Define the question and intended decision
Agree what needs to be decided, who will use the work and what a useful output should contain.
Evidence you can follow.
Judgment you can question.
The Sage Evidence & Quality Protocol is our working process for keeping research focused, traceable and useful. It helps us check the work and explain the reasoning behind it.
Assess scope, information and competence
Review available sources, data permissions, time and the expertise needed. Narrow the scope or identify specialist help where necessary.
Plan the research method
Set the research approach, comparison criteria, inclusion rules and review points before gathering evidence.
Collect and register evidence
Record sources, dates and relevant context. Keep observations, supplied information and assumptions distinguishable.
Analyse and verify
Check calculations and comparisons, trace important claims to sources and look for contrary evidence or missing context.
Prepare recommendations
Connect the findings to options, risks and practical next steps. Explain where judgment is being used.
Review accuracy, limitations and confidentiality
Check the output against the agreed question. Make uncertainty visible and remove information that should not be disclosed.
Deliver the agreed outputs
Hand over the brief, report or framework in the agreed format, with the evidence and limitations needed to use it responsibly.
Tools and responsibilityAI can assist.
AI can assist.
People remain accountable.
AI may assist discovery, organisation, analysis or drafting. Outputs still need source verification, appropriate data handling and professional oversight. An AI-generated statement is not treated as evidence by itself.
Confidential materials require an agreed handling arrangement. We do not assume that permission to read a document includes permission to upload it to an external AI tool.
Our confidentiality boundaries