How we work
How Inside MECCA assesses evidence and reaches conclusions
Accounts, records, public sources, company statements and findings are not interchangeable. This page explains how Inside MECCA keeps them separate, tests material fairly and decides whether a claim can be published.
Different sources do different jobs
First-hand testimony
What a person experienced or directly witnessed.
Contemporary records
Messages, notes, emails, rosters or calendar entries created near an event.
Supporting documents
Material that supports, challenges or adds context to part of an account.
Independent corroboration
A genuinely separate source supports a material detail.
Company statements
What MECCA publicly or internally stated about its policies, systems or commitments.
Public records
Official documents, court or regulator records and other primary public material.
Media reports
Attributed reporting that records what a publisher reported, not an Inside MECCA finding.
Public commentary
Public posts and reviews used as research leads, not verified testimony or prevalence evidence.
How to read a claim or conclusion
First-hand
Based on direct personal experience or observation.
Supported
A reliable source supports a material detail.
Corroborated
A genuinely independent source supports a material detail.
Disputed
A relevant party has supplied conflicting information.
Unverified
The available material is not sufficient to support the claim.
Unable to determine
The evidence does not permit a responsible conclusion.
Finding
A defined conclusion supported by the available evidence after contradiction, fairness and response checks.
A label applies to a particular claim or source. It is not a judgement about the whole person who supplied it.
From account to publication decision
- 1
Log
Preserve the source, date, origin, access conditions, consent choices and known limitations.
- 2
Assess
Separate direct knowledge from hearsay, inference, documents and information supplied by others.
- 3
Test
Check dates, detail, authenticity, relevance, source independence and plausible alternative explanations.
- 4
Corroborate and challenge
Look for independent support, but also search for contradictions and evidence pointing in another direction.
- 5
Seek response
Give people or organisations facing material criticism enough information and a fair opportunity to respond.
- 6
Publish, narrow or hold
Publish only wording proportionate to the evidence. Otherwise narrow the claim, continue reviewing it or do not publish it.
How recurring accounts are counted without overstating them
- One eligible submission from one contributor is the primary unit.
- Follow-up messages and documents from that contributor do not become additional independent accounts.
- Duplicate, linked, coordinated or copied accounts must be identified before counting.
- A recurring theme is a research flag, not proof.
- Repetition does not establish accuracy, prevalence or a systemic pattern.
- Contributors are self-selecting.
- Submission counts cannot be presented as a workforce survey or incidence estimate.
- Small numbers may be combined or suppressed where disclosure could identify someone.
View the research theme codebook
The codebook helps to categorise qualitative accounts safely.
- Leadership & Culture: Bullying, psychological safety, leadership accountability.
- Employment Systems: Career progression, training, job security, performance metrics.
- Accountability: Speaking up, retaliation, policy gaps.
The investigation must test its own assumptions
Accuracy
Check material facts and preserve relevant context.
Fairness
Do not omit credible conflicting evidence or reasonable alternative explanations.
Clear attribution
Distinguish verified facts, testimony, allegations, company statements, analysis and opinion.
Source independence
Check whether apparently separate accounts share an incident, discussion, source or wording.
Minimise harm
Consider privacy, trauma, re-identification and foreseeable harm.
Confidentiality
Restrict sensitive information to people who need it for verification, editorial or legal work.
Independence
Contributors, subjects, sponsors and advocacy objectives do not determine the findings.
Right of reply
Provide the substance of proposed serious criticism and a reasonable opportunity to respond.
Detailed right of reply protocol
Inside MECCA provides a fourteen-day starting period for serious, non-urgent criticism before publication.
We provide the substance of the claims to give a fair opportunity to verify records and prepare a response.
Publication is a decision, not an automatic outcome
Receiving, recording or coding material does not automatically authorise publication. General contact permission, identification permission, quotation permission and publication permission must remain separate.
Publication decisions must consider:
- evidence strength
- wording and scope
- contradictory material
- right of reply
- contributor permission
- identification risk
- privacy and source safety
- public interest
- legal-review requirements
Corrections
- Material factual errors should be corrected promptly.
- Significant corrections should identify what changed and when.
- Corrections must not disappear into silent edits.
- A correction request should identify the statement, concern, supporting material and requested outcome.
Limitations
- Memory can be incomplete.
- Contributors are self-selecting.
- Similar accounts may not be independent.
- Documents may have authenticity, provenance or publication-clearance gaps.
- Company policies establish stated expectations, not consistent implementation.
- Public reporting establishes what was reported, not necessarily the truth of every allegation.
- No digital system can guarantee absolute anonymity or security.