CAD Study Plan, Traps, and Practice Strategy

CAD Study Plan, Traps, and Practice Strategy

Quick Answer: Spend the most final-review time on Managing Applications (25%, the heaviest domain), then the three 20-percent domains — Security, Working with Data in ServiceNow, and Automating Applications — and treat Application User Experience (10%) and Application Design and Development Concepts (5%) as smaller but still-tested domains. The CAD exam delivers 60 questions in 90 minutes with a working pass line around 70%, so you need both blueprint coverage and fast tool-selection judgment.

CAD is not a memorization-only exam. It rewards candidates who read a scenario and choose the platform artifact that solves the requirement with the least risk. There is no formal prerequisite, but ServiceNow recommends the Certified System Administrator (CSA) foundation first, and the exam is delivered through Pearson VUE (test center or OnVUE online proctoring) with a USD $300 registration fee (plus local sales tax). Verify current fees on the ServiceNow University site before you book.

Pacing Math

Ninety minutes across 60 questions is 90 seconds per item on average. Spend 45-60 seconds on the straightforward recall items so you bank time for the multi-tool scenarios that need 2-3 minutes. If you flag and skip 6 hard items on the first pass, you still have time for a second pass. To pass at 70% you may miss roughly 18 questions, so do not let two slow scenarios cost you five easy points elsewhere.

Domain-Weighted Final Review

For a 12-hour final-review block, start from the official weights, then shift time toward weak areas after a practice test.

If you already passed CSA, do not over-spend on navigation basics. Pour the extra time into the application lifecycle (the 25-percent Managing Applications domain), ACL edge cases, and Business Rule timing.

High-Yield CAD Traps

• Module role vs. ACL: a module role only hides or shows navigation; ACLs enforce table and field access. Hiding a module is not security.
• UI Policy vs. Data Policy: a UI Policy changes form behavior in the browser; a Data Policy can enforce required/read-only beyond the form, including imports and APIs.
• Flow vs. Business Rule: Flow Designer orchestrates readable, multi-step process; a Business Rule handles server-side record timing near the commit.
• before vs. after vs. async vs. display: before edits the record before save, after reacts to committed data, async offloads heavy work to the scheduler, display populates g_scratchpad at form load.
• GlideAjax vs. client GlideRecord: client code should call a Client Callable Script Include via GlideAjax, not query tables from the browser.
• GlideRecord vs. GlideRecordSecure: GlideRecord is the trusted server table API; GlideRecordSecure enforces ACLs for user-context access.
• Update set vs. source control vs. App Repository: update sets promote captured config, Git tracks history and collaboration, the App Repository distributes versioned scoped apps.
• ATF vs. Instance Scan: ATF proves behavior through repeatable tests; Instance Scan reports findings on quality, security, upgrade, and maintainability.

Hands-On Practice Strategy

Build one small scoped app in a free Personal Developer Instance (PDI) or a sub-production instance. Create a custom table with a reference field, a role, a module, table and field ACLs, a UI Policy, a Data Policy, a before Business Rule, a Flow Designer approval path, a reusable Script Include, and a GlideAjax client call. Add at least one ATF form test, one server-side test target, and one app or update-set scan.

Then rehearse the release path end to end. Commit app files to source control, review the diff, publish or package the scoped app for installation, and move any separate global configuration through an update set with preview and scan checks. Open System Logs, Application Logs, flow execution details, and ATF results so you know exactly where evidence appears when something fails.

For practice questions, write down why each of the three wrong choices is wrong. If the wrong reason is tool-boundary confusion, turn it into a flashcard. If the wrong reason is timing, rebuild that behavior in the instance until the trigger order is obvious. That habit turns a missed question into permanent knowledge instead of a memorized position.

A Seven-Day Final Plan

Use the last week to convert knowledge into recall speed rather than to learn new topics:

1. Day 1-2: Build the hands-on scoped app above; every artifact you touch becomes a recallable mental model.
2. Day 3: Drill Security and Working with Data — ACL evaluation order, roles versus modules, transform-map coalesce behavior.
3. Day 4: Drill Automation — Business Rule timing and order, Flow versus subflow versus action, GlideAjax wiring.
4. Day 5: Drill Managing Applications — the nine-step update-set promotion order, Git plus App Repository, ATF versus Instance Scan, dependencies.
5. Day 6: Take a full timed practice test at 90 seconds per item, then analyze every miss by trap category.
6. Day 7: Light review of flagged traps only; sleep, hydrate, and arrive early.

Day-Of Reminders

Read the full stem before the options, because CAD scenarios often hide the deciding clue ("must block the save," "must honor the user's ACLs," "global scope") in the last sentence. When two options both look correct, pick the one with the least code and the tightest scope. Trust the declarative-before-scripted instinct, watch for the tool-boundary traps catalogued above, and keep moving so pacing never costs you reachable points.