SOC2Scout
SOC2Scout
DirectoryMatch WizardCompareGuidesFor AuditorsGet Matched Free
Home/Qualified Opinion/Control Failures

The 10 Most Common SOC2 Control Failures (and How to Remediate Them)

These 10 controls fail with disproportionate frequency in SOC2 Type 2 audits. For each: what the failure looks like in practice, why auditors flag it, how to remediate it, and a realistic timeline.

Updated: March 2026

Most SOC2 qualified opinions trace back to a small set of control categories. They are not obscure or technically complex — they fail because they require consistent operational execution over 12 months, not just a policy document. Access reviews, offboarding, change approvals: these are calendar-dependent controls that slip when nobody owns them explicitly. The list below covers the failures by their Trust Services Criteria reference, in order of how frequently auditors cite them.

01
CC6.3

Access Review Failures

WHAT IT LOOKS LIKE

User access is not reviewed on the schedule defined in your access control policy — typically quarterly. The auditor finds no evidence that access reviews were performed, or finds them performed only once or twice in a 12-month observation period.

WHY AUDITORS FLAG IT

Auditors test this by requesting evidence of each review — typically a dated export from your identity provider showing accounts reviewed, with a sign-off. If the evidence is missing for any review period, it is an exception. This is the single most common failure because it requires consistent operational execution, not just a policy document.

HOW TO REMEDIATE

Perform access reviews on the defined schedule and retain evidence: a dated CSV export of active user accounts, a comparison against HR records, and a documented approval from the access review owner. Automate the evidence collection with your identity provider (Okta, Azure AD) to generate a timestamped export.

TIMELINE:30–60 days to implement the process and gather 1–2 documented reviews
02
CC6.1

Terminated Employee Access Not Removed Promptly

WHAT IT LOOKS LIKE

Former employee accounts remain active in your production systems, cloud infrastructure, or SaaS tools after their termination date. The auditor cross-references your HR termination log against access provisioning records and finds accounts that were not revoked within your defined SLA.

WHY AUDITORS FLAG IT

This is both a process failure and a high-severity security finding. Auditors pull a sample of terminations from the period and verify that access was revoked within the timeframe specified in your offboarding policy (typically 24–48 hours). Even one miss in the sample can generate an exception.

HOW TO REMEDIATE

Integrate HR offboarding with IT access revocation. The minimum: HR notifies IT on termination date, IT revokes access and logs the ticket, and both are retained as paired evidence. Automate with tools like Okta Lifecycle Management or Rippling if volume warrants it.

TIMELINE:2–4 weeks to implement the process; 1–3 months of clean execution before re-audit
03
CC8.1

Change Management Without Documented Approval

WHAT IT LOOKS LIKE

Code or infrastructure changes are deployed to production without documented approval from an authorized reviewer. The auditor reviews a sample of production deployments and finds commits merged without pull request approvals, or tickets closed without sign-off.

WHY AUDITORS FLAG IT

Auditors look for evidence that every production change went through an approval step before deployment. The evidence is typically: PR approval history in GitHub/GitLab, deployment pipeline configuration showing required approvals, or JIRA/Linear ticket approval status. Hotfixes pushed directly to main without a PR are common exception triggers.

HOW TO REMEDIATE

Enforce branch protection rules requiring at least one approval on all pull requests to main/production. Document the configuration. Retain deployment logs showing the approval chain. For infrastructure changes, require approval tickets in your ITSM tool.

TIMELINE:1–2 weeks to configure branch protection; immediate effect on subsequent deployments
04
CC3.1

Missing or Incomplete Risk Assessment

WHAT IT LOOKS LIKE

The annual risk assessment was not performed, or was performed but lacks documentation of identified risks, risk ratings, and risk treatment decisions. Auditors ask for the most recent formal risk assessment and find either no document or a document that was never reviewed and approved.

WHY AUDITORS FLAG IT

CC3.1 requires that management perform a risk assessment to identify risks to the achievement of objectives. Auditors look for: a dated risk register, risk owner assignments, risk severity ratings, and evidence that management reviewed and approved the assessment.

HOW TO REMEDIATE

Perform a formal risk assessment annually. Use a template if you do not have one: list all assets, identify threats and vulnerabilities, assign likelihood and impact ratings, determine treatment (accept, mitigate, transfer, avoid), and get sign-off from a named owner. Date and version the document.

TIMELINE:2–4 weeks to complete initial assessment; ongoing annual cadence
05
CC1.4

Security Awareness Training Lapses

WHAT IT LOOKS LIKE

Not all employees completed annual security awareness training during the observation period. The auditor requests an LMS completion report and finds employees hired during the year who completed training late, or long-tenured employees who missed the annual cycle.

WHY AUDITORS FLAG IT

CC1.4 requires that the organization communicate information security responsibilities to personnel. Auditors test this with an LMS (or equivalent) export showing every active employee's training completion date. Any active employee with no training record for the period is an exception.

HOW TO REMEDIATE

Enroll new hires in security awareness training within their first 30 days and document completion. Run annual training for all employees on a calendar cycle. The LMS should retain dated completion records per employee. If you use a manual approach (PDF acknowledgment), retain signed records.

TIMELINE:Immediate for new hires; full compliance within 90 days if current employees are behind
06
CC7.4

Inadequate Incident Response Documentation

WHAT IT LOOKS LIKE

The incident response policy exists, but there is no evidence that it was ever exercised. No incident log entries exist (even for minor events), no tabletop exercise was conducted, and no post-incident reviews are documented.

WHY AUDITORS FLAG IT

An empty incident log is itself a finding — auditors are suspicious that incidents occurred but were not logged, or that the incident response process is theoretical rather than operational. CC7.4 requires detection, containment, and recovery processes. Evidence of execution is required, not just the policy.

HOW TO REMEDIATE

Log every incident — including minor events like failed login alerts, brief service interruptions, or phishing simulation results. Conduct at least one tabletop exercise annually with a written agenda, attendee list, and findings summary. Retain all evidence.

TIMELINE:30 days to begin logging; 90 days to complete a tabletop exercise and document findings
07
CC9.2

Vendor and Subservice Organization Reviews Missing

WHAT IT LOOKS LIKE

The company uses critical third-party services (cloud infrastructure, SaaS tools that process customer data) but has no evidence of reviewing those vendors' security posture. No SOC2 reports from subservice organizations were obtained or reviewed during the period.

WHY AUDITORS FLAG IT

CC9.2 requires monitoring of vendors who provide services that affect your controls. For cloud-native companies, this primarily means obtaining and reviewing SOC2 or ISO 27001 reports from AWS, GCP, Azure, and critical SaaS vendors annually. If you cannot produce a dated review for each critical vendor, it is an exception.

HOW TO REMEDIATE

Build a vendor inventory with criticality ratings. For critical vendors (those with access to production data or infrastructure), obtain their current SOC2 or ISO 27001 report annually. Document your review with a dated memo noting the opinion, scope, and any exceptions in the vendor's report relevant to your environment.

TIMELINE:2–4 weeks to compile inventory; 30–60 days to obtain and document all vendor reviews
08
CC6.7

Encryption Gaps

WHAT IT LOOKS LIKE

Data at rest is not encrypted, or encryption is implemented but not documented in the system description. Alternatively, the system description claims encryption is in place but auditors cannot find configuration evidence confirming it is enabled.

WHY AUDITORS FLAG IT

CC6.7 covers protecting information from unauthorized disclosure. Auditors look for: database encryption settings (RDS encryption, disk encryption), object storage encryption (S3 bucket settings), backup encryption, and encryption key management practices. A claim in the policy without configuration evidence is an exception.

HOW TO REMEDIATE

Audit your actual encryption settings against your stated controls. Enable encryption at rest for all databases, object storage, and backups if not already done. Document the configuration with dated screenshots or infrastructure-as-code settings. For key management, document your KMS configuration.

TIMELINE:1–2 weeks to audit and enable encryption; documentation is immediate
09
A1.2

Backup Testing Not Performed

WHAT IT LOOKS LIKE

Backups exist and run automatically, but there is no evidence that a restore was ever tested. The auditor asks for proof of backup restoration testing and finds none — or finds a test performed three years ago.

WHY AUDITORS FLAG IT

A1.2 (Availability TSC) requires that recovery procedures are tested. An untested backup is not a control — it is an assumption. Auditors require at minimum one documented restore test per year, including the date, what was restored, and confirmation that the restoration was successful.

HOW TO REMEDIATE

Perform a documented restore test. Use a non-production environment. Test full database restore from backup. Record: backup date used, restore start time, restore completion time, confirmation that data integrity was verified, and name of person who performed the test. Repeat annually.

TIMELINE:1–2 days to perform and document the first test; annual calendar event thereafter
10
CC7.2

Monitoring and Alerting Gaps

WHAT IT LOOKS LIKE

Security monitoring tools are running, but there is no evidence that alerts were reviewed and acted upon. Log aggregation exists but nobody reviewed the logs. Intrusion detection is configured but there is no documented process for who reviews alerts and what constitutes an escalation.

WHY AUDITORS FLAG IT

CC7.2 requires ongoing monitoring of the environment to detect potential security events. Auditors test this by asking for evidence of alert review — typically ticket records for alerts, a monitoring dashboard with usage history, or documented on-call procedures. Tool existence without evidence of review is an exception.

HOW TO REMEDIATE

Establish a documented alert review process: who reviews alerts, at what frequency, and what the escalation criteria are. Retain records of alert review activities — even a weekly log of 'reviewed, no action required' creates a defensible evidence trail. Route critical alerts to a ticketing system where response is automatically logged.

TIMELINE:2–4 weeks to implement documented review process; ongoing evidence collection

Frequently Asked Questions

How many control exceptions typically result in a qualified opinion?

There is no hard number — auditors exercise judgment about the materiality and pervasiveness of exceptions. A single severe exception (encryption not implemented across all production databases) can result in a qualified opinion. Multiple minor exceptions (access reviews performed 3 out of 4 quarters) may or may not, depending on the auditor's risk assessment. Generally, 3–5 exceptions in the same control domain signal a pervasive control weakness and increase the likelihood of qualification.

Can I fix control failures during the audit to avoid an exception?

Auditors test controls as they operated during the observation period — not as they operate on the day of fieldwork. If your quarterly access review was not performed in Q2, implementing it on the day auditors ask for evidence does not retroactively eliminate the exception. The control failed to operate effectively during the period. You can note remediation in the management response section of the report, which some buyers factor positively into their evaluation.

Which of these failures is considered most severe by enterprise buyers?

In procurement reviews, encryption gaps (CC6.7) and terminated employee access (CC6.1) are treated as highest severity because they represent operational security failures, not documentation gaps. Access review failures (CC6.3) and missing backup tests (A1.2) are also taken seriously but are more commonly understood as process maturity issues. Missing training records (CC1.4) and missing risk assessment documentation (CC3.1) are typically viewed as administrative gaps.

How can I avoid these failures between audits?

The controls that fail most often are all calendar-dependent: they require someone to do something at a regular interval and retain evidence of having done it. The most reliable prevention method is a compliance calendar with explicit owners and evidence artifacts defined for each task. Many companies use compliance automation platforms (Vanta, Drata, Thoropass) that automatically collect evidence and alert when recurring tasks are overdue.

Qualified Opinion: 72-Hour GuideAutomate Controls Evidence →