What should you ask for in a penetration test report?
Ask for two useful views: an understandable account of scope and important findings, and enough technical detail for your engineers to act. Agree how limitations, remediation guidance and later retest results will be represented before the engagement begins.
Use a sample report to test the buying promise
NIST SP 800-115 describes reporting identified vulnerabilities, risk ratings and mitigation guidance. Use that foundational framing to ask for a redacted sample deliverable, then assess whether it works for your readers. The following worksheet is our suggested buying checklist.
On a small screen, scroll the table sideways to see every column.
| Reader | What to ask the sample to show |
|---|---|
| Engineering | Affected component, supporting evidence, clear finding description and useful remediation direction |
| Business owner | Scope, dates, important impacts and decisions that need attention |
| Customer reviewer | What was assessed and what the document actually concludes |
| Future maintainer | Stable finding IDs and a way to reconcile retest results |
Source: NIST SP 800-115: Technical Guide to Information Security Testing and Assessment (2008)
Example: two audiences, one engagement
A fictional SaaS company needs its engineering lead to plan fixes and its sales team to answer a customer’s security review. A detailed technical report may be appropriate for the first audience, while a shareable summary may be useful for the second.
Request both outputs if needed, with permitted recipients and handling agreed. Do not assume that a badge, certificate or summary contains enough information to make a remediation decision.
Ask how severity becomes a work queue
Have the provider explain how it communicates urgent findings during the engagement and how it describes business context in the final output. Your team should be able to identify the owner and next action for each finding.
Do not turn a count of findings into a provider league table. Different scopes, environments and reporting approaches can produce different totals. Evaluate whether the report answers the agreed question and supports useful action.
Keep the report and retest story together
Before the test starts, ask what happens to the deliverable after fixes are checked: a new version, an appendix or a separate confirmation. Choose a process that preserves the original finding and clearly records the later outcome.
Ask the receiving customer what evidence they expect before promising that a specific document will satisfy them. This page is a report-buying guide and makes no claim that a report automatically satisfies SOC 2 or another assurance requirement.
An executive summary with a decision attached
Fictional example—not a finding from a real engagement:
We assessed build portal-2026.11.02 in staging on November 3–5, 2026, using two synthetic customer accounts and one support account. Finding PT-03 concerns the invoice authorization boundary. Testing identified a tenant-isolation weakness that allowed a test customer account to retrieve another test tenant’s invoice. This creates a risk of unauthorized disclosure of customer billing information. We did not establish that the issue had been exploited outside the test.
The engineering owner should correct the authorization check, assess whether related invoice paths share the same weakness and arrange focused retesting. Leadership needs to assign remediation capacity and decide how the unresolved risk will be managed before the next release. The payment processor and employee identity environment were outside the agreed scope, so this report does not assess their security.
This summary separates observation, implication and decision. It does not invent a financial loss estimate or claim a real breach. Actual reports should replace the example with supported findings, agreed dates, named owners and evidence from the engagement.
Turn severity into a work queue
Illustrative work item: Finding PT-03 becomes an engineering ticket owned by the API team. The ticket links to the restricted report, states the affected invoice authorization boundary and identifies the expected permission behavior. The team sets a remediation deadline through its own risk process and requests a retest after deployment.
Avoid copying sensitive report details into a widely accessible project board. Give each owner enough information to act while keeping exploit evidence and customer data in the appropriate restricted location.
Severity helps prioritize remediation, but the queue also needs dependencies and ownership. A medium-rated issue with an exposed critical workflow may need attention before an isolated issue whose remediation is already contained. Record the rationale rather than changing labels to make the dashboard look better.
Expanded November 16, 2026 with an illustrative executive summary and remediation work item. Original report-buying guidance retained.