Most Microsoft security exams test whether you can configure a feature. SC-400 tests something harder: whether you can pick the right feature when five of them look like they do the same job.
Sensitivity labels, retention labels, DLP policies, retention policies, records, eDiscovery holds — Microsoft Purview is a toolbox full of overlapping instruments, and the Information Protection Administrator exam lives in the overlaps. Candidates who study each feature in isolation walk in confident and walk out confused, because the exam rarely asks "how do you create a sensitivity label?" It asks which control stops a specific leak in a specific scenario, and three of the four answers will be technically real features that solve the wrong problem.
That is the exam. Everything below is built around it.
The exam in one paragraph
SC-400 earns the Microsoft Certified: Information Protection and Compliance Administrator Associate credential. It is 40–60 questions, you need 700 out of 1,000 to pass, the fee is $165, and there are no formal prerequisites — though Microsoft expects working familiarity with Microsoft 365 workloads and the Purview portal. The certification is valid for one year and renews free through an online assessment on Microsoft Learn, like all current Microsoft role-based certs. The exam is organized around three big capability areas: implementing information protection, implementing data loss prevention, and implementing information governance. Information protection is the heaviest of the three — sensitivity labels and encryption carry the most weight — but no single area is small enough to skip.
Information protection: labels are a decision, not a setting
The first capability area covers sensitive information types, trainable classifiers, sensitivity labels, and message encryption. This is where most candidates under-study, because labels feel simple: pick a classification, apply a label, done. The exam disagrees.
A sensitivity label is a bundle of decisions. Encryption, content marking, access controls, and retention can all ride on one label, and the exam tests whether you know what happens when those decisions collide. What takes precedence when a label's encryption conflicts with a DLP policy? What happens to a labeled document when it is shared externally, downloaded, or opened in a desktop app instead of the web? What does a label do — and not do — for data already sitting in SharePoint versus data in an email attachment?
Trainable classifiers are the other trap in this area. The exam expects you to know the difference between a built-in sensitive information type (a pattern, like a credit card number) and a trainable classifier (a machine-learning model trained on examples, like a specific contract type). Candidates who treat them as interchangeable miss questions about when each is the right detection method — and about the practical reality that a classifier needs seed content and time to train before it is reliable.
The study rule for this area: for every feature, learn what it does not do. Labels do not stop screenshots. Encryption does not stop an authorized user from forwarding content the label permits. A classifier does not work the day you create it. The exam rewards people who know the boundary of each tool.
Data loss prevention: the most practical third of the exam
DLP is where the exam gets concrete, and it is the section working admins tend to score best on — because the scenarios mirror real incidents. A user tries to email a spreadsheet of customer data to a personal address. A file containing passport numbers is shared from SharePoint to an external guest. Someone pastes source code into a Teams chat.
For each scenario, the exam wants the right DLP scope: Exchange, SharePoint, OneDrive, Teams, endpoints, or Defender for Cloud Apps. The policy that catches the email does not catch the endpoint copy. Endpoint DLP has its own prerequisites and its own blind spots. DLP for Teams chat is not the same policy surface as DLP for SharePoint.
Three distinctions decide most DLP questions:
- Policy tips vs. blocking. A policy tip warns; an override with business justification lets the user proceed anyway. The exam tests whether you know which enforcement mode actually stops the data, and what an override does to the audit trail.
- Conditions vs. exceptions. A policy that fires on everything except one group is a different design from a policy scoped to one group — and the exam will present a scenario where the wrong scoping creates either a gap or a flood of false positives.
- Where the data lives. DLP policies apply per workload. A rule that covers Exchange does nothing for a file copied to a USB drive from a managed device. That is endpoint DLP's job, and it needs onboarding and its own policy set.
If you have ever tuned a DLP policy in production and watched the alert volume, you already think the way this section wants you to think. If you have only read about DLP, build a trial tenant and break things — the exam's scenarios are written by people who have seen policies fail in production.
Information governance: the boring third that decides pass or fail
Retention policies, retention labels, records management, and eDiscovery are the least glamorous part of SC-400 and the part candidates most often cram last. That is a mistake. Governance is roughly a quarter to a third of the exam, and it is the section where rote memorization fails hardest, because the questions are about conflicts and precedence.
The core conflict to master: retention labels vs. retention policies, and what happens when both apply to the same content. A retention policy keeps everything in a location for a set period; a retention label applies to specific items and can override, extend, or trigger disposition review. When a label says delete after three years and a policy says retain for seven, which wins — and why? The exam asks exactly this kind of question, and the answer depends on understanding that labels are more specific than policies and that the longest retention generally wins when the principles collide.
Records management adds another layer: declaring a record locks it against edits and deletion, and regulatory records add proof of disposition. eDiscovery is the investigation side — Standard vs. Premium, holds vs. retention, and what you can actually search when content is encrypted or labeled.
The practical study move: build a retention scenario in a trial tenant with real dates. Create a label that retains for one year, a policy that retains for three, apply both to the same document library, and watch what the system does. One afternoon of that teaches more than a week of reading definitions.
The two habits that separate passers from repeaters
First, study in scenarios, not in features. Do not ask "what is a sensitivity label?" Ask "a finance team shares quarterly results with an external auditor; the file must be encrypted, watermarked, and expire after 30 days — what do you configure, and what breaks if you pick the wrong tool?" Every SC-400 question is a scenario wearing a feature's name. Practice translating scenarios into tool choices, and the exam starts to read like a normal workday.
Second, use the trial tenant as your lab. A Microsoft 365 E5 trial gives you the full Purview surface: sensitivity labels, DLP, retention, eDiscovery. The exam assumes you have touched these tools, not just read about them. Candidates who have actually created a label, watched it apply, tried to override a DLP block, and run an eDiscovery search recognize the exam's scenarios on sight. Candidates who only read the documentation have to reason everything out from first principles under time pressure.
A six-week plan that works
Weeks 1–2: Information protection. Build sensitivity labels with encryption and content marking in a trial tenant. Create a custom sensitive information type and a trainable classifier, and watch how long each takes to become usable. Learn the precedence rules between labels, DLP, and retention.
Weeks 3–4: DLP. Write policies for Exchange, SharePoint, Teams, and endpoints. Test each one with a real violation attempt. Learn the difference between policy tips, overrides, and blocks — and what each one writes to the audit log.
Weeks 5: Governance. Retention labels and policies, records declaration, disposition review, eDiscovery Standard vs. Premium. Practice the conflict scenarios: label vs. policy, hold vs. retention, record vs. delete.
Week 6: Timed practice. Take full practice tests under exam conditions — 40–60 questions, no notes, no pausing to look things up. The SC-400 clock is generous, but the questions are dense; the skill to practice is reading a scenario once and committing to an answer. Review every wrong answer by tracing it back to the tool boundary you misunderstood, not just the fact you missed.
Bottom line
SC-400 is not a test of whether you can click through the Purview portal. It is a test of whether you understand what each control actually guarantees — and what it does not. The candidates who pass are the ones who stopped memorizing features and started thinking in scenarios: which tool, for which data, in which location, against which threat. Study that way, lab it in a trial tenant, and the exam reads like a week at work.
If you want to check your readiness before booking, the free SC-400 practice test at Study4Pass covers the same ground — 189 practice questions with instant feedback: https://study4pass.com/study-material/microsoft/sc-400
Top comments (0)