Chapter 1
Start with need
Use credible local evidence and affected voices rather than generic goodwill.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Write a need statement with geography, group and baseline. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.
Chapter 2
Connect to the contract
Keep commitments relevant and proportionate to the subject matter and delivery.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Explain the causal link between contract resources and benefit. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.
Chapter 3
Choose a partner
Confirm competence, capacity, safeguarding, data and costs.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Complete mutual due diligence before promising delivery. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.
Chapter 4
Design the evidence chain
Separate input, activity, output, outcome and longer-term impact.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Define each metric before mobilisation. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.
Chapter 5
Deliver without harm
Prevent displacement, exploitation, double counting and unsafe volunteering.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Agree stop triggers and participant protections. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.
Chapter 6
Report honestly
Show delivery, variance, learning and corrective action without inflated monetisation.
Why this matters
A strong result begins with an explicit decision and a proportionate evidence boundary. Record what is known, what is reported by another party, what is inferred and what remains untested. This prevents enthusiasm, seniority or polished presentation from being mistaken for operating proof.
Do the work
Publish a concise performance record with caveats. Name the owner, affected people, dependencies, risk and review date. Where another organisation or person is involved, confirm their role, permission and expectations rather than assuming agreement.
Review questions
- What decision will this evidence support?
- Whose experience, data or opinion is it?
- What could make the conclusion inaccurate or unsafe?
- What is the smallest useful next test?
- Who can challenge or approve the result?
Practical checkpoint
Do not move on because a document exists. Sample a real record, observe the process where possible and ask whether a capable third party could reproduce the reasoning. Record a proceed, change, pause or stop decision and the evidence behind it.