What Observed uses, weighs and excludes.
Observed works from lawful, public and attributable information. Every source is classified by type, treated according to its reliability and limits, and weighed against the wider evidence base before any benchmark comparison or publication decision is made.
Evidence principles
The source rules are part of the methodology.
Observed does not treat every public source as equally strong. The methodology separates formal records, reported information, anonymous experience signals, organisation-owned claims and stakeholder material before drawing any comparison.
Public and lawful
Sources must be publicly available and lawfully accessible. Observed does not use surveillance, impersonation, hacking, private databases, covert recordings or unlawfully obtained material.
Attributable and dated
Sources must be identifiable enough to record where they came from, when they were accessed and how they should be classified. Unsupported material is not enough.
Classified before weighed
A source is first classified by type and role. It may be formal evidence, reported information, anonymous signal, organisational positioning or stakeholder material.
What counts as a public signal?
A signal is not a finding. It is a classified input.
A public signal is a lawful, accessible and attributable piece of public information that may indicate something relevant about organisational conduct, governance, culture, public claims, stakeholder experience or public-interest accountability.
Signals are not treated as automatic proof. They are classified, compared with other source types, checked for contradiction and assessed against recognised benchmarks before any finding is developed.
| Source type | How Observed treats it | Main limitation |
|---|---|---|
| Formal public records | Regulator notices, tribunal decisions, public registers, public funding records and formal accountability material are treated according to their legal or official status. | Formal records may be narrow, historic or limited to matters that reached an official pathway. |
| Media reporting | Media reporting is classified as reported public information, not automatically treated as independently verified fact. | Reporting may rely on selected sources, limited documents or unresolved allegations. |
| Workplace review platforms | Glassdoor, SEEK and similar platforms are classified as anonymous employee-experience signals. | Anonymous reviews cannot support named-organisation publication alone, regardless of volume. |
| Organisation-owned material | Websites, annual reports, impact reports and social media are treated as public claims, positioning or self-description. | These materials are not independent verification of organisational practice. |
| Stakeholder material | Public submissions, stakeholder reports, parliamentary material and official information responses are assessed by context and source role. | Stakeholder material may reflect a specific interest, perspective or campaign context. |
| Client-supplied material | Client-supplied material may be considered only where lawful, relevant and assessed for reliability. | It does not replace the public-source requirement and may not be suitable for publication. |
Source categories
The evidence base is deliberately broad.
Observed looks across different public source categories because organisational patterns are rarely visible in one place. Source diversity is central to confidence, fairness and publication decisions.
Government and regulator records
Public enforcement notices, regulator findings, public registers, tribunal decisions, ombudsman material and other formal public accountability records.
Media publications
Public reporting from media outlets, treated as reported information and weighed according to context, source quality and corroboration.
Workplace review platforms
Public employee-experience signals from platforms such as Glassdoor and SEEK, classified as anonymous signals rather than verified findings.
Organisation-owned publications
Websites, annual reports, impact reports, promotional material, public statements and social media posts used to understand public positioning.
Funding and procurement information
Public grants, contracts, procurement material, government funding records and public reporting obligations where available.
Submissions and public responses
Public submissions, parliamentary material, stakeholder reports, official information responses and community-facing documents.
Legal or tribunal material
Publicly available tribunal, court, employment, complaints or regulator material, handled according to legal context and publication risk.
Academic and professional benchmarks
Peer-reviewed academic frameworks, recognised good-practice standards and professional guidance used for comparison.
Client-supplied material
Considered only where lawful, relevant and reliable. It is assessed carefully and does not override the public evidence model.
How sources are weighted
Not all sources carry the same confidence.
Source weight depends on independence, formality, public accessibility, specificity, recency, relevance, corroboration and whether contradictory information is available.
Higher-weight source characteristics
- Formal public records or regulator material.
- Specific, dated and attributable information.
- Independent sources that point to similar issues.
- Evidence that can be checked against public records.
- Material that is current enough to remain relevant.
Lower-weight source characteristics
- Anonymous, uncorroborated or single-platform material.
- Generalised statements without specific context.
- Material that may reflect a single personal dispute.
- Historic information with unclear current relevance.
- Claims that cannot be checked against other public sources.
Confidence factors
Confidence is tied to source strength, not certainty.
Observed does not use confidence ratings to claim factual certainty. Confidence reflects whether the public evidence base is broad, attributable, relevant and strong enough to support a careful benchmark comparison.
Source diversity
Signals from multiple independent source types are stronger than signals concentrated in one place.
Source hierarchy
Formal records, regulator material and tribunal decisions carry different weight from anonymous reviews.
Specificity
Specific, dated and attributable material is stronger than broad or generalised commentary.
Corroboration
Confidence increases where different public source types point to related patterns.
Contradiction
Contradictory or balancing material must be included where it exists.
Recency
Older signals may still matter, but current relevance must be assessed.
Public-interest fit
The source base must support an organisational accountability question, not merely private dispute content.
Publication risk
Legal, privacy and proportionality risks affect what can responsibly be published.
Named-organisation findings are suppressed unless signals are drawn from at least three independent source types. A concentration of signals from a single source type, including multiple anonymous reviews on one platform, does not meet the threshold regardless of volume.
This threshold protects against single-source patterns, thinly sourced findings, small-sample identification and confidence ratings that exceed what the available evidence can support.
Where the threshold is not met, a matter may be held pending further public evidence, added to a sector monitoring watchlist, used only as background context for sector-level analysis or declined.
Anonymous review platforms
Useful signals, clear limits.
Public workplace review platforms can provide useful employee-experience signals. They may indicate patterns worth examining, especially when similar themes appear across different time periods, roles or public sources.
But anonymous reviews are not treated as findings of fact. They are not enough on their own to support named-organisation publication. They must be classified, limited and weighed against other independent source types.
This protects the model from over-reliance on anonymous material while still recognising that employee-experience signals can be relevant to public-interest organisational analysis.
Evidence boundaries
Some material is excluded, even if it appears relevant.
Observed’s evidence boundary is absolute. Public-interest purpose does not justify using private, unlawful, unreliable or disproportionate material.
Accepted source conditions
- Publicly available and lawfully accessible.
- Attributable to a source, platform, register, publication or public record.
- Relevant to organisational behaviour, governance, culture, public claims or accountability.
- Capable of being classified by source type and reliability limits.
- Suitable for comparison against an academic or good-practice benchmark.
Excluded material
- Private, hacked, leaked or unlawfully obtained information.
- Covert recordings, impersonation, surveillance or private databases.
- Unsupported claims without public-source grounding.
- Material focused on naming or targeting natural persons.
- Content that would create disproportionate privacy, safety or legal risk.
Public evidence only. Source diversity required. Confidence limited by the evidence base.
What appears in outputs
Published work must show its evidence handling.
Observed outputs should make source treatment visible. Readers should be able to understand what was searched, what was found, how sources were classified and what the analysis can and cannot support.
Source register
Outputs identify source categories, access dates, source roles and how each source type was classified.
Evidence limits
Outputs state where sources are anonymous, reported, organisation-owned, formal, historic, narrow or incomplete.
Confidence explanation
Outputs explain why a finding carries the confidence level assigned and what would be needed to strengthen it.
Why this matters
The evidence standard protects the purpose of the work.
Observed exists to make organisational accountability more evidence-based, fair and accessible. That only works if the source base is handled carefully.
The evidence standard prevents the work from becoming private investigation, unsupported allegation, personal targeting, reputational attack or advocacy detached from verifiable public material.
The result is a clearer and more defensible comparison between publicly visible organisational signals and recognised research or good-practice benchmarks.
Have public sources that may warrant review?
Observed can assess whether the available public evidence is lawful, relevant, diverse and strong enough to justify an Evidence Check or deeper benchmark analysis.
Public evidence. Academic benchmarks. Better accountability.
