How field surveys are conducted
A survey reports what an authorised researcher observed at a particular place and time. It does not prove that a broker is safe, solvent, regulated or suitable.
Before a visit
The broker, the legal entity and the address are matched before travel, and the visit is planned around a local safety assessment and a lawful public-access plan. Researchers do not trespass, bypass security, misrepresent authority or record where prohibited.
The checklist
Every visit follows the same physical-address checklist: locating the address, matching the building, external signage, the building directory, reception, the floor or suite, office signage, staff interaction, business materials, any appointment, photo and video capture, and the coordinates recorded. Each check has a fixed set of result labels, and the report shows them all, including the ones that could not be completed.
What the researcher writes
The report describes observable facts and attributes statements to their speaker. A receptionist's statement is recorded as an attributed observation, not as an independent finding. Researchers do not infer staff numbers, business scale, solvency or regulatory activity from the appearance of a location.
The outcome
The public badge must reflect the evidence recorded for that visit. A shared-office result is not converted into an office not found, and lobby identification alone is not converted into a full operating-office confirmation. A directory listing, receptionist statement or sign is one piece of evidence; the final status reflects the combined observations and limitations.
Photographs and video
Original files are retained privately with their metadata and a source hash. Public derivatives are edited only for exposure, white balance and straightening, and redacted for privacy: faces, licence plates, access codes, badges, screens, documents and uninvolved personal data. No generative fill or removal of relevant details is used, and public derivatives are never described as untouched originals. Real visit photographs and videos are never fabricated or replaced with generated evidence. The illustrations on the survey pages are interface media and are never shown as evidence of a visit.
Captions and alt text
A caption states what the image shows, where relevant, and when it was captured, and is never used to make a conclusion the image cannot support. Alt text describes the meaningful visible content and its role in the survey.
Review before publication
Each report passes a media-integrity check, a privacy review, a factual review that separates observations from statements and database facts, a status review against the taxonomy, a check that the report is attached to the correct broker and legal entity, and a publication review of captions, alt text, map precision and limitations.
After publication
Corrections, broker responses and address changes create dated updates rather than silent rewrites. A later visit creates a new survey record linked to the earlier report; the historical visit is never overwritten. A broker's response is published beside the report and does not change the recorded observations.
Privacy and safety
Exact coordinates or identifiable access details may be withheld where publication would create a safety, privacy or building-security concern. Researcher identity, travel pattern and real-time location are not exposed where that creates risk, and publication may be delayed where immediate release could create a safety or interference concern.