Methodology
How the research is actually done
Investigation credibility rests on method, not on marketing language. This page sets out what sources are used, how a claim becomes a corroborated finding, how confidence is assigned, and — just as importantly — what is never done.
Sources
Source categories used
Research draws only on lawful, publicly available information and material you provide.
- Public social media profiles and their visible history
- Public business and corporate registries where accessible
- Public websites, domain registration and archived web pages
- Publicly indexed images and reverse-image comparison
- Public news, adverse media and regulatory publications
- Publicly visible forum, marketplace and review activity
- Information and documents you supply yourself
- Publicly available sanctions, litigation and insolvency listings, where accessible
The corroboration standard
A finding needs independent support
One source is a lead. Two independent sources that agree are the beginning of a finding.
Source verification — illustrative
- Public corporate registryCorroborated
- Archived web contentCorroborated
- Public social profilePartially consistent
- Claimed employer websiteNot corroborated
A claim is not reported as established simply because it appears somewhere online. Before a finding is treated as reliable, it is checked against at least one additional, independent source. Where sources genuinely disagree, that disagreement is reported rather than resolved by picking whichever version is more convenient to the case.
Confidence levels
How confidence is assigned
Every significant finding in a report carries one of these labels, applied consistently.
- Corroborated
- Supported by two or more independent, reliable sources that are consistent with each other.
- Partially corroborated
- Supported by one credible source, or by sources that agree on part but not all of a claim.
- Unverified
- A claim that could not be checked against any independent source with the information available.
- Contradicted
- A claim for which available public information conflicts with what was stated or expected.
- Not established
- No reliable public information could be located either supporting or contradicting the claim.
Negative findings
An absence of information is reported as an absence
Failing to find something is not the same as proving it is false, and the two are never conflated.
When research does not turn up supporting information for a claim, the report states that plainly, together with what was searched and why the absence is or is not significant. It is not dressed up as a definitive negative finding unless the available evidence genuinely supports that conclusion.
Quality control
Analyst review, documentation and citation
Before a report is issued
- A human analyst reviews the full body of research collected
- Findings are cross-checked against each other for internal consistency
- Every source used is documented as research is carried out
- Confidence levels are assigned individually to each significant finding
- Conclusions that overstate what the evidence supports are revised before delivery
Limits
What is never used
These limits exist regardless of price, urgency or how the case is described.
- No access to private accounts, private messages or friends-only content
- No access to restricted, government or law-enforcement databases
- No pretexting, impersonation or deception to obtain information
- No physical or live surveillance of any person
- No location tracking of any device or individual
- No hacking, unauthorised access, or circumvention of security measures
Frequently asked questions
Is this the same as a background check website or people-search tool?
No. Automated lookup tools return raw, uncontextualised records with no verification of accuracy. Every case here involves a human analyst who cross-checks sources against each other and reports an explicit confidence level rather than a raw data dump.
What does 'corroborated' actually mean?
It means at least one additional, independent source supports the same finding. A single source — even a plausible one — is not treated as established fact on its own and is reported as unverified or partially corroborated instead.
Can you access private accounts or messages?
No. Research is limited to lawful, publicly available information and material you supply yourself. There is no access to private accounts, restricted databases, or anything requiring deception, hacking or unauthorised access.
Do you use surveillance or tracking?
No. There is no live location tracking, physical surveillance, or monitoring of ongoing communications. Research draws on the public record and information already provided by the client.
What happens if evidence contradicts itself?
Contradictions are reported, not smoothed over. Where sources disagree, the report explains the disagreement and reflects it in a lower confidence level rather than picking the more convenient version.
Who actually reviews the findings?
A human analyst reviews every case before a report is issued. Findings are checked against each other for consistency, sources are checked for reliability, and confidence levels are assigned deliberately rather than generated automatically.
Related pages
- How it worksThe six stages, from submission to report.
- ConfidentialityHow your case information is handled.
- Sample reportA worked example of findings and confidence levels.
- OSINT InvestigationCustom open-source research.
- PricingStarting prices for every service.
- About Fact InvestigationWho this service is for.
See the method applied to your case
Submit a case assessment to find out what can realistically be established, and how it would be verified.