Directly supported by a strong primary source or multiple independent sources.
Methodology
Evidence that can be traced. Conclusions that can be tested.
A disciplined method protects the quality of the decision. Every material finding should show what was observed, how it was interpreted and how much uncertainty remains.
Six stages
From a useful question to a defensible answer.
- 01
Define the decision
Translate the business need into a precise research question, scope, target and evidence threshold.
- 02
Build the source plan
Identify primary, technical, regulatory, commercial and visual sources likely to reduce the key uncertainty.
- 03
Collect with provenance
Record source origin, publication context, date and relevant extracts so material remains traceable.
- 04
Verify and challenge
Cross-reference independent sources, check recency, identify contradictions and test alternative explanations.
- 05
Interpret proportionately
Separate direct observation from assessment and inference. Assign confidence that reflects the evidence.
- 06
Deliver an evidence trail
Present conclusions, limitations, sources, information gaps and the questions that would most improve confidence.
Evidence standard
Quality depends on provenance, context and corroboration.
Source reliability and recency
Sources are assessed for authority, proximity to the subject, publication purpose and date. A source may be authoritative for one fact and weak for another. Old material is not discarded automatically, but its relevance to the current question is stated.
Contradictions and gaps
Conflicting evidence is retained and explained. Reports identify missing information instead of smoothing it over. Where more than one explanation fits the evidence, alternatives remain visible.
Observation and interpretation
Directly observable facts are recorded separately from analytical judgement. Assessments explain their evidential basis; inferences are labelled as possibilities that need further verification.
Confidence model
Plain language for uncertainty.
Confidence refers to the support for a specific finding—not to the overall professionalism of the work.
Supported by consistent evidence, with limited uncertainty.
Plausible and evidence-supported, but incomplete.
A useful signal that requires further verification.
Insufficient reliable evidence for an assessment.
Image and video analysis
What can be seen is not the same as what can be proved.
Public photographs and video can support observations about visible equipment, facility layout, process sequence, product configuration and change over time. Technical interpretation may use manuals, supplier material and independent sources to test what those features could indicate.
Visual analysis does not automatically prove hidden specifications, material composition, internal tolerances, production yield or processes outside the frame. Reports state these limitations and avoid converting visual resemblance into certainty.
Ethical and legal boundary
Lawful public-source research only.
Within scope
- Publicly accessible documents and websites
- Public corporate and professional activity
- Lawfully published photographs and video
- Patents, tenders, certifications and public records
Outside scope
- Hacking or unauthorised access
- Impersonation or social engineering
- Stolen, leaked or unlawfully acquired material
- Unnecessary collection of sensitive personal data
See the method in practice