Playbooks Level: Practitioner
Delivery and change
How to make the DAM Policy stick: in the yearly business-as-usual cycle, and inside the programmes and projects that create, change and retire data assets.
In brief
- Most critical data assets are created or replaced by change programmes, so controls belong in them.
- Each item lists actions, evidence, roles and the rules behind it.
- Tick actions as you go; progress stays in your browser.
What the policy leaves to you
The policy states its requirements briefly. These capabilities are needed to meet it but are only implied, or not stated at all. They are where organisations most often need help.
| ID | Capability | In the policy | Portal coverage | Practices |
|---|---|---|---|---|
| CR-02 | Data strategy and prioritisation by reuse value | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-05 |
| CR-09 | Data quality issue management | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | See the capability model |
| CR-12 | Data architecture and data services | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | See the capability model |
| CR-13 | Catalogue and quality tooling | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | See the capability model |
| CR-14 | Reference data, master data and authoritative sources | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Partial | See the capability model |
| CR-15 | Records, retention and disposal | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-04 |
| CR-20 | Third-party and supplier data | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-06 CH-16 |
| CR-21 | Data skills, capability and literacy | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-09 |
| CR-22 | Data culture and engagement with users | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-09 |
| CR-23 | Maturity assessment and continuous improvement | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-12 |
| CR-25 | Integration with programmes, spend control and assurance | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-02 CH-03 |
| CR-26 | Legacy migration and decommissioning | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-04 CH-05 CH-14 |
| CR-27 | Ownership transition and handover to business as usual | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-08 |
| CR-29 | Benefits realisation and evaluation | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-10 |
| CR-30 | Equality and inclusion in data | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-05 |
| CR-31 | Cyber resilience and incident response for data assets | Implicit: The text depends on it to meet an explicit requirement but does not name it as one. | Covered | CH-15 |
| CR-32 | Avoiding future legacy: open standards, portability and exit | Inferred: Not in the text. This portal's judgement from good practice and adjacent rulebooks. | Covered | CH-14 CH-16 |
DAM across the delivery life cycle
Stages follow the GOV.UK Service Manual phases and the project delivery standard. Use it as an overlay on your own programme plan.
Strategy and business case
- List the critical data assets the change will create, change, depend on or retire.
- Cost ownership, cataloguing and quality work in the economic and financial cases.
- Add data asset risks to the programme risk register.
Discovery
- Check the catalogue for assets that already meet the need before collecting new data.
- Agree who will own each new or changed asset.
Alpha
- Design metadata and quality rules alongside the data model.
- Run a data protection impact assessment where personal data is involved.
Beta and build
- Catalogue new assets before go-live, with the core metadata.
- Write and test the DQAP; reconcile migrated data against the source.
Live and business as usual
- Hand ownership, the DQAP and the runbook to named people in business as usual.
- Report the asset in the next annual return.
Migrate or retire
- Start from the asset register; decide to migrate, archive or dispose of each asset.
- Update or close catalogue entries and transfer ownership formally.
Programme roles with data responsibilities
Medium and large programmes bring in roles the DAM Policy and its guidance do not name. They decide most of what happens to critical data during change, so each needs explicit data responsibilities. This is this portal's assessment; the policy names only the chief data officer, data owners and the Chief Data Officer Council.O: This playbook's own synthesis of published guidance
| Role | Data responsibilities | Stages | Works with | Rules |
|---|---|---|---|---|
| Senior responsible owner |
| Strategy and business case, Live and business as usual | Data owner Chief data officer | Role of the senior responsible ownerV: Verified against the primary source on 2026-11-05; Implementing integrated assurance for major projectsV: Verified against the primary source on 2026-11-05 |
| Programme or project manager |
| Strategy and business case, Discovery, Beta and build, Live and business as usual | Data owner Lead policy coordinator Data steward | Government Functional Standard GovS 002: Project DeliveryV: Verified against the primary source on 2026-11-05 |
| Service or product owner |
| Discovery, Alpha, Beta and build, Live and business as usual | Data owner Catalogue manager Data quality lead or analyst | What each role does in a service teamV: Verified against the primary source on 2026-11-05; Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05 |
| Design authority or solution architect |
| Alpha, Beta and build | Data custodian Catalogue manager Data steward | The Technology Code of PracticeV: Verified against the primary source on 2026-11-05; API technical and data standardsV: Verified against the primary source on 2026-11-05; Open standards for governmentV: Verified against the primary source on 2026-11-05 |
| Data migration lead |
| Beta and build, Migrate or retire | Data owner Data quality lead or analyst Catalogue manager Data custodian | Managing legacy technologyV: Verified against the primary source on 2026-11-05; Retiring your serviceV: Verified against the primary source on 2026-11-05 |
| Commercial lead |
| Strategy and business case, Discovery, Migrate or retire | Data owner Data protection officer Lead policy coordinator | Model Services ContractV: Verified against the primary source on 2026-11-05; The Sourcing PlaybookV: Verified against the primary source on 2026-11-05; PPN 020: Guidance on data protection legislationV: Verified against the primary source on 2026-11-05 |
| Programme office and assurance |
| Strategy and business case, Beta and build, Live and business as usual | Lead policy coordinator Data governance board | Implementing integrated assurance for major projectsV: Verified against the primary source on 2026-11-05; National Infrastructure and Service Transformation AuthorityV: Verified against the primary source on 2026-11-05 |
These roles also appear on the roles list, the programme or project lead route and in the practices below.
Service blueprint for a programme
What happens to critical data at each stage, front stage and back stageO: This playbook's own synthesis of published guidance.
| Lane | Strategy and business case | Discovery | Alpha | Beta and build | Live and business as usual | Migrate or retire |
|---|---|---|---|---|---|---|
| What the programme does | Programme board agrees scope and business case | Team explores needs and existing data | Team designs and tests options | Team builds, migrates and goes live | Service runs; programme closes | Old systems are switched off |
| What users and boards see | Business case names critical assets and costs DAM work | Catalogue search before new collection | Data model and quality rules designed together | New assets catalogued; validation built from rules | Quality published to users | Users told what happens to their data |
| Behind the scenes | Asset list drawn from the catalogue; risks added to the register | Contracts and data sharing options reviewed | Schemas, identifiers and exit route approved by design authority | Migration profiled and reconciled; lineage recorded | DQAP, runbook and contracts handed over | Assets migrated, archived or disposed of under records rules |
| Data roles | Data owners and the chief data officer agree priorities | Owners confirmed for each asset in scope | Stewards agree definitions and targets | Owners sign off quality against targets | Business-as-usual owner and steward named | Catalogue entries updated or closed |
| Support and assurance | Assurance plan includes DAM checkpoints | Commercial lead brings data clauses into the route to market | DPIA started where personal data is involved | Gate review checks DAM evidence | Asset included in the next annual return | Records transferred where the Public Records Act applies |
Practices
-
Not started
CH-01 · Change programmes · Strategy and business case
Put data assets in the business case
The policy says data must be valued and invested in; business cases are where that money is decided.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Name each critical data asset in scope of the change in the strategic case.
- Cost owner, steward and catalogue effort and DQAP remediation in the financial case.
- Count avoided duplicate collection and cleansing as benefits in the economic case.
- Name the data owner in the management case.
Evidence
- Asset list in the business case
- Costed DAM activities
- Named owners
Rulebook
- The Green Book and accompanying guidanceV: Verified against the primary source on 2026-11-05
- The AQuA BookV: Verified against the primary source on 2026-11-05
-
Not started
CH-02 · Change programmes · Alpha
Use spend control and service assessments as checkpoints
Spend control and service assessments already stop work that is not ready; adding DAM evidence costs little.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Add three questions to internal spend control: which assets, who owns them, are they catalogued.
- At alpha, show the planned metadata and quality rules; at beta, the catalogue entry and DQAP.
- Show the panel that data was considered: who owns each asset, how it is described and quality-planned, the DPIA, and that data professionals and users were involved. You do not need to hand in every template; keep them as your team's working record.
- Record the answers so the next return reuses them.
Evidence
- Spend control submission notes
- Assessment evidence pack
Rulebook
- Set up a commercial or digital and technology spend controls pipelineV: Verified against the primary source on 2026-11-05
- Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05
- The Technology Code of PracticeV: Verified against the primary source on 2026-11-05
- Service StandardV: Verified against the primary source on 2026-11-05
-
Not started
CH-03 · Change programmes · Strategy and business case
Write DAM into programme assurance
Large programmes have integrated assurance plans; DAM checks in them are reviewed with everything else.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Add DAM checkpoints to the integrated assurance and approval plan.
- Ask reviewers to check asset ownership, catalogue entries and quality plans at each gate.
- Give the senior responsible owner a single line on data asset readiness in highlight reports.
Evidence
- Assurance plan entries
- Gate review findings
Rulebook
- Implementing integrated assurance for major projectsV: Verified against the primary source on 2026-11-05
- National Infrastructure and Service Transformation AuthorityV: Verified against the primary source on 2026-11-05
- Government Major Projects PortfolioV: Verified against the primary source on 2026-11-05
- Government Functional Standard GovS 002: Project DeliveryV: Verified against the primary source on 2026-11-05
-
Not started
CH-04 · Change programmes · Migrate or retire
Plan legacy exits from the asset register
Legacy systems hold large concentrations of hard-to-reach data; the legacy guidance asks for an asset register first.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Build or update the register of data assets held in the legacy system.
- Profile quality before migration and set reconciliation rules for counts, totals and keys.
- Decide for each asset: migrate, archive under records rules, or dispose.
- Keep the catalogue and ownership in step as each asset moves.
Evidence
- Asset register for the system
- Profiling and reconciliation results
- Disposition decisions
Rulebook
- Managing legacy technologyV: Verified against the primary source on 2026-11-05
- Prevent technical debt and legacyV: Verified against the primary source on 2026-11-05
- Retiring your serviceV: Verified against the primary source on 2026-11-05
- Information management guidanceV: Verified against the primary source on 2026-11-05
-
Not started
CH-05 · Business as usual and change · Strategy and business case
Target legacy where it blocks outcomes and widens inequalities
Where poor or locked-in data delays access to services, the people most affected are often those already disadvantaged.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Score each critical asset on legacy risk, quality, reuse value and impact on outcomes.
- Check whether gaps or errors fall harder on some groups, using the equality duty.
- Fund fixes first where risk, value and inequality overlap.
- Report the choice and its reasons to the data governance board.
Evidence
- Prioritised asset heatmap
- Equality analysis
Rulebook
- Public sector equality dutyV: Verified against the primary source on 2026-11-05
- Managing legacy technologyV: Verified against the primary source on 2026-11-05
- Data Maturity Assessment for Government: FrameworkV: Verified against the primary source on 2026-11-05
-
Not started
CH-06 · Business as usual and change · Discovery
Build DAM into contracts
Suppliers often hold or process critical assets; without contract terms the owner cannot meet the policy.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- State who owns the data and its metadata.
- Require metadata in an agreed format and quality reporting against the DQAP.
- Set data return, format and deletion terms for exit.
- Include data protection terms from the procurement policy note.
Evidence
- Contract schedule for data
- Exit plan
Rulebook
- The Sourcing PlaybookV: Verified against the primary source on 2026-11-05
- The Digital, Data and Technology PlaybookV: Verified against the primary source on 2026-11-05
- PPN 020: Guidance on data protection legislationV: Verified against the primary source on 2026-11-05
-
Not started
CH-07 · Business as usual and change · Live and business as usual
Manage data asset risk on the same registers
The policy relies on risk controls; separate data risk lists are easy to ignore.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Record risks to critical assets on the corporate or programme register.
- Link each to the asset, its owner and the senior information risk owner.
- Review alongside DQAP results each quarter.
Evidence
- Risk register entries
- Quarterly review notes
Rulebook
- The Orange Book: management of riskV: Verified against the primary source on 2026-11-05
- Government Security ClassificationsV: Verified against the primary source on 2026-11-05
-
Not started
CH-08 · Change programmes · Live and business as usual
Hand over to business as usual
Assets created by projects lose their owner when the project closes unless handover is planned.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Confirm a named business-as-usual owner and steward before the project closes.
- Hand over the DQAP, quality rules, runbook and open issues.
- Use the ownership model's transition steps and set the next review date.
Evidence
- Signed handover note
- Updated catalogue entry
Rulebook
- Government Functional Standard GovS 002: Project DeliveryV: Verified against the primary source on 2026-11-05
-
Not started
CH-09 · Business as usual and change · Strategy and business case
Plan people change, not just process change
New duties fail when people are not told, trained, given time or recognised for them.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Assess who is affected: owners, stewards, custodians, analysts, suppliers.
- Put data duties into role descriptions and objectives.
- Train each role at the right level and give stewards protected time.
- Communicate why the policy matters in terms of services and people, not compliance.
Evidence
- Change impact assessment
- Training plan
- Communications plan
Rulebook
- Success ProfilesV: Verified against the primary source on 2026-11-05
- Data skills for allV: Verified against the primary source on 2026-11-05
- Government Digital and Data Profession Capability FrameworkV: Verified against the primary source on 2026-11-05
-
Not started
CH-10 · Business as usual and change · Live and business as usual
Show the benefits
Continued funding depends on evidence that the work changed something.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Agree a few measures before starting, such as reuse requests met, duplicate collections stopped or time spent cleansing.
- Take a baseline, then measure each year with the return.
- Use proportionate evaluation methods.
Evidence
- Benefits map
- Baseline and annual measures
Rulebook
- The Magenta BookV: Verified against the primary source on 2026-11-05
-
Not started
CH-11 · Business as usual · Live and business as usual
Keep the yearly cycle running in business as usual
The policy is annual; without a calendar it becomes a scramble before each return.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Set fixed dates for asset review, metadata updates, DQAP review and the board.
- Track each owner's actions between returns.
- Feed lessons into the next year's priorities.
Evidence
- Annual calendar
- Action log
-
Not started
CH-12 · Business as usual · Strategy and business case
Assess maturity and set a target
A shared maturity measure turns the policy into a plan that can be resourced and tracked.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Run the government data maturity assessment or use the self-assessment here as a first pass.
- Set a target level for each DAM strand with dates.
- Repeat every one to two years.
Evidence
- Assessment results
- Target levels
Rulebook
- Data Maturity Assessment for Government: FrameworkV: Verified against the primary source on 2026-11-05
- Government data maturity hubV: Verified against the primary source on 2026-11-05
-
Not started
CH-13 · Business as usual and change · Discovery
Share and reuse safely across organisations
Reuse is the point of the policy, but it never removes legal duties or forces sharing.O: This playbook's own synthesis of published guidance
Actions done0 of 3 doneActions
- Check the catalogue for reusable assets before collecting new data.
- Use the sharing framework and code for agreements.
- Record sharing arrangements against the asset in the catalogue.
Evidence
- Data sharing agreements
- Catalogue links
Rulebook
- Data Sharing Governance FrameworkV: Verified against the primary source on 2026-11-05
- ICO data sharing code of practiceS: Found by web search; the site blocked automated checking, so open it before relying on it
- Digital Economy Act 2017V: Verified against the primary source on 2026-11-05
-
Not started
CH-14 · Change programmes · Alpha
Design out future legacy and lock-in
Systems bought or built today become tomorrow's legacy unless their data can be moved out without loss or delay.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Require data in open, documented formats with a published schema.
- Ask suppliers to show how a full export with metadata works before you sign.
- Prefer open standards and APIs that follow the government API standards.
- Record each system's exit route against its assets in the catalogue.
Evidence
- Exit and portability test results
- Schema and API documentation
Rulebook
- Open standards for governmentV: Verified against the primary source on 2026-11-05
- API technical and data standardsV: Verified against the primary source on 2026-11-05
- Choosing technology: an introductionV: Verified against the primary source on 2026-11-05
- Prevent technical debt and legacyV: Verified against the primary source on 2026-11-05
-
Not started
CH-15 · Business as usual · Live and business as usual
Prepare for a cyber incident affecting a critical asset
A ransomware attack or breach can corrupt or expose a critical asset; quality plans and owners must be ready.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Mark critical data assets in the incident response plan and asset inventory.
- Agree recovery point and recovery time with each data owner.
- Test restoring an asset and check its quality afterwards against the DQAP rules.
- Agree who tells data users when an asset may be affected.
Evidence
- Incident plan entries
- Restore test and post-restore quality results
Rulebook
- Incident managementV: Verified against the primary source on 2026-11-05
- Cyber Assessment Framework (CAF)V: Verified against the primary source on 2026-11-05
- 10 Steps to Cyber SecurityV: Verified against the primary source on 2026-11-05
- Government Functional Standard GovS 007: SecurityV: Verified against the primary source on 2026-11-05
-
Not started
CH-16 · Change programmes · Discovery
Use the buying routes and contract terms that protect data
Most catalogue, quality and platform tools are bought; the route and terms decide who controls the data.O: This playbook's own synthesis of published guidance
Actions done0 of 4 doneActions
- Buy through an established framework where one fits, such as G-Cloud.
- Use the model contract schedules for data, security and exit rather than supplier terms.
- Make metadata delivery, quality reporting and data return measurable obligations with service credits.
- Publish transparency information as the Procurement Act requires.
Evidence
- Procurement route decision
- Contract data, security and exit schedules
Rulebook
- G-Cloud 15 frameworkV: Verified against the primary source on 2026-11-05
- Model Services ContractV: Verified against the primary source on 2026-11-05
- The Mid-Tier Contract: Schedule 20 (Processing Data)V: Verified against the primary source on 2026-11-05
- Procurement Act 2023V: Verified against the primary source on 2026-11-05
- Procurement Act 2023: guidance documentsV: Verified against the primary source on 2026-11-05
- The Sourcing PlaybookV: Verified against the primary source on 2026-11-05
Download every action as CSV to load into your own plan or tracker.
Stages, gates and assessments
Service assessments and gate reviews ask whether data management has been thought about, not for every template. Show that the critical data assets the service creates, changes or retires are known, owned, described, of known quality, protected and handed over, and keep the templates as your working evidence.
The Service Standard, service assessments, spend control and NISTA gate reviews are published elsewhere and linked here, not repeated. This section adds the critical data evidence for eachO: This playbook's own synthesis of published guidance. Assessors and reviewers judge against their own published criteria.
Four principles
- Show decisions and their reasons, not documents for their own sake. One page of evidence per point is usually enough.
- Use the same asset list, owners and catalogue records the organisation reports under the DAM Policy, so nothing is written twice.
- Name the business-as-usual owner of each critical asset before beta, not at handover.
- Bring the data people to the assessment: the owner or steward can answer questions a delivery manager cannot.
Stage
Strategy and business case
- Gates and assessments
- Spend control
- Gate 0 strategic assessment (programmes)
- Gate 1 business justification
- Service Standard points that touch critical data
- Before the Service Standard applies
- What reviewers will want to see about critical data
- Which critical data assets the programme creates, changes, depends on or retires, and who owns them.
- The cost of the data work, and of not doing it, in the business case.
- Data risks on the programme register.
- Evidence, and where it comes from here
- Data assets section of the business case (Data assets section for a business case)
- Critical assets ranked by value and risk (Rank critical assets by what they make possible)
- Data risks recorded (Record the main risk to each critical asset)
- Benefits with baselines (Pick a few measures and record the baseline)
- Common gaps
- Data appears as an IT cost line with no named assets
- No owner is consulted before the case is approved
- Read the published guidance
- Set up a commercial or digital and technology spend controls pipelineV: Verified against the primary source on 2026-11-05; Gate Review 0: Strategic AssessmentS: Found by web search; the site blocked automated checking, so open it before relying on it; The Green Book and accompanying guidanceV: Verified against the primary source on 2026-11-05
Stage
Discovery
- Gates and assessments
- Discovery peer review, where your organisation runs one
- Spend control
- Service Standard points that touch critical data
- point 1: Understand users and their needs, point 2: Solve a whole problem for users, point 3: Provide a joined up experience across all channels
- What reviewers will want to see about critical data
- That the team searched the catalogue for existing data before planning to collect more.
- Who holds the data the service needs today, inside and outside the organisation, and under what agreements.
- The people the data describes, and what they would expect.
- Evidence, and where it comes from here
- Catalogue searched and existing assets listed (Cataloguing step 2: Discover candidates)
- Third-party and supplier data found (Find the critical assets held or supplied by third parties)
- Users of the critical assets and their problems (Find out who uses each critical asset and what gets in their way)
- Reuse checked before new collection (Share and reuse safely across organisations)
- Common gaps
- Data collected again because nobody looked for it
- Supplier-held data discovered only at contract end
- Read the published guidance
- Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05; Service StandardV: Verified against the primary source on 2026-11-05
Stage
Alpha
- Gates and assessments
- Alpha service assessment
- Gate 2 delivery strategy
- Spend control
- Service Standard points that touch critical data
- point 2: Solve a whole problem for users, point 6: Have a multidisciplinary team, point 9: Create a secure service which protects users' privacy, point 11: Choose the right tools and technology, point 13: Use and contribute to open standards, common components and patterns
- What reviewers will want to see about critical data
- A data model with entities, relationships, identifiers and authoritative sources.
- Quality rules and targets agreed with owners before build.
- Open standards and government identifiers chosen, and how data would leave the chosen tools.
- Data roles in the team, and the protection and lawful basis for personal data.
- Evidence, and where it comes from here
- Entities, relationships and authoritative sources (List the entities and relationships in each critical asset)
- Quality rules and targets defined (Quality step 2: Define rules and targets)
- Identifiers and standards chosen (Use the identifiers and standards government expects)
- Exit questions answered for each new tool (Ask three questions of every new system)
- Request for information with ready-made requirements, if buying (Run a request for information with ready-made requirements)
- Common gaps
- Local code lists invented where national ones exist
- A tool chosen before its export was tested
- Read the published guidance
- Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05; Service StandardV: Verified against the primary source on 2026-11-05; Open standards for governmentV: Verified against the primary source on 2026-11-05; API technical and data standardsV: Verified against the primary source on 2026-11-05
Stage
Beta and build
- Gates and assessments
- Beta service assessment
- Gate 3 investment decision
- Spend control
- Service Standard points that touch critical data
- point 5: Make sure everyone can use the service, point 9: Create a secure service which protects users' privacy, point 10: Define what success looks like and publish performance data, point 13: Use and contribute to open standards, common components and patterns, point 14: Operate a reliable service
- What reviewers will want to see about critical data
- Catalogue records for every critical asset before go-live, owned and reported.
- A quality baseline and a data quality action plan, with results the owner has seen.
- Data protection, sharing agreements and security controls in place for the classification.
- Migration reconciled and signed off, and the business-as-usual owner named.
- Contracts that keep the data and metadata yours and let you leave.
- Evidence, and where it comes from here
- Assets described in the catalogue (Cataloguing step 5: Describe)
- Quality baseline measured (Quality step 3: Measure a baseline)
- Proof of concept with your own data, if buying (Prove it with your own data before you sign)
- Data clauses in contracts (Put data terms in every new contract)
- Migration reconciled (Migrate with reconciliation)
- Business-as-usual owner named (Name the business-as-usual owner early)
- Common gaps
- Catalogue records written after go-live
- Quality known only from test data
- No named owner after the project
- Read the published guidance
- Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05; Service StandardV: Verified against the primary source on 2026-11-05; Government Security ClassificationsV: Verified against the primary source on 2026-11-05; Data sharing: a code of practice - data sharing agreementsS: Found by web search; the site blocked automated checking, so open it before relying on it; Understanding accessibility requirements for public sector bodiesV: Verified against the primary source on 2026-11-05
Stage
Live and business as usual
- Gates and assessments
- Live service assessment
- Gate 4 readiness for service
- Gate 5 operations review and benefits realisation
- Service Standard points that touch critical data
- point 8: Iterate and improve frequently, point 10: Define what success looks like and publish performance data, point 14: Operate a reliable service
- What reviewers will want to see about critical data
- The yearly cycle running in business as usual: review, catalogue upkeep, quality checks and the return.
- Handover complete, with budget, and checked after six months.
- Benefits measured against the baseline, and data risks reviewed.
- An exit test each year for tools holding critical data.
- Evidence, and where it comes from here
- Handover checklist signed (Hand over with a checklist)
- Records kept current (Cataloguing step 8: Maintain and review)
- Benefits reported (Track and report the benefits)
- Data risks reviewed (Review data risks with the asset review)
- Exit tested (Test the exit every year)
- Common gaps
- The project team leaves and nobody reviews the asset
- Benefits claimed with no baseline
- Read the published guidance
- Service assessments and applying the Service StandardV: Verified against the primary source on 2026-11-05; Gate Review 5: Operations Review and Benefits RealisationS: Found by web search; the site blocked automated checking, so open it before relying on it; The Magenta BookV: Verified against the primary source on 2026-11-05
Stage
Migrate or retire
- Gates and assessments
- Retirement decision
- Gate 5 operations review, where it applies
- Service Standard points that touch critical data
- point 14: Operate a reliable service
- What reviewers will want to see about critical data
- Every critical asset in the system migrated, archived or disposed of under the retention schedule.
- Catalogue records marked retired, with what replaced them.
- The owner's sign-off before switch-off.
- Evidence, and where it comes from here
- Assets in the legacy system listed (List the critical assets in each legacy system)
- Switch-off signed off (Switch off safely)
- Legacy exit planned from the asset register (Plan legacy exits from the asset register)
- Common gaps
- Data needed for records or legal reasons lost at switch-off
- Retired assets vanish from the catalogue
- Read the published guidance
- Retiring your serviceV: Verified against the primary source on 2026-11-05; Manage information: guidance for governmentV: Verified against the primary source on 2026-11-05; Managing legacy technologyV: Verified against the primary source on 2026-11-05
Download the evidence by stage as CSV.
These practices are this portal's synthesisO: This playbook's own synthesis of published guidance of the rulebook entries listed with each one. They add no new requirements; where they differ from your department's own delivery or assurance rules, follow your department.
Was this page useful? Tell us Tell us privately, without GitHub