A supplier hands you a clean SOC 2 Type 2 report on the same day your rating platform shows its score falling, and the third-party risk team has to decide which one to believe. Both can be right, because they look at different things over different windows: the SOC 2 report describes tested controls inside a defined system boundary for a past period, and the security rating describes what the supplier's internet-facing footprint looks like now. This guide is about reading the two side by side for one vendor: where the report's scope, carve-outs and bridge letter leave gaps, which of those gaps an outside-in rating can and cannot fill, and how to turn both into a tiered decision. It is for TPRM analysts, procurement security reviewers and audit teams. For the wider question of ratings against questionnaires and assessments in general, read Security Ratings vs Vendor Assessments; for a section-by-section walk through a SOC 2 report, read the How to Read a SOC 2 Report Guide. Sources were checked on 14 November 2026.
Key takeaways
- The report and the rating describe different objects. A SOC 2 report covers a named system, chosen trust services categories and a past period; a rating covers assets a provider attributes to the supplier's organisation, today.
- Scope is the first thing to check, not the opinion. An unqualified opinion on a system that does not include the service you buy tells you almost nothing about that service.
- Carve-outs and bridge letters are where assurance stops. Carved-out subservice organisations were not tested, and a bridge letter is a management statement with no auditor involvement.
- A rating cannot rescue a weak report, or sink a strong one. Use rating changes to ask targeted questions about the gap periods and the untested parts, never to override audited evidence.
- Tier decides how hard you push. Critical suppliers need a current, correctly scoped report, the subservice chain and a rating watch; low-tier suppliers may need only one of the two.
The short answer
Treat the SOC 2 report as your primary evidence about how the supplier operates the service, and the rating as a continuous, shallow signal about whether its external estate is being looked after between reports. When they disagree, the disagreement is a question to put to the supplier, not a verdict. Read the report's scope first, map every untested area (carved-out subservice organisations, complementary user entity controls you own, the gap since the period ended), then decide whether any rating findings sit in those areas. US banking regulators' 2023 Interagency Guidance on Third-Party Relationships says banks may review SOC reports where relevant and available, and should check that the scope and results of those reports match the planned activity. That scope check is the heart of this page.
| Dimension | Security rating (outside-in) | SOC 2 report (audited controls) |
|---|---|---|
| Who produces it | A commercial rating provider, from observed data | An independent CPA firm, under AICPA attestation standards |
| Object | Internet-facing assets attributed to the organisation | A defined system and its controls, described by management |
| Time window | Now, refreshed continuously | A point in time (Type 1) or a past period (Type 2) |
| What it tests | Observable hygiene: exposed services, certificates, email authentication, patch signals, leaked data | Design and, for Type 2, operating effectiveness of controls against the trust services criteria |
| Supplier involvement | None needed; disputes are optional | Supplier commissions, scopes and pays for the examination |
| Main blind spot | Internal controls, identity, cloud tenant settings, how your data is handled | Anything outside the system boundary, carved-out parties, and the time since the period ended |
| Typical failure | Mis-attributed assets and noisy findings | Scope that misses the service, template reports, stale periods |
| Best use | Change detection and triage across many suppliers | Decision on a specific relationship and its contract terms |
Threat model: what each one is built to catch
The security rating
Threats it addresses: neglect and drift on the external estate. An expired certificate, an exposed remote access service, a missing DMARC policy or a burst of leaked credentials are visible from outside and correlate with an organisation that is struggling to keep up. A rating catches change quickly, which is its real value: a sudden drop is a reason to ask what happened.
Threats it misses: almost every route that hurts you through a SaaS supplier. Identity provider configuration, MFA coverage for admins, how API keys and OAuth tokens issued to your tenant are stored, backup isolation, logging and change control are invisible from the internet. A supplier with a strong external score can still hold your data behind a support account without MFA.
Fairness limits: the US Chamber of Commerce's Principles for Fair and Accurate Security Ratings call for transparency of methodology and data sources, a dispute, correction and appeal process with disputed ratings marked as such, validation of models against past performance, notice before methodology changes, and independence from commercial relationships. Ask any rating provider how it meets each one before you let its score drive decisions.
The SOC 2 report
Threats it addresses: control failure inside the service. A SOC 2 is an examination of a service organisation's system description and the controls relevant to security, availability, processing integrity, confidentiality or privacy, as the AICPA describes it. A Type 2 report adds a test of whether the controls operated effectively across the period, which is what lets you rely on it for access reviews, change management and incident response.
Threats it misses: anything the supplier left out of the boundary, anything a carved-out subservice organisation runs, controls the supplier expects you to operate, and everything after the period end date. It also misses quality problems in the report itself. The Journal of Accountancy reported in February 2026 on concern that fast, cheap, highly automated engagements may rely too heavily on inquiry, producing reports that read like templates. A report is only as good as the work behind it; see the SOC 2 Auditor Selection page for what good looks like.
Reading SOC 2 scope against the rating
The rating provider attributes assets to the supplier's whole organisation. The SOC 2 report describes one system. Before comparing them, line them up.
- Name the service you buy. Write down the product, region and hosting model in your contract. A supplier may run several platforms with different reports.
- Check the system description. Section III of a typical report (management's description) lists infrastructure, software, people, procedures and data. Confirm your product, region and data flows are inside it.
- Check the trust services categories. Security (the common criteria) is always in scope. Availability, confidentiality, processing integrity and privacy are optional. If you rely on uptime, an availability-free report gives you no assurance on it; see the SOC 2 Availability Criteria Guide.
- Check the type and period. Type 1 covers design at a date; Type 2 covers operating effectiveness over a period, commonly six to twelve months. The SOC 2 Type 1 vs Type 2 page explains why a Type 1 is weak evidence for a critical supplier.
- Map the rating's assets to the boundary. For each high-severity rating finding, ask whether the asset is inside the described system, part of corporate IT, a marketing site, an acquired subsidiary or not the supplier's at all.
| Where the rating finding sits | What it tells you | Action |
|---|---|---|
| Inside the SOC 2 system boundary | A possible control failure the report did not see, or one that started after the period | Ask for the related control, the ticket and the fix date; raise it at the next report review |
| Corporate IT (email, VPN, staff SaaS) | Hygiene of the estate that staff use to reach the service; out of report scope but relevant to account takeover | Ask about admin access paths into production and MFA for support staff |
| Acquired subsidiary or separate product | Usually unrelated to your service | Confirm separation in writing; lower priority |
| Mis-attributed (not the supplier's asset) | Noise | Dispute with the rating provider; record the outcome |
| Carved-out subservice organisation | Something in a party the auditor did not test | Get that party's own report or other assurance |
Carve-outs, CSOCs and CUECs: the untested parts
Carve-out method: the subservice organisation (often a cloud provider, data centre or payments processor) is named and its services described, but its controls are excluded and not tested, and the opinion covers only the service organisation's own controls. Under the inclusive method the subservice organisation's controls are in scope and tested; it needs that party's agreement and access, so large infrastructure providers rarely take part.
CSOCs (complementary subservice organisation controls) are the controls the supplier expects the carved-out party to run. They are a disclosure, not a tested assertion. CUECs (complementary user entity controls) are the controls you, the customer, are expected to run for the criteria to be met. They appear in every report.
A rating rarely sees carved-out parties as part of the supplier: it scores the hyperscaler or processor separately, if at all. So the carve-out is a gap the rating does not fill. Close it with the subservice organisation's own report, checked for period and scope against the CSOCs, and escalate if there is no report or it has relevant exceptions. For mapping the chain beyond one tier, see the Fourth-Party and Nth-Party Risk Guide.
CUECs are a gap neither source fills: they are yours. Typical ones are user provisioning and removal in your tenant, MFA for your own users, protecting API credentials the service issues to you and reviewing the audit logs it exposes. The Snowflake breach shows the pattern: investigators found no compromise of the platform itself, and the affected customer accounts had no MFA and unrotated credentials, which were customer-configurable controls. No supplier report or rating would have flagged them.
Bridge letters and the gap since the period ended
A Type 2 report covering January to December, delivered in March, is already three months old on arrival. A bridge letter (also called a gap letter) is a short statement signed by the supplier's management that refers to the last report and says that, to management's knowledge, the control environment has not changed significantly. The service auditor does not prepare, sign or review it. Grant Thornton notes that it carries no independent testing or opinion, should not be used after significant control changes, and that an auditor would generally not support one covering more than about three months.
This is the window where a rating earns its keep. Pull the rating history for the gap period and compare it with the bridge letter's claim of no significant change:
- Stable rating, short gap, bridge letter received: accept for the current cycle and diary the next report date.
- Material drop during the gap: ask what changed before accepting the bridge letter. A new exposed service or a breach-related signal in the gap contradicts "no significant change" until explained.
- Gap longer than about three months, or repeated bridge letters: treat the report as stale. Ask for an interim report or an earlier audit cycle, and raise oversight for the relationship.
- Known incident at the supplier: a bridge letter is not enough. Use your incident questions and contract notice terms; see the vendor breach response playbook.
Failure modes for each, and the control
| Failure | Source | How it shows up | Control |
|---|---|---|---|
| Scope miss | SOC 2 | Report covers a different product, region or platform | Match contract service to system description before reading the opinion |
| Category miss | SOC 2 | Security only, but you rely on availability or confidentiality | Require the categories you depend on in the contract |
| Template report | SOC 2 | Generic controls, thin test procedures, mostly inquiry | Review test procedures and the CPA firm's peer review and capacity |
| Stale evidence | SOC 2 | Old period, long or repeated bridge letters | Track period end dates; set a maximum gap per tier |
| Hidden exceptions | SOC 2 | Exceptions in access reviews or change control treated as minor | Read Section IV exceptions and management responses for your risks |
| Mis-attribution | Rating | Findings on assets the supplier does not own | Confirm ownership before raising any finding; use disputes |
| Score as proxy for safety | Rating | High score used to skip evidence for a critical supplier | Policy: a rating never lowers a tier or replaces evidence |
| Methodology change | Rating | Scores shift across the portfolio overnight | Require notice of model changes; annotate trend charts |
Combining them in third-party risk tiers
The Interagency Guidance frames oversight as commensurate with the risk of each relationship, with more rigorous oversight for critical activities. NIST SP 800-161 Rev. 1 places supplier assessment inside cybersecurity supply chain risk management. In practice, set a minimum evidence package per tier. Tiering itself is covered in the Vendor Tiering and Inherent Risk Questionnaire Template.
| Tier | SOC 2 expectation | Rating use | Gap rules |
|---|---|---|---|
| Critical (sensitive data, privileged access, or operations depend on it) | Type 2, scope matched to service, categories you rely on, exceptions reviewed, subservice reports obtained for carve-outs | Continuous watch with alert on material drop or new critical finding in the system boundary | Bridge letter accepted for up to about three months; longer needs interim evidence |
| High | Type 2 or equivalent attestation, scope checked, CUECs mapped | Watch; review on drop | Bridge letter plus rating check for gap |
| Moderate | Type 2 accepted on scope check; Type 1 with a plan to Type 2 | Periodic review | Annual refresh |
| Low | Not required | Optional, portfolio level | None |
Small versus large programmes: a small team should spend effort on scope matching and CUECs for its few critical suppliers and use ratings mainly as an alarm. A large programme can automate report intake (period dates, opinion, exceptions, carve-outs as structured fields) and join them with rating feeds, so that a rating alert on a critical supplier automatically shows the report's period, the bridge letter date and the carved-out parties.
Worked scenario: one supplier, two signals
A customer support SaaS supplier is tiered critical: it holds customer contact data and an integration token with write access to your CRM.
- Report intake: SOC 2 Type 2, security and confidentiality, period 1 October to 30 September, unqualified opinion, two exceptions in quarterly access reviews. A cloud provider and an email delivery service are carved out. CUECs include protecting API credentials and reviewing admin activity logs.
- Scope check: the system description covers the EU region you use. Availability is not in scope, so uptime assurance comes from the contract SLA only.
- Carve-outs: you obtain the cloud provider's current report and confirm its period covers yours. The email service has no report available; you record the gap and ask the supplier how it monitors that party.
- Gap: in February you receive a bridge letter for October to December. The rating dropped in January after a new remote desktop service appeared on a host attributed to the supplier.
- Reconcile: the supplier confirms the host is a support jump server inside the production boundary. That contradicts the bridge letter. You ask for the change record, the remediation date and confirmation of MFA on the jump server, and link it to the access review exceptions.
- Your side: you rotate the CRM integration token, scope it down and enable log export, meeting the CUECs. See the third-party NHI due diligence questions for the token questions.
- Decision: relationship continues with a dated remediation commitment and a requirement for the next report to test the jump server control.
Where identity fits
The highest-impact supplier failures of recent years ran through identities: support accounts, integration tokens and OAuth grants. SOC 2 logical access criteria (CC6) are where these sit; read the SOC 2 CC6 Logical Access Guide for what to look for in tests and exceptions. Ratings see almost none of it, beyond leaked credential signals. Ask specifically how the supplier stores tokens it holds for your tenant, how its staff reach production, and whether those paths sit inside the report boundary. The Salesloft Drift breach showed how stolen OAuth tokens from one integration reached many customer tenants, a risk that lives in CC6 and in your own CUECs rather than in any external score.
Detection and measurement
- Scope match rate: share of critical suppliers whose current report is confirmed to cover the contracted service and region.
- Evidence currency: share of critical suppliers with a report period ending within the last 12 months, plus bridge letter cover no longer than your tier limit.
- Carve-out coverage: share of carved-out subservice organisations for critical suppliers with their own current report on file.
- CUEC implementation: share of CUECs for critical suppliers mapped to an owner and evidenced in your environment.
- Rating alert closure: median days from a material rating drop on a critical supplier to a recorded supplier explanation.
- Dispute accuracy: share of rating findings raised with suppliers that turned out to be mis-attributed (falling over time means your ownership check works).
Framework mapping
| Framework | Reference | Relevance |
|---|---|---|
| NIST CSF 2.0 | GV.SC-06, GV.SC-07, ID.RA-10 | Due diligence before contracting, ongoing supplier risk assessment, assessing critical suppliers before acquisition |
| NIST SP 800-53 Rev. 5 | SR-6 Supplier Assessments and Reviews; SA-9 External System Services | Assessing suppliers and defining oversight of external services |
| NIST SP 800-161 Rev. 1 | C-SCRM practices | Supplier assessment within supply chain risk management |
| ISO/IEC 27001:2022 Annex A | 5.19 to 5.22 | Supplier relationships, agreements, ICT supply chain, and monitoring and review of supplier services |
| CIS Controls v8.1 | Control 15 Service Provider Management | Inventory, classification, assessment and monitoring of service providers |
| AICPA trust services criteria | CC6, CC9.2 (vendor risk), plus chosen categories | What a SOC 2 report actually tests |
Decision checklist
- Does the report's system description include the exact service, region and data flow you contract for?
- Are the trust services categories you rely on in scope?
- Is it a Type 2, and does the period end within your tier's currency limit?
- Do the exceptions touch access, change or incident controls that matter to you?
- Do you hold current reports for every carved-out party that matters?
- Is any bridge letter short, and is the rating stable across the gap?
- Have you confirmed ownership of every rating finding you plan to raise?
- Have you implemented and evidenced the CUECs on your side?
Practitioner checklist
- Record service, region, categories, type, period, opinion, exceptions and carve-outs for every critical supplier report as structured data.
- Read scope before the opinion, every time.
- Obtain subservice reports for carve-outs that touch your data or access.
- Assign an owner to every CUEC and keep evidence.
- Set a maximum bridge letter period per tier and refuse back-to-back letters for critical suppliers.
- Join rating alerts to report data so analysts see the gap period and carve-outs immediately.
- Never let a rating lower a tier or stand in for a report.
- Check the CPA firm and test procedures when a report looks generic.
Standards and references
- AICPA: System and Organization Controls (SOC) suite of services
- AICPA: SOC 2 Reporting on an Examination of Controls at a Service Organization
- Federal Reserve, FDIC and OCC: Interagency Guidance on Third-Party Relationships: Risk Management
- NIST: SP 800-161 Rev. 1, Cybersecurity Supply Chain Risk Management Practices
- US Chamber of Commerce: Principles for Fair and Accurate Security Ratings
- Journal of Accountancy: Promises of fast and easy threaten SOC credibility
- Grant Thornton UK: How bridge letters help mind the gap between audit periods
- Compass IT Compliance: Subservice organizations in SOC reports, carve-out vs inclusive method
Related NHI Mgmt Group resources: Third-Party Risk Management Guide · TPRM and Security Ratings Buyer's Guide · SOC 2 Guide · SOC 1 vs SOC 2 vs SOC 3 · Attack Surface Management Guide