MNEMARK   Independent thinking for a changing world
Research / Perspective / Guides

From the journal

Ideas worth understanding.

The Small-Team SaaS Security Questionnaire: 35 Questions Before You Sign

The Small-Team SaaS Security Questionnaire: 35 Questions Before You Sign
On this page

    Choosing a SaaS product is not only a feature decision. It is also a decision about who can access your information, how long the provider keeps it, what happens during an outage, and how quickly your team can respond when something goes wrong.

    This 35-question SaaS security questionnaire is designed for a small team that needs useful evidence without turning a vendor review into a six-month procurement project. It is not a certification checklist or legal advice. Use it to structure a conversation, identify missing evidence, and decide which risks need contract, technical, or leadership review.

    Small team reviewing a SaaS vendor security questionnaire on a laptop
    A practical vendor questionnaire connects security evidence to the way a small team will actually use the SaaS product.

    Start with scope, ownership, and data

    Before sending questions to a provider, write down what you are buying and what the service will touch. A vendor may be low risk for public project tracking but high risk when it stores customer records, source code, credentials, or regulated information. Your own scope prevents a generic security packet from becoming a substitute for judgment.

    1. What business process will the service support, and what happens if it is unavailable for one day?
    2. Which data classes will enter the service: public, internal, confidential, personal, or regulated?
    3. Will the provider process production data, test data, or both?
    4. Which employees, contractors, customers, or partners will have accounts?
    5. Who owns the vendor relationship and who approves security exceptions?
    6. Which integrations will connect the service to identity, storage, repositories, email, or payment systems?
    7. What is the planned exit date or review date if the product no longer fits?

    These questions also expose hidden dependencies. If the service becomes part of login, deployment, support, or billing, its failure and compromise scenarios deserve more attention than a standalone convenience tool.

    Ask for security evidence, not reassuring language

    A vendor may say that security is a priority. That statement is not useless, but it is not evidence. Ask for the document, control description, test date, scope, and owner behind the claim. NIST’s supply-chain due diligence guidance describes supplier assessment as an investigation of available information and highlights areas such as provenance, resilience, foundational cyber practices, and supply-chain tiers.

    1. Which independent security assessment or audit report covers the service?
    2. What product, environment, and legal entity are included in the report’s scope?
    3. What period does the report cover, and when was it last updated?
    4. Which important controls are outside the report’s scope?
    5. Can the vendor explain unresolved findings, exceptions, or complementary user controls?
    6. Does the vendor provide a security whitepaper, trust center, or control matrix with named owners?
    7. Which security claims are commitments, and which are only product intentions or roadmap items?

    A SOC report can provide useful assurance information for outsourced services, but it is not a universal security guarantee. The report must be read in context: check the service boundary, period, control objectives, exceptions, and the responsibilities left to your team. For a broader assessment, the Cloud Security Alliance’s Cloud Controls Matrix and CAIQ provide a structured set of security domains and yes-or-no questions that can help organize follow-up questions.

    For a deeper explanation of how to interpret evidence before adoption, see our guide to evaluating a SaaS vendor’s security evidence.

    Review identity, access, and administrative control

    Most SaaS risk is not determined by the marketing page. It is determined by who can enter the system, what administrators can change, and how quickly access can be removed. Prefer a provider that supports a clear identity lifecycle and makes privileged actions visible.

    1. Does the service support SSO with your chosen identity provider?
    2. Can you enforce phishing-resistant MFA or at least strong MFA for administrators?
    3. Can the provider separate ordinary users, support staff, and privileged administrators?
    4. Are administrator actions recorded in an audit log that your team can review?
    5. Can you export logs, and how long are they retained?
    6. Can access be removed promptly when an employee or contractor leaves?
    7. Does customer support use controlled, time-limited access instead of permanent broad access?

    Review every OAuth scope and API token connected to the product. A service that only needs to read calendar availability should not receive permission to manage an entire workspace. Our practical guide to auditing an AI agent’s permissions applies the same least-privilege reasoning to connected tools.

    Check data handling, retention, and recovery

    Data security includes the full lifecycle, not only encryption during a normal session. Ask how the provider stores, copies, searches, exports, and deletes information. Request a concrete answer for backups and derived data, because deleting a primary record may not delete every copy immediately.

    1. Is customer data encrypted in transit and at rest, and what exceptions exist?
    2. Where is data stored and which subprocessors can access it?
    3. Is customer data used to train or improve a shared model or service?
    4. How are production and test environments separated?
    5. What is the retention period for primary data, logs, backups, and support tickets?
    6. Can the customer export data in a usable format before termination?
    7. How does deletion work across primary storage, replicas, backups, and subprocessors?
    8. What recovery point and recovery time objectives apply to the service?

    Do not accept “we use backups” as the end of the discussion. Ask when recovery was last tested, what customers should do during an outage, and which features degrade first. If your team stores secrets or credentials in the service, remove that use case unless the provider’s controls and recovery model are appropriate. For operational handling, see our guide to rotating API keys without breaking a small team’s workflow.

    Prepare for incidents and the exit decision

    A vendor review is incomplete if it only describes normal operation. Your team needs to know how the provider communicates a security incident, what evidence it will share, and how you can leave without losing control of your information.

    1. How will the provider notify customers about a suspected or confirmed incident?
    2. Which contact receives urgent security notices outside business hours?
    3. What evidence can the provider share after an incident, such as timelines or affected systems?
    4. How does the provider coordinate with customers, investigators, and regulators where applicable?
    5. What happens to data and integrations when the contract ends?
    6. Can the provider demonstrate deletion or return of data after termination?

    Record unanswered questions as risks instead of silently converting them into approval. A small team can assign each risk an owner, due date, treatment, and acceptance authority. If a vendor cannot explain its boundary, access model, notification process, or exit path, pause the rollout until the uncertainty is understood.

    Turn the answers into a decision

    Do not score every answer as if it has equal importance. A missing public security page may be inconvenient; an unbounded administrator account connected to customer data may be a release blocker. Group findings into four outcomes:

    • Accept: the evidence and remaining risk fit the intended use.
    • Mitigate: add a technical control, contract term, monitoring step, or narrower data scope.
    • Defer: the service may be useful later, but essential evidence is missing.
    • Reject: the provider cannot meet a non-negotiable security, privacy, recovery, or exit requirement.

    Review the questionnaire again when the data classification, integration set, provider ownership, or contract changes. Security evidence has a date, a scope, and a limit. The goal is not to make a SaaS vendor prove that risk is impossible. The goal is to make the remaining risk visible enough for the right person to decide.

    Sources

    PASS IT FORWARD

    Found this useful? Share it with your network.

    Filed under
    Original MNEMARK editorial work. Brief quotations require clear attribution and a direct source link; republication, bulk extraction, and full-article translation require written permission.Copyright & DMCA policy

    Comments