A penetration testing RFP needs five sections (company background, scope, requirements, evaluation criteria, and timeline) and a short list of questions that make each vendor say what it will actually do. Penetration testing proposals tend to read alike. Every vendor follows a recognized methodology, has certified testers, and delivers an executive summary. A request for proposal exists to get underneath that language. It lets you compare what each firm will actually do, who will do it, and what you will hold at the end. This template gives you the structure, the 25 questions that separate the answers, a scoring matrix, and a version you can copy.
You do not need a forty-page document. A two-page RFP with precise scope and pointed questions gets better responses than a long one. Send it to a shortlist you have already screened using our guide to choosing a penetration testing company, and let the responses decide. If you are still deciding whether to hire a firm at all, start with the outsourcing penetration testing guide.
The RFP structure
Five sections cover everything a vendor needs to respond well.
1. Company background
Give the vendor enough context to size the work and judge fit: what your company does, its size, its industry, and the regulatory frameworks that apply. State why you are testing now, whether that is a compliance deadline, a customer requirement, a new product, or a program you are building. Name the internal owner of the engagement and who will receive the report.
2. Scope
This section determines the price and the timeline, so be specific. List each target with the detail a tester needs:
- Web applications: URL, number of user roles, what the application does, and whether testing is authenticated
- APIs: number of endpoints, authentication scheme, and whether documentation exists
- Networks: number of external hosts, and for internal testing the number of hosts, sites, and Active Directory domains
- Cloud: provider, number of accounts or subscriptions, and the services in use, plus any Kubernetes clusters, which a Kubernetes penetration test scopes cluster by cluster
- Mobile: platforms, and whether the backend API is included
- Social engineering or red team: objectives, the people or systems in play, and any constraints
- Vulnerability scanning: if you want scans and the test in one proposal, ask for vulnerability assessment and penetration testing services as a single scope
Add the environments available for testing, the testing windows you can accept, and anything explicitly out of scope. If you cannot describe the scope precisely, ask vendors to propose one from a written description. A good firm responds with scoping questions rather than a guess. Our guide to scoping your first penetration test covers what to gather, and the penetration testing checklist covers what to have ready before kickoff.
3. Requirements
State what a compliant response must include:
- A named testing team with certifications listed per person
- A described methodology with reference to a public standard
- A redacted sample report
- A fixed price for the stated scope, with any assumptions listed
- Retest terms in writing
- Compliance mapping for the framework you name
- Certificates of insurance and a willingness to sign your NDA and MSA
4. Evaluation criteria
Tell vendors how you will score them, and set the weighting before responses arrive. Otherwise the loudest proposal sets it for you. A sensible order: manual depth and methodology first, reporting and retest terms second, team and continuity third, compliance fit fourth, and price last among the scored items. To check whether a quote is in range, benchmark it against the US penetration testing price index. Treat insurance and legal terms as pass or fail.
5. Timeline and process
List the dates: RFP issued, deadline for vendor questions, response deadline, shortlist interviews, sample report review, decision date, and the testing window you need.
The 25 questions
Each question comes with what a good answer looks like.
Team and credentials
1. Who will perform the testing on our engagement, and what certifications do they hold? A good answer identifies the individual testers and gives each person’s hands-on credentials and experience, not a company-wide summary. Our testers are senior, in-house employees, never subcontractors.
2. Are your testers employees or subcontractors, and where are they located? A good answer states that testers are employees, names where they work from, and commits that the assigned team will not change without notice.
3. How much offensive security experience does the assigned lead have, and in what kinds of environments? A good answer gives a specific figure for the lead and describes environments like yours, rather than an average across the firm.
4. Can you provide references from clients in our industry or under our compliance framework? A good answer offers contactable references, with the client’s permission, and describes comparable engagements without naming clients.
5. Will the same testers be available for our retest and for future engagements? A good answer explains how continuity works, so the person verifying your fixes understands the original findings.
Methodology and manual depth
6. What share of the engagement is manual testing, and what is automated? A good answer states that manual testing is the majority of the effort, with tools used for discovery and coverage. Our breakdown of automated vs manual penetration testing explains what the humans do that the tools cannot. If a vendor’s answer is mostly a platform, read autonomous penetration testing vendors compared before you score it.
7. Which methodology do you follow, and will the report show coverage against it? A good answer names public standards such as PTES, the OWASP Web Security Testing Guide, or NIST SP 800-115. It also commits to a coverage section in the report. Our methodology page describes ours.
8. How do you test business logic and access control between user roles? A good answer describes manual testing with one account per role, and checks of every function and object for horizontal and vertical privilege escalation.
9. Do you attempt exploitation and chain findings, or stop at identification? A good answer confirms exploitation to demonstrate real impact within agreed rules of engagement. Chained findings should be reported as an attack path, not a list of separate items.
10. How do you handle a critical finding discovered during testing? A good answer names a notification window, a contact on the vendor side, and the rule for pausing testing if something dangerous is found.
Scope and pricing model
11. Is the price fixed, hourly, or credit-based, and what could change it? A good answer is a fixed price, agreed in writing before work starts. The scope statement should be specific enough that only a scope change you request can move it. That is how we price every engagement; starting prices are on the pricing page. Credit-based pricing is mostly a platform model, and PTaaS companies compared shows how each vendor’s credits and subscriptions work.
12. What does the price include? A good answer itemizes it: testing, the executive and technical reports, a findings platform, a readout call, the retest, and an attestation letter. Anything sold separately should be listed with its price.
13. How do you size an engagement, and what do you need from us to do it? A good answer describes a scoping questionnaire or call covering applications, roles, endpoints, hosts, and environments, and gives a turnaround for the proposal.
14. What is your lead time to start, and how long will testing run for our scope? A good answer gives a start date tied to your access being ready and a duration tied to your scope. A generic promise that applies to every client is not an answer.
15. Can you test in production, staging, or both, and what safeguards apply? A good answer describes testing windows, rate limits, excluded actions, and how the tester coordinates with your team when something unexpected happens.
Reporting and retest
16. Can we see a redacted sample report before we decide? A good answer is yes, without conditions. You can request ours and hold every other response to the same standard.
17. What does the report contain, and is there a separate executive summary? A good answer describes an executive report for leadership and auditors, plus a technical report with reproduction steps, evidence, severity ratings, and stack-specific remediation guidance. Our penetration testing report template lists every section to check.
18. Is a retest included, what is the window, and does the final report reflect the fixes? A good answer includes the retest, states a reasonable window, and commits to an updated report showing each finding closed. Ours is free on every engagement, with phishing as the exception, since a phishing campaign produces no findings to remediate.
19. How are findings delivered: a PDF only, or a platform where we can track remediation and request retests? A good answer offers both, with the platform included rather than licensed per seat.
Compliance mapping
20. Will the report map findings to our framework? A good answer names the frameworks the firm maps to, such as SOC 2, PCI DSS, HIPAA, ISO 27001, NYDFS 23 NYCRR 500, or CMMC. For UK contracts, add Cyber Essentials Plus to the list. It should also describe experience with the evidence auditors under your framework expect.
21. Do you provide an attestation letter we can share with customers and auditors? A good answer is yes, as a standard deliverable: a short letter confirming scope, dates, methodology, and outcome without technical detail.
22. Are your testers independent of any systems we run, and will you confirm that in writing? A good answer confirms the firm does not manage, host, or build anything in scope, and offers an independence statement for your auditor.
Insurance and legal
23. What professional liability, errors and omissions, and cyber liability insurance do you carry? A good answer offers a certificate of insurance on request and confirms the limits meet the requirements in your vendor contracts.
24. Will you sign our NDA and master services agreement, and what are your authorization terms? A good answer accepts a mutual NDA, provides its own agreement for comparison, and requires written authorization before any testing begins. You can read our master services agreement before you send the RFP.
25. How is our data handled during and after the engagement? A good answer describes where evidence and credentials are stored, who can access them, how they are encrypted, and when they are destroyed after the retest. Ask for the vendor’s own SOC 2 report or ISO 27001 certificate if it has one, and a data processing agreement (DPA) if testers will see personal data. Our trust center covers how we handle client data.
Extra questions for PTaaS and automated testing platforms
If the RFP is for a penetration testing as a service (PTaaS) subscription, a continuous testing program, or an automated platform, keep the 25 questions and add these five.
P1. How many human tester days does the subscription include, and on which assets? A good answer gives days or hours per asset and names who performs them. “Continuous” with no human days stated usually means scheduled scans.
P2. What does a credit buy, do credits expire, and what happens when they run out mid-test? A good answer converts credits into tester days or asset sizes in writing, states the expiry date, and gives the price of additional credits before you sign.
P3. Is retesting capped? A good answer states how many retests each finding gets, how long the window stays open, and whether a retest consumes credits.
P4. Which findings does a human validate before they reach us? A good answer says every finding is confirmed by a tester and rated with a stated scale such as CVSS, with false positives removed before delivery.
P5. What does the platform integrate with, and what do we keep if we leave? A good answer lists the ticketing and chat integrations, confirms that findings export in a standard format, and states that reports and the attestation letter remain yours after the contract ends.
Scoring the responses
Score each question against the rubric from your evaluation criteria and total the themes, not the individual questions. A response that is strong on methodology and reporting but weak on price is usually the better buy. A response that is cheapest and vague on questions 6, 9, 16, and 18 is usually a scan with a cover page. For named vendors and how their models compare, our comparison pages put Invadel next to platforms, crowdsourced programs, scanners, and other testing firms.
RFP scoring matrix: weights by theme
These example weights follow the order in the evaluation criteria. Adjust them, but fix them before the first response arrives. Score each theme from 0 to 5, multiply by the weight, and add the results.
| Theme | Questions | Example weight | What earns a 5 |
|---|---|---|---|
| Methodology and manual depth | 6 to 10 | 30% | Mostly manual effort, exploitation and chaining, a named standard with coverage shown in the report |
| Reporting and retest | 16 to 19 | 25% | Sample report supplied, retest included with a stated window, updated final report |
| Team and credentials | 1 to 5 | 20% | Named employees, a lead with experience in environments like yours, the same team at retest |
| Compliance mapping | 20 to 22 | 15% | Findings mapped to your framework, attestation letter, written independence statement |
| Scope and pricing model | 11 to 15 | 10% | Fixed price for the stated scope, itemized inclusions, a start date tied to your access |
| Insurance and legal | 23 to 25 | Pass or fail | Certificate of insurance, signed NDA and MSA, written data handling terms |
Copy the penetration testing RFP template
Paste this into a document, replace everything in square brackets, and delete what does not apply. Or download the same template as a Word file: penetration testing RFP template (DOCX). The winning proposal then becomes the basis for the statement of work (SOW).
## Request for Proposal: Penetration Testing
### 1. Company background
[Company name] is a [size] [industry] company. Frameworks that apply: [SOC 2 / PCI DSS / HIPAA / ISO 27001 / other].
Reason for testing: [compliance deadline / customer requirement / new product / annual program].
Engagement owner: [name, title, email]. Report recipients: [names or roles].
### 2. Scope
- Web applications: [URL, user roles, purpose, authenticated or not]
- APIs: [number of endpoints, authentication scheme, documentation available]
- Networks: [external hosts; internal hosts, sites, Active Directory domains]
- Cloud: [provider, accounts or subscriptions, services, Kubernetes clusters]
- Mobile: [platforms, backend API included or not]
- Social engineering or red team: [objectives, people or systems in play, constraints]
- Environments: [production / staging]. Testing windows: [dates, hours]. Out of scope: [list].
### 3. Requirements
A compliant response includes: a named testing team, a described methodology with a public standard, a redacted sample report, a fixed price with assumptions listed, retest terms in writing, compliance mapping for [framework], certificates of insurance, and agreement to sign our NDA and MSA.
### 4. Evaluation criteria
Methodology and manual depth [30%], reporting and retest [25%], team [20%], compliance mapping [15%], scope and pricing model [10%]. Insurance and legal terms are pass or fail.
### 5. Timeline and process
RFP issued: [date]. Vendor questions due: [date]. Responses due: [date]. Interviews and sample report review: [dates]. Decision: [date]. Testing window: [dates].
### 6. Questions
Team and credentials
1. Who will perform the testing on our engagement, and what certifications do they hold?
2. Are your testers employees or subcontractors, and where are they located?
3. How much offensive security experience does the assigned lead have, and in what kinds of environments?
4. Can you provide references from clients in our industry or under our compliance framework?
5. Will the same testers be available for our retest and for future engagements?
Methodology and manual depth
6. What share of the engagement is manual testing, and what is automated?
7. Which methodology do you follow, and will the report show coverage against it?
8. How do you test business logic and access control between user roles?
9. Do you attempt exploitation and chain findings, or stop at identification?
10. How do you handle a critical finding discovered during testing?
Scope and pricing model
11. Is the price fixed, hourly, or credit-based, and what could change it?
12. What does the price include?
13. How do you size an engagement, and what do you need from us to do it?
14. What is your lead time to start, and how long will testing run for our scope?
15. Can you test in production, staging, or both, and what safeguards apply?
Reporting and retest
16. Can we see a redacted sample report before we decide?
17. What does the report contain, and is there a separate executive summary?
18. Is a retest included, what is the window, and does the final report reflect the fixes?
19. How are findings delivered: a PDF only, or a platform where we can track remediation and request retests?
Compliance mapping
20. Will the report map findings to our framework?
21. Do you provide an attestation letter we can share with customers and auditors?
22. Are your testers independent of any systems we run, and will you confirm that in writing?
Insurance and legal
23. What professional liability, errors and omissions, and cyber liability insurance do you carry?
24. Will you sign our NDA and master services agreement, and what are your authorization terms?
25. How is our data handled during and after the engagement?
Frequently asked questions
Do we need a formal RFP for a penetration test? Not always. A short request for information with the 25 questions above works for a single engagement. A formal RFP makes sense when procurement requires it or when you are selecting a firm for a multi-year program.
What if a vendor will not answer a question? Treat it as the answer. Every question above has a straightforward response from a firm doing the work properly.
What is the difference between an RFP, an RFI and an RFQ for penetration testing? A request for information (RFI) asks vendors what they do so you can build a shortlist. A request for proposal (RFP) asks a shortlist how they would test your stated scope and at what price. A request for quotation (RFQ) asks only for a price against a scope and method you have already fixed.
How long should a penetration testing RFP be? Two to four pages. The scope section and the questions do the work; long background sections add reading time without changing the answers.
How many vendors should receive a penetration testing RFP? Three to five that you have already screened. Fewer than three gives you nothing to compare, and more than five means scoring responses you were never going to pick.
How long does the RFP process take before testing can start? Plan for about four weeks from issue to decision: one week for vendor questions, one to two weeks for responses, and a week for interviews, sample report review and scoring. Contracting and access setup follow. On our side, onboarding begins within 24 hours of a signed proposal and testing typically starts within a week.
Want to see how we respond to all 25? Get a fixed penetration testing quote and we will return a fixed-scope proposal that answers them, or request a sample report and start with the deliverable.
Written by
Azad Khan, Senior GRC & Compliance
26+ years in the field. Focus: GRC, ISO 27001 & PCI DSS consulting. Part of the team that scopes, tests, and reports for our clients. Full profile →