Evidence behind every result.
A score is easier to trust when the answers, supported files, revisions, reviewer decisions and release history remain connected to the result instead of being scattered across systems.
Preserve the evidence behind the workflow.
The evidence model keeps structured answers, supported image/PDF files and workflow history attached to the programme record, with correction and release state kept separate.
Questionnaire answers
Tie each submission to the published questionnaire version used for that work.
Supported files
Attach supported photographs, screenshots or PDF documents to the relevant answer.
Revisions
Keep correction lineage rather than silently overwriting the prior submission.
Use the evidence the current workflow can preserve reliably.
RetailHawk keeps native evidence deliberately bounded. Today, file evidence is limited to JPEG/JPG, PNG and PDF. Screenshots and photographs are therefore handled as images; structured questionnaire answers preserve written observations alongside the published question set.
Native file evidence
Attach JPG/JPEG, PNG and PDF files to file questions. Uploaded evidence is type-checked and passes through the evidence scanning workflow before it becomes usable.
Attributable context
Keep answer text, submission and review timestamps, questionnaire version, assignment and outlet context connected to the same record. Where the programme already carries location context, that context remains attributable to the work.
Scoped rather than implied
Audio, video, call recordings and signatures are not presented as native evidence uploads today. Where a programme needs those evidence forms, confirm the required integration or external workflow during scoping rather than assuming unsupported capture.
Separate submission, review and release.
RetailHawk keeps reviewer decisions attributable and treats client release as its own controlled event. Later correction does not retroactively redefine the work that was previously released.
Reviewer decisions
Record approve, return or reject decisions separately from field submission.
Release history
Preserve which approved result was released to the client audience and when.
Score snapshots
Keep released scoring tied to the released result rather than recalculating historical output from changed questions.
One result, multiple attributable states.
A synthetic audit submission includes structured answers and a PDF. A reviewer returns one answer for correction, the revised submission is approved and the approved result is released. The prior revision and decisions remain part of the history.
1. Submit
Capture the answer set and supporting evidence.
2. Review
Return, correct and approve without deleting the earlier state.
3. Release
Make the approved result visible to the authorised client audience.
Keep each decision connected to the evidence that supported it.
The diagram explains the attributable workflow; it is not a product screenshot.
Bound evidence access to authorised use.
Evidence is stored privately and accessed through platform controls. Sensitive evidence access and relevant operational events can be recorded in bounded audit history without exposing internal security architecture in public copy.
Private evidence storage
Keep evidence out of public paths and serve it through authorised access.
Bounded audit events
Record relevant security and evidence-access events without presenting an unlimited surveillance log.
UK hosting
Built and hosted in the UK.