or STOP.

Leadership checkpoint Leaders see contradictions and gaps before option comparison.

Human control point The proof owner accepts classification; uncertainty is not removed by arithmetic.

Failure modes and safeguards Consensus theatre and duplicated evidence are controlled by stable sources, dissent and explicit proof states.

Change, stop or escalation condition Escalate source-owner disputes; narrow any claim beyond its evidence boundary.

Handoff to the next step Pass decision-fit claims to Step 10.

Step 10: Compare the smallest coherent options

Purpose and decision served Compare viable scope choices using common outcome, burden, risk and dependency evidence.

Accountable owner Economic buyer.

Delegated lead and contributors Scope lead, finance, operating, technology, data, people, risk, procurement and affected-owner contributors.

Preconditions, inputs and access Reconciled findings, class tests, current commitments, constraints and baseline status.

Actions

1. Include do nothing, stop, direct repair, improve, automate, modernise and transform where credible.
2. Define component treatments beneath each primary class.
3. Compare expected outcome, evidence confidence, burden, capacity, dependencies, reversibility, risk and opportunity cost.
4. Specify the next evidence or investment gate for each viable option.

Evidence to retain Option definitions, comparison logic, assumptions, rejected alternatives and conflicts.

Decision or scoring rule Prefer the smallest coherent option that meets the outcome and guardrails; a weighted score informs but never authorises.

Output and template Option and classification rows in the Scope Classification and Component Matrix.

Acceptance gate ACCEPT options only when evidence, dependencies, owner, burden and stop logic are visible.

Leadership checkpoint The economic buyer challenges sunk cost, supplier preference and false comprehensiveness.

Human control point Leadership judges trade-offs and risk appetite; the tool does not recommend automatically.

Failure modes and safeguards False precision and hidden dependencies are controlled by ranges, confidence, narrative evidence and component IDs.

Change, stop or escalation condition Return to validation when options depend on unresolved material claims; stop options without owners.

Handoff to the next step Pass viable options and dissent to Step 11.

Step 11: Make the executive scope decision

Purpose and decision served Authorise one primary scope class and bounded next gate.

Accountable owner Named executive decision owner.

Delegated lead and contributors Decision secretary, sponsor, operating owner, finance, CIO, benefit owner, control owners and independent challenger.

Preconditions, inputs and access Decision pack, source links, option comparison, dissent, conflicts and forum authority.

Actions

1. Review outcome, evidence limits, alternatives, burden and guardrails.
2. Select improve, automate, modernise, transform, validate, defer or stop.
3. Name component treatments, resource commitment, owner and next gate separately.
4. Record dissent, conditions, expiry and decision rationale.

Evidence to retain Signed decision, attendance, authority, rationale, conditions, dissent and linked records.

Decision or scoring rule One primary disposition per decision; no “hybrid”, “proceed” or “fund” without a defined class and next gate.

Output and template Decision rows in the Executive Scope Decision and Review Register.

Acceptance gate ACCEPTED only when class, rationale, owner, conditions, resource boundary and dated next gate are complete.

Leadership checkpoint The economic buyer confirms what has and has not been authorised.

Human control point Only the mandated forum accepts the disposition and risk.

Failure modes and safeguards Ambiguous conditional approval is controlled by one enum, explicit conditions and a decision record.

Change, stop or escalation condition Defer for absent quorum or authority; stop if material conflicts cannot be mitigated.

Handoff to the next step Pass the accepted decision to Step 12.

Step 12: Handover the bounded next gate

Purpose and decision served Translate classification into owned downstream work without implying the downstream method is complete.

Accountable owner Operating owner, with the sponsor retaining benefit accountability.

Delegated lead and contributors Handover lead, downstream owner, delivery, architecture, process, automation, finance, people and control contributors as selected.

Preconditions, inputs and access Accepted decision, conditions, component map, retained evidence and approved resource boundary.

Actions

1. Define the next decision, inputs, outputs, acceptance gate and exclusions.
2. Assign stable record ownership and approved access.
3. Route detailed work to the applicable A2 or A4 guide or narrower specialist.
4. Confirm capability transfer, support, exception and escalation expectations.

Evidence to retain Handover acceptance, next-gate brief, owner, controls, access transfer and unresolved gaps.

Decision or scoring rule Handover passes only when the receiving owner can explain the class, outcome, conditions and evidence limits.

Output and template Handover rows in the Executive Scope Decision and Review Register.

Acceptance gate HANDOVER ACCEPTED or RETURN FOR CHANGE; no silent transfer.

Leadership checkpoint The sponsor confirms that classification has not been expanded into unauthorised delivery.

Human control point The receiving owner accepts operational accountability and qualified controls remain with their owners.

Failure modes and safeguards Scope inflation and evidence loss are controlled by exclusions, stable IDs and receipt acceptance.

Change, stop or escalation condition Return when the recipient lacks authority or capacity; defer rather than create an ownerless programme.

Handoff to the next step Pass accepted ownership and measure definitions to Step 13.

Step 13: Review scope fitness in operation

Purpose and decision served Test whether evidence still supports the class as new information and operating results emerge.

Accountable owner Benefit owner and operating owner.

Delegated lead and contributors Measurement owner, downstream lead, finance, customer/service, technology, data and control owners.

Preconditions, inputs and access Accepted baseline, measures, source access, review cadence, guardrails and decision authority.

Actions

1. Review learning, delivery, operating, outcome and guardrail measures separately.
2. Compare actual component needs with the accepted primary class.
3. Test new evidence, exceptions, dependencies and unintended effects.
4. Decide continue, change class, validate, defer, stop or commission a new charter.

Evidence to retain Measure extracts, review record, exceptions, new evidence, decision and retired access.

Decision or scoring rule Change the class when a necessary condition no longer holds; do not protect the original label.

Output and template Review and measure rows in the Executive Scope Decision and Review Register.

Acceptance gate Every review ends CONTINUE, CHANGE, VALIDATE, DEFER, STOP or NEW CHARTER with owner and date.

Leadership checkpoint The sponsor reviews whether the next gate remains proportionate to the outcome and evidence.

Human control point Named owners interpret measures, harms and attribution before changing scope.

Failure modes and safeguards Label persistence and delivery-only reporting are controlled by scope-condition and outcome review.

Change, stop or escalation condition Escalate guardrail breach immediately; stop activity whose risk or burden exceeds accepted bounds.

Handoff to the next step Close or renew the controlled cycle; materially changed outcomes require a new charter at Step 01.

4. The Evidence & Measurement Checklist

Preflight and authority

• One bounded outcome and affected population are named without prescribing a solution. Evidence: accepted scope charter. Owner: executive sponsor.
• The dated decision uses improve, automate, modernise, transform, validate, defer or stop. Evidence: decision-question row. Owner: decision owner.
• Economic buyer, operating owner and benefit owner are named. Evidence: authority register. Owner: executive sponsor.
• Decision, access, control, funding and escalation mandates are explicit. Evidence: authority matrix. Owner: decision owner.
• Supplier interests, sunk-cost interests and recusals are declared. Evidence: conflict register. Owner: independent challenger.

Evidence access and quality

• Every material claim links to one retained evidence row. Evidence: condition and evidence register. Owner: evidence custodian.
• Population, period, source, method and limitation are recorded. Evidence: source-lineage fields. Owner: source owner.
• Performed work and material exceptions have been observed or marked for validation. Evidence: observation records. Owner: operating owner.
• Duplicated reports are not counted as independent corroboration. Evidence: source-dependency review. Owner: evidence lead.
• Missing or disputed evidence is visible with an owner and decision date. Evidence: evidence-gap rows. Owner: decision owner.

Method execution

• Each disposition has approved necessary-condition tests. Evidence: classification matrix. Owner: decision owner.
• Improvement has been tested before a larger scope is accepted. Evidence: option comparison. Owner: operating owner.
• Automation fitness includes rule, exception, control and fallback evidence. Evidence: automation-test rows. Owner: operating owner.
• Modernisation claims link technical condition to business or service consequence. Evidence: technology-condition rows. Owner: CIO.
• Transformation requires multiple necessary operating changes and rejection of smaller coherent routes. Evidence: operating-model tests. Owner: economic buyer.

Governance, privacy, security and risk

• Evidence access and retention have approved purposes and custodians. Evidence: access approvals. Owner: information authority.
• Customer and employee participation has appropriate protection. Evidence: research-control approval. Owner: qualified control owner.
• Privacy, security, legal and employment conclusions remain with qualified authorities. Evidence: specialist referral log. Owner: executive sponsor.
• Material incidents use immediate containment routes outside this method. Evidence: incident escalation record. Owner: risk executive.
• Human authority is retained for consequential exceptions and final classification. Evidence: decision-rights record. Owner: decision owner.

Deliverables and acceptance

• The charter, evidence register, classification matrix and decision register are complete. Evidence: working-pack index. Owner: scope lead.
• One primary scope class governs each decision. Evidence: accepted decision row. Owner: decision owner.
• Component treatments are linked beneath the primary class. Evidence: component matrix. Owner: operating owner.
• Rejected alternatives retain rationale and evidence. Evidence: option rows. Owner: independent challenger.
• The next gate states inputs, outputs, exclusions, owner and acceptance test. Evidence: handover brief. Owner: downstream owner.

Measurement

• Learning, delivery, operating, outcome and guardrail measures remain separate. Evidence: measurement register. Owner: benefit owner.
• Every measure has a stable definition, source, baseline and owner. Evidence: measure-contract rows. Owner: measurement owner.
• Missing baselines say baseline required rather than zero. Evidence: baseline status. Owner: finance and source owner.
• Thresholds specify the decision response rather than an invented target. Evidence: threshold fields. Owner: decision owner.
• Attribution limits and time lags are visible. Evidence: measure limitations. Owner: benefit owner.

Decision and handover

• The authorised forum, quorum and decision date are retained. Evidence: decision record. Owner: decision secretary.
• Funding and resource commitment are separate from the scope disposition. Evidence: resource-boundary row. Owner: economic buyer.
• Conditions, dissent and conflicts accompany the decision. Evidence: signed decision pack. Owner: decision owner.
• The receiving owner has accepted the evidence and exclusions. Evidence: handover acceptance. Owner: operating owner.
• Access, custody and unresolved gaps transfer or close explicitly. Evidence: custody log. Owner: evidence custodian.

Review, change and stop controls

• Each scope condition has a review cadence or event trigger. Evidence: review schedule. Owner: benefit owner.
• New evidence can change the class without protecting sunk cost. Evidence: review decision. Owner: economic buyer.
• Guardrail breaches have immediate escalation and stop routes. Evidence: control-response record. Owner: risk executive.
• Obsolete claims, access and working copies are retired under policy. Evidence: closure log. Owner: evidence custodian.
• Materially changed outcomes trigger a new charter. Evidence: new-charter decision. Owner: executive sponsor.

Measurement contracts

Measure A: Classification evidence coverage

Measure B: Next-gate acceptance

Measure C: Scope-condition stability

Measure D: Accepted operating outcome movement

Measure E: Scope-related harm and control exceptions

Measure F: Avoidable scope expansion

Evidence classifications used in this guide

• VERIFIED: directly supported by retained, evidence relevant to the decision or an identified authoritative source within its stated boundary.
• INFERRED: a reasoned interpretation linked to evidence, with alternative explanations and limits visible.
• PROPOSED: a working definition, option, classification or action awaiting authorised acceptance.
• REQUIRES VALIDATION: a material claim, baseline, consequence or condition that cannot yet support the decision.

Engagement records additionally distinguish observed fact, corroborated interpretation, single-source statement, assumption, evidence gap and constraint. The absence of evidence is not proof that a condition is absent.

Sources and evidence notes

1. GOV.UK Service Manual: How the discovery phase works supports problem definition, solution challenge and learning before build. VERIFIED official source at the source cut-off; its public-service context limits direct transfer.
2. GOV.UK Service Manual: User research in discovery supports end-to-end user context and research before design. VERIFIED official source at the source cut-off; it does not define corporate scope classes.
3. GOV.UK Service Manual: Measuring service benefits supports comparing options and connecting discovery evidence to benefits. VERIFIED official source at the source cut-off; benefits still require local definitions and baselines.
4. NIST Cybersecurity Framework 2.0 frames cybersecurity outcomes around governance and risk management. VERIFIED official source at the source cut-off; qualified owners must interpret applicability.
5. NIST Privacy Framework supports managing privacy risk arising from data processing. VERIFIED official source at the source cut-off; it is not legal advice or a certification.

The seven-disposition model, smallest-coherent-scope rule, component treatment model, steps, checklists and visual structures are Abrasim-created working methods. They are PROPOSED, not external proof or a promised result.