When a Partner Sends You a Security Questionnaire
A partner security questionnaire is a gate on revenue disguised as a technical exercise. What the other side is actually testing, why optimistic answers are the worst option, and what a small biotech can realistically have in place.
A partner security questionnaire is a structured way for a large pharma company, a CDMO, or a prospective customer to extend its own regulatory and reputational exposure to your company before it trusts you with its data. Partners expect gaps in a small biotech's answers — what they don't accept is inaccuracy, because an inaccurate answer is a contractual representation that comes back as a breach-of-contract problem if an incident later exposes it.
I have filled these in for companies that had no security function and a deadline attached to a contract. The questionnaire looks terrifying and is mostly tractable, provided you answer it honestly and have a few basics in place. What follows is what I tell people who are staring at a 200-question spreadsheet and a date certain.
See where your regulated environment actually stands.
Fifteen questions, about five minutes, scored on screen with a remediation roadmap emailed to you.
What is the partner actually trying to establish?
Reframe it from their side. A pharma partner is extending its own regulatory and reputational exposure to you the moment it shares data, integrates a system, or lists you as a vendor. If you mishandle their data, the finding lands on them — in a regulatory observation, in a breach notification, in a press cycle. The questionnaire is how they reduce that exposure to something they can defend.
They are trying to establish four things: that your access controls are real (not just passwords on a spreadsheet), that your data handling is deliberate (you know where their data sits and who can reach it), that you would detect an incident if one happened, and that you would tell them within a window they can plan around. They are not expecting enterprise maturity from a 40-person biotech. They are expecting that you know the difference between what you have and what you don't, and that you'll say so. The pharmaceutical industry page covers the regulatory backdrop that makes partners this careful.
Why answering optimistically is the worst option
This is the central point and the one most companies get wrong under deadline pressure. An answer on a security questionnaire is a contractual representation. If an incident later reveals the answer was wrong — you said you had MDM on all endpoints and an investigation shows three laptops weren't enrolled — it moves from a security problem to a breach-of-contract problem, and it poisons everything else you represented. The partner now has grounds to question every other answer, and the relationship is in a worse place than if you had been honest from the start.
"No, and here is our plan and date" is routinely accepted. An overstated yes is not recoverable. Reviewers who handle vendor assessments for a living have seen every version of optimistic answering, and they discount it accordingly. What they can work with is a clear statement of where you are, what you're doing about it, and when it will be done. This is the same principle that governs IT due diligence: the gaps are not the problem, the misrepresentation is.
What the questions actually cluster into
A 200-question spreadsheet looks like 200 separate problems. It is really eight or so things asked in different ways:
- 1.Identity and access management. Who can access what, how is access granted and removed, is multi-factor authentication in place, how are privileged accounts controlled.
- 2.Endpoint security. What's on the laptops — antivirus or EDR, disk encryption, patch management, mobile device management.
- 3.Data classification and handling. Do you know what data you hold, where it lives, how it's encrypted in transit and at rest, and how it's disposed of.
- 4.Third-party and subprocessor management. Who else touches the data — your cloud provider, your SaaS tools, any subcontractors — and do you have agreements with them.
- 5.Business continuity and recovery. Backups, recovery time objectives, and whether you've actually tested restoration.
- 6.Incident response and notification obligations. Would you detect an incident, do you have a plan, and how quickly would you tell the partner.
- 7.Secure development, if you build anything. Code review, vulnerability scanning, separation of environments, how you handle open-source dependencies.
- 8.Physical security. Office access controls, device security in transit, whether equipment is secured when unattended.
What an honest small-company answer looks like in each cluster is simpler than people fear. For identity and access: yes, we use a single identity provider, MFA is enforced, access is reviewed quarterly, and privileged accounts are limited to two people. For endpoint: yes, we have EDR on all company devices, disk encryption is enforced by policy, and we patch on a monthly cycle. For the rest, a mix of yes, partially, and "not yet — here is our plan." That mix is normal and acceptable.
What you need before you can answer anything
There is a practical blocker that catches companies before they've answered a single question. You can't answer a security questionnaire accurately — even with good intentions — without a few things in place:
- A current system inventory. What systems you run, what they hold, where they're hosted. You cannot answer "where is our data stored" if you don't know what systems you have.
- A data flow understanding. What of the partner's data you'd hold, where it would sit, who would access it, and what would leave your environment. This is the question partners care about most, and "we'll figure that out" is not an answer.
- A named person responsible for security. Not a full-time CISO — a person whose name goes on the answers and who understands what they mean. At a small company this is often an operations lead or an external partner.
- Any written policy at all. An acceptable use policy, an access control policy, an incident response outline. One-page policies that reflect what you actually do are better than nothing and better than borrowed boilerplate you can't follow.
Companies without these can't answer the questionnaire accurately even with good intentions. The first step is not filling in the spreadsheet. It's building the picture the spreadsheet asks for.
Do we need SOC 2 or ISO 27001?
Everyone asks this, usually because a partner mentioned it or a competitor has one. The direct answer: usually not at this stage, and pursuing one under a questionnaire deadline is the wrong response.
SOC 2 is an AICPA attestation framework focused on the Trust Services Criteria — security, availability, processing integrity, confidentiality, and privacy. A Type I report describes your controls at a point in time; a Type II describes them operating over a period, typically six to twelve months. ISO/IEC 27001 is an international standard for an information security management system — a broader, certifiable framework that covers how you establish, maintain, and continually improve security across the organization. Both are real achievements and both take real time: a SOC 2 Type I can take a few months with focused effort, a Type II needs the controls operating for the observation period, and ISO 27001 is a larger undertaking that typically spans the better part of a year for a first-time implementation.
Neither is worth pursuing under a contract deadline. The partner isn't asking for the certification — it's asking the questions. What you can offer instead is a completed questionnaire with honest answers and a remediation plan, plus any controls evidence you already have. If a partner eventually requires a SOC 2 or ISO 27001, that's a decision to make on its own timeline with budget and leadership buy-in, not a panic purchase. The NIST Cybersecurity Framework is a useful free reference for structuring your controls while you decide — it is not a certification, it is a way to think about what you have.
Frameworks like the SIG (Standardized Information Gathering questionnaire, maintained by Shared Assessments) and CAIQ (Consensus Assessments Initiative Questionnaire, maintained by the Cloud Security Alliance) exist precisely because the same questions get asked in different formats. Knowing which one a partner used tells you something about how their third-party risk program is structured, but the answers underneath are the same.
What to do with the questions you fail
A remediation plan with owners and dates is a legitimate response and often satisfies the partner. The key is triage:
- Cheap and fast gaps. Enforcing MFA on a system that should have had it, writing down the policy you already follow, adding a backup test to the calendar. Do these now and answer "done" with a date.
- Genuinely material to the partner's risk. Endpoint coverage on devices that touch their data, a signed agreement with a subprocessor, a tested restore of the system holding their data. These need a plan with a date and an owner, and they need to be real.
- Reasonable to defer. A formal ISO 27001 program, a 24/7 SOC, capabilities that are genuinely disproportionate to your size and the data you hold. State that you don't have them and explain why; most partners will accept this from a small company if the basics are solid.
The format that works: for each gap, "We do not currently [X]. We plan to [Y] by [date], owned by [person]." That sentence is accepted far more often than people expect.
What happens after you submit
This part is often overlooked. Submitting the questionnaire is not the end. Expect follow-up questions, evidence requests — screenshots, policy documents, configuration exports — and sometimes a call with the partner's security team. Expect annual re-attestation, where they send the questionnaire again and expect updated answers. And expect the contractual obligations that follow: notification windows (tell us within X hours of a confirmed incident), audit rights (we may review your controls annually), and subprocessor approval (you need our consent before engaging a new vendor that touches our data). These now bind you. Read them before you sign, because they are the real commitments — the questionnaire was the interview.
How to stop starting from zero every time
The strategic point. If you will receive more than one questionnaire — and any company with several partners will — maintain a standing answer set and an evidence pack, kept current, so the next questionnaire is an editing job rather than a fire drill. The answer set is a document with your standard responses to the common questions, written once and reviewed quarterly. The evidence pack is a folder with your current policies, screenshots of key configurations, your system inventory, and your data flow diagram, kept up to date.
For a company with several partners, this is the difference between a week and an afternoon. It also makes the answers better, because you wrote them when you had time to think, not when you were under a deadline and inclined to be optimistic. Cybersecurity and data protection is not a one-time exercise; it is a standing practice, and the questionnaire is a chance to build that practice deliberately.
What they ask, what they're really testing, and an honest answer
| What they ask | What they're really testing | An honest answer for a company your size |
|---|---|---|
| "Do you enforce multi-factor authentication?" | Are access controls real or theoretical? | "Yes, on all systems with external access, via our identity provider. Privileged accounts require hardware keys." |
| "Do you have endpoint detection and response?" | Would you catch malware before it spreads? | "Yes, EDR on all company devices with centralized alerting. We do not have a 24/7 SOC." |
| "Is data encrypted at rest and in transit?" | Do you know where the data lives and how it's protected? | "Yes. At rest via cloud provider encryption; in transit via TLS. We hold a data flow diagram showing where partner data resides." |
| "Do you have an incident response plan?" | Would you know what to do, and would you tell us? | "Yes, a written plan with defined roles. Notification to partners within 72 hours of a confirmed incident. Not yet tested via a full tabletop exercise." |
| "Do you perform background checks on employees?" | Is the people risk considered? | "Yes, reference and background checks for new hires with system access. We do not use a third-party screening service for every role." |
| "How do you manage third-party and subprocessor risk?" | Do you know who else touches the data? | "We maintain a vendor inventory and require DPAs with subprocessors that handle sensitive data. Formal annual review of all vendors is not yet in place." |
| "Do you have a business continuity plan?" | Could you actually keep operating and recover data? | "Yes, backups are daily and stored in a separate region. We have tested restoration of critical systems; RTO is defined for partner-facing systems." |
| "Do you have SOC 2 or ISO 27001 certification?" | Have you invested in formal assurance? | "Not currently. We have completed this questionnaire with a remediation plan. We are evaluating SOC 2 for next year based on partner requirements." |
| "Do you conduct regular security training?" | Is the human layer considered? | "Yes, onboarding training and annual refresher. We do not currently run phishing simulations." |
| "How do you dispose of data when a contract ends?" | Is the end of the relationship handled? | "Per our data retention policy, partner data is deleted within 30 days of contract termination, with written confirmation. Process documented and followed." |
The reframe
A security questionnaire is not a test you pass or fail. It is the start of a relationship in which you have now made written commitments — to notify, to maintain controls, to manage who touches the data, to tell the truth about what you have. That is why the answers have to be true. The partner is not looking for perfection; it is looking for a vendor it can defend if something goes wrong later. A company that answers honestly, with a remediation plan for its gaps and the discipline to maintain it, is that vendor. A company that answers optimistically to get through the form is a liability that neither side has properly priced.
Photo by Kelly Sikkema on Unsplash.
Frequently asked questions
Don't see your question? Email info@isectra.com and we'll answer it directly.

Brian has filled in security questionnaires for companies that had no security function and a deadline attached to a contract. He writes for leaders staring at a 200-question spreadsheet and a date certain.
Get the next post in your inbox.
Signal, once or twice a month. No spam — just the posts worth reading.
Let's see what your IT could feel like.
A senior engineer will review your environment and respond within 24 hours. No sales script, no obligation.

