Scope disputes rarely begin during the assessment. They begin when the report arrives and someone asks why one system was tested, why another was left out, or why a finding was rated critical without evidence. A written scope removes those questions because every boundary was agreed before the first request was sent.
This template is for a Bangladeshi SME buying its first or second vulnerability assessment. Copy it into a document, complete it with the vendor, and attach the signed version to the purchase order. The cost implications of each choice are covered separately in the vulnerability assessment cost guide; this page is about deciding what is and is not being tested.
Why the scope document belongs in the contract
Most vendors quote from a short call, and most buyers approve that quote without checking the assumptions behind it. The assumptions are where disagreements live: how many applications, whether staging is included, whether test accounts will be provided, whether retesting is covered, and what happens if a critical finding appears on the first day. A scope checklist forces those assumptions into writing while the commercial terms are still open.
It also protects the tester. An assessment without documented authorisation can look like an attack to a hosting provider or an internet service provider. Where a business hosts with a regional provider or uses a third-party payment gateway, the boundary between "our system" and "someone else's platform" is rarely obvious.
1. Authority, contacts and legal cover
- Signed authorisation letter naming the legal entity, domains, IP ranges and testing dates. Owner: managing director. Evidence: signed PDF attached to the contract. Review: before every test window.
- Named technical contact who can approve a change mid-test, plus a deputy who answers when that person is unavailable. Owner: IT lead. Evidence: names and phone numbers in the scope document. Review: each engagement.
- Emergency stop procedure describing how the tester pauses work if a production service fails. Owner: IT lead. Evidence: agreed phone tree and escalation times. Review: before testing.
- Hosting provider permission where the application sits on shared or managed infrastructure. Owner: IT lead. Evidence: provider ticket or written confirmation. Review: before testing.
- Insurance and liability check for accidental damage during testing. Owner: finance. Evidence: policy clause noted in the file. Review: annually.
2. Assets and application boundaries
List targets individually. "Our website" is not a target; a domain, an application, an API base URL and a mobile build are targets, and each needs a different testing method.
- In-scope domains and subdomains, including staging or admin hosts reachable from the internet. Owner: IT lead. Evidence: DNS export dated before the test. Review: each engagement.
- APIs and mobile applications with versions, authentication method and test environment. Owner: product or IT lead. Evidence: endpoint list and build numbers. Review: each engagement.
- Third-party components such as payment gateways, SMS providers and courier APIs, marked "configuration review only" unless the provider has authorised testing. Owner: IT lead. Evidence: list with authorisation status. Review: each engagement.
- Out-of-scope assets written down explicitly, including staff laptops, the office network and any sister company. Owner: IT lead. Evidence: exclusions section signed by both parties. Review: each engagement.
- Source code and configuration stated as included or excluded. Owner: engineering lead. Evidence: repository access list. Review: each engagement.
3. Depth of testing and permitted techniques
Two vendors can both say "vulnerability assessment" and deliver very different work. Record the method, not just the label.
- Authenticated or unauthenticated testing, and which roles will be used. Owner: IT lead. Evidence: test account matrix. Review: each engagement.
- Automated scanning versus manual verification, with the expected proportion stated in the proposal. Owner: vendor, approved by IT lead. Evidence: methodology section. Review: each engagement.
- Exploitation limits, such as proof-of-concept only, no persistence, no data exfiltration and no denial-of-service testing. Owner: both parties. Evidence: rules of engagement. Review: each engagement.
- Rate limits and maintenance windows so scanning does not starve production traffic or trigger provider defences. Owner: IT lead. Evidence: agreed thresholds. Review: before testing.
- Social engineering and physical testing explicitly included or excluded. Owner: managing director. Evidence: written decision. Review: each engagement.
4. Schedule, access and escalation
- Test calendar with start and end dates, blackout periods around month-end, campaigns or exams, and a named approver for any extension. Owner: operations lead. Evidence: shared calendar entry. Review: weekly during the engagement.
- Access provisioning for test accounts, VPN or IP allow-listing, with a named person who can fix access failures within an agreed time. Owner: IT lead. Evidence: access checklist. Review: before testing.
- Daily status note during active testing, even when nothing has been found. Owner: vendor. Evidence: short email or message thread. Review: daily.
5. Deliverables and acceptance criteria
Define the report before the work starts. A finding should state the affected asset, the reproduction steps, the evidence, the business impact, a severity rating and a remediation recommendation. Ask which severity method will be used and whether the vendor will recalculate ratings after your team provides context.
- Report format with an executive summary, a technical register and an appendix of evidence. Owner: vendor. Evidence: sample report reviewed at selection. Review: each engagement.
- Retest terms covering how many findings, within what period and at what additional cost. Owner: procurement. Evidence: contract clause. Review: each engagement.
- Raw scanner output delivered or withheld, and in what format. Owner: vendor. Evidence: deliverable list. Review: each engagement.
- Acceptance meeting where the business owner signs that the agreed scope was covered. Owner: IT lead. Evidence: signed acceptance note. Review: at closure.
6. Data handling and confidentiality
Testing produces screenshots, exports and sometimes copies of customer records. State where that material is stored, who can see it, how long it is kept and how it is destroyed. Confirm whether the vendor will use the findings in marketing material and require written approval before any client name is published.
7. Sign-off and change control
Close the document with a sign-off block: business owner, IT owner, vendor lead, date and version. Any change to targets, dates or techniques should create a new version rather than an email amendment. Keep the signed scope with the report so an auditor can see what was agreed.
Using the checklist in an RFQ
Send the completed checklist with the request for quotation and ask each vendor to mark every line as included, excluded or priced separately. The responses will separate serious bidders from those who quote a scanner run and call it an assessment. Compare on the same target list and the same deliverables, not on the headline number alone.
Where this fits in our work
Scoping sits inside Managed Technology & Cybersecurity, alongside hardening and ongoing monitoring. The pricing page explains how assessment work is separated from monthly support. For the budget side of the same decision, read the vulnerability assessment cost guide.
Sources
- NIST SP 800-115, Technical Guide to Information Security Testing and Assessment, accessed 17 November 2026.
- OWASP Web Security Testing Guide, accessed 17 November 2026.
- Penetration Testing Execution Standard, accessed 17 November 2026.
- CISA, Cyber Essentials, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.