Home

Solutions

QSA evidence workflow

Audit evidence for PCI DSS payment page controls

Build a traceable package that shows what the control is designed to do, who owns it, how it operated during the review period, and how exceptions and changes were resolved.

For PCI program owners, control operators, internal audit, and QSA-facing teams preparing evidence for requirements 6.4.3 and 11.6.1.

A complete sample should connect

The scoped payment journey and approved control design.

Observed scripts, owners, authorization, justification, and integrity method.

Baseline, detected change, alert, triage decision, response, and closure.

Review period, retention, reviewer, and reproducible export or source record.

Updated: July 26, 2026

Review basis

Reviewed against PCI DSS v4.0.1, PCI SSC payment-page guidance, and PCI SSC sampling guidance. Artifact examples support preparation; the assessor defines the final sample and sufficiency.

Design evidence

Procedures, scope, RACI, baseline rules, approval model, and retention requirements.

Operating evidence

Inventory snapshots, observed changes, alerts, reviews, decisions, and periodic control checks.

Traceable sampling

A reviewer can move from requirement to control, event, owner, action, and retained source record.

Separate design from operating effectiveness

A procedure or product screenshot can explain how a control is intended to work, but it does not demonstrate that the control operated throughout the assessment period. Conversely, a raw event export without scope, approval rules, and responsible roles is difficult to interpret.

Prepare both layers: design evidence describing the control and operating evidence showing representative execution, exceptions, review, and follow-up.

Create one chain from observation to decision

For script governance, connect each observed item to its owner, authorization, business justification, integrity method, and latest review. For page changes, connect the comparison to the baseline version, alert, responsible reviewer, release or incident context, decision, action, and closure.

Stable identifiers for pages, scripts, baseline versions, and events reduce manual reconciliation and make samples reproducible across exported reports and source systems.

Build a sample set that shows normal and adverse operation

A useful package does not contain only clean month-end screenshots. Include a normal inventory review, an approved release that changed the observed state, an unexplained deviation or controlled negative test, a closed exception, and evidence that the monitoring mechanism ran at the required or risk-defined frequency.

For each sample, state why it was selected and what population it represents. Sampling is an assessor decision: do not imply that a small convenient selection proves the whole period or every payment journey.

Preserve provenance and export context

Every exported file should identify its source system, generation time, scope filters, period, record count, schema or report version, and the stable identifiers needed to find the authoritative records. If files are transformed, record the transformation and preserve the source.

Checksums or signed packages can help demonstrate that a review bundle did not change after generation, but they do not prove that the underlying control was complete or effective. Access history, approval authority, and exception handling remain part of the evidence story.

Define retention and access before the audit

Agree how long evidence is retained, where authoritative records live, who can alter approvals or close events, and how access changes are logged. Retention should reflect the assessment period and the organization's policy rather than a generic marketing promise.

Run a sample retrieval before the formal review: select a page, script, date range, and change event, then verify that an independent reviewer can reconstruct the decision without oral context.

Run a dry review with someone outside the control team

Give an internal reviewer the requirement mapping and evidence index, but do not provide an oral walkthrough. Ask them to find the source inventory entry, baseline, event, owner decision, action, and closure for one selected journey and date.

Record every question they cannot answer from the package. Missing context, inconsistent time zones, broken references, unexplained record-count differences, and inaccessible source systems are evidence defects to resolve before the formal assessment.

Retrieval exercise

Pass a 15-minute evidence reconstruction test

Select one payment journey and one material change. An independent reviewer should be able to reconstruct what was expected, what was observed, who decided, what happened next, and where the authoritative records live.

Definition of done

A dated evidence index and manifest whose references resolve to source records, with scope, period, owner, reviewer status, exceptions, retention, and export provenance clearly stated.

01

Choose the population and sample

State the journeys, period, systems, event population, and rationale for the selected normal, changed, and exception records.

02

Link design to operation

Map the requirement and procedure to the inventory snapshot, baseline, event, decision, response, and periodic review records.

03

Capture export provenance

Record source system, query or filters, generation time, schema or report version, record count, and source identifiers.

04

Verify access and immutability controls

Confirm who can approve scripts, change baselines, close events, generate exports, and alter the authoritative history.

05

Reconstruct without oral context

Have someone outside the operating team follow the references and log missing, inconsistent, or inaccessible evidence.

Example evidence manifest

The manifest describes the exported package and points back to authoritative records; it is not itself proof that the control operated.

{
  "manifest_version": "1.0",
  "generated_at": "2026-07-26T09:00:00Z",
  "scope": {
    "journey": "checkout-card-entry",
    "period": ["2026-07-01", "2026-07-26"]
  },
  "source": {
    "system": "payment-page-control-system",
    "report_version": "inventory-and-events-v3",
    "record_count": 18
  },
  "records": {
    "inventory_snapshot": "INV-SNAPSHOT-2026-07-26",
    "baseline": "BASE-2026-07-20.2",
    "change_event": "EVT-1161-0042",
    "owner_decision": "DEC-0042",
    "closure": "CLOSE-0042"
  },
  "review": {
    "reviewer": "internal-audit",
    "status": "pending"
  }
}

PCI DSS payment-page evidence index

CSV fields for requirement, objective, journey, owner, source system, period, source record, reviewer, exception, retention, and access class.

CSV · UTF-8 · includes one clearly marked example row

Download

Agree the assessment period, population, sample, retention, and acceptable source systems with the assessor. Do not replace authoritative records with screenshots or manually edited extracts.

Evidence assembly workflow

Build the package continuously instead of reconstructing it shortly before the assessment.

01

Map

Map requirements to controls, owners, systems, and expected artifacts.

02

Collect

Capture inventory, baselines, events, decisions, and periodic reviews.

03

Correlate

Link observed items to approvals, releases, incidents, and responsible roles.

04

Review

Check completeness, exceptions, access, timestamps, and reviewer sign-off.

05

Export

Produce a scoped, dated package while preserving the authoritative source record.

Artifact map for 6.4.3 and 11.6.1

The exact sample and period must be agreed with the assessor.

#Control objectiveDesign artifactOperating artifactReviewer question
01Script inventoryInventory schema and scope procedureDated export of observed scriptsDoes the register match the live page?
02Authorization and justificationApproval workflow and role modelApproved records and sampled exceptionsWho approved each script and why?
03Integrity or authenticitySelected methods and exception criteriaChecks, signatures, hashes, or monitoring resultsHow is substitution or unauthorized change addressed?
04Page baselineBaseline approval and update procedureVersioned baseline with owner and releaseWhat state was expected on this date?
05Change detectionComparison scope and notification rulesEvent, diff, timestamp, affected journeyWas the change detected with useful context?
06ResponseTriage and escalation procedureOwner decision, action, and closureWas the event handled by an authorized role?
07RetentionRetention and access policyRetrievable sample and access historyCan the review period be reproduced?
08Export provenanceSource, schema, filter, and generation procedureManifest, record count, source IDs, generation timeCan each exported item be traced to its source?
09Exceptions and failuresException and control-failure procedureOpen/closed exceptions, owner, remediation, retestAre gaps visible rather than omitted from the sample?

Avoid screenshots as the only source of truth. Keep exports and source records machine-readable where practical and preserve their scope and generation time.

Evidence supports assessment; it does not make the conclusion

Cartelta can organize inventory, comparison, event, and export data. The organization and its assessor determine applicability, sampling, sufficiency, operating effectiveness, and the final PCI DSS result.

Audit evidence questions

Usually not by itself. A useful sample includes scope, time, source record, owner, control context, and any related decision or response. The assessor decides sufficiency for the specific assessment.

Keep the authoritative history according to policy, but prepare a scoped sample and exception summary for review. The sample should show normal operation as well as meaningful deviations and their resolution.

Subject to the assessor's sampling method, useful candidates include normal control operation, an approved release, an unexplained deviation or controlled test, an exception, a failed or missed run and its resolution, and a periodic owner review.

No. A checksum can support package integrity after generation, but sufficiency still depends on scope, population, provenance, authoritative source records, owners, decisions, operating period, and assessor testing.


Build the underlying controls

Implementation checklist

Track scope, controls, owners, evidence, retention, and RACI in one place.

Open

How to build an evidence pack

Use a practical article focused on review preparation and sample quality.

Open

JSIR monitoring methodology

Understand observation, baseline, limitations, and the path from change to evidence.

Open

Test whether one sample is independently reviewable

A focused evidence review exposes missing owners, timestamps, approvals, and response records before the formal assessment.

Discuss an evidence review