Salesforce Certified Data Architect

What to Expect on Exam Day

The Platform Data Architect exam is 60 scored multiple-choice/multiple-select questions, with up to five additional unscored questions mixed in to gather data for future versions — you won’t know which five, so there’s no point trying to guess them. You get 105 minutes, which works out to under two minutes per question, and the pass mark is 58%. That sounds low until you sit with the questions: they’re almost all scenario-based, several paragraphs long, describing a fictional company’s org setup, data volume, and business constraint, then asking what a data architect should recommend. There’s rarely a single “correct” answer so much as a “best” one among several technically valid options — which is what actually makes this exam hard, not the reading length.

Registration runs US$400 plus applicable local tax, and a retake (if you need one) is US$200. There’s no prerequisite exam required, though Salesforce recommends candidates have roughly 2–3 years of hands-on Salesforce experience and 5+ years working with data-centric systems generally, which tracks with how the questions are written — they assume you’ve actually designed a data model under  constraints, not just read about one.

Domain Breakdown

Data Modeling and Salesforce Data Management together make up half the exam, so if your prep time is limited, that’s where it should go first. Master Data Management is only 5%, but don’t skip it entirely — a handful of MDM questions tend to be some of the most conceptually dense on the exam, because MDM isn’t a Salesforce-native skill most admins or developers pick up on the job.

The Topics Candidates Consistently Underestimate

Data skew and its three flavors. Most candidates know “data skew is bad” but can’t distinguish ownership skew, parent-child skew, and lookup skew on sight — and the exam will describe a symptom (record locking during bulk updates, sharing recalculation taking hours) and expect you to name which skew type caused it and what to do about it.

Big objects vs. custom objects, specifically. It’s not enough to know big objects exist for high-volume, append-mostly data. The exam tests the trade-offs: no standard relationship fields, no triggers in the traditional sense, async SOQL for querying at scale, and when that limitation actually matters for a given business case versus when a custom object with careful indexing would do.

Virtualized data. Salesforce Connect and similar virtualization approaches show up in Large Data Volume Considerations, and candidates who’ve only ever worked with data physically stored in Salesforce tend to blank on when virtualizing data (rather than migrating or archiving it) is the right call.

GDPR as architecture, not legal trivia. The Data Governance section isn’t asking you to recite GDPR articles — it’s asking how you’d model a system to support right-to-be-forgotten requests, data minimization, and consent tracking inside an existing Salesforce org without redesigning everything.

How to Use Practice Tests Effectively for Salesforce  Exam

Because nearly every  question is scenario-based, drilling flashcard-style recall (“What does MDM stand for?”) won’t move your score much. What helps more is practicing the habit of eliminating answers — most scenario questions give you four options where two are obviously wrong, one is a reasonable answer that violates a constraint stated earlier in the scenario, and one is correct. Training yourself to re-read the scenario for the constraint you missed, rather than just picking the answer that “sounds like a best practice,” is the actual skill being tested.

Work through practice sets by domain rather than randomly shuffled, at least for your first couple of passes — it’s easier to notice which domain is actually weak when you’re not alternating between Data Migration and Master Data Management every other question. Once you’re consistently clearing 70%+ within a domain, switch to mixed sets that mimic the pacing, so you’re used to context-switching between domains under the clock.

Common Mistakes That Cause Retakes

• Treating this as a developer exam. It isn’t. Candidates with strong Apex/declarative skills but no architecture-level experience often fail because the exam asks “which approach is best for this business constraint,” not “which feature exists.”

• Skipping Large Data Volume prep because it feels theoretical. At 20% of the exam, LDV is the second-heaviest domain after Data Modeling and Salesforce Data Management — ignoring it because you’ve never personally managed a multi-million-record org is a common and costly mistake.

• Running out of time on long scenarios. With under two minutes per question and some scenarios running several paragraphs, candidates who read too slowly on the first 20 questions often rush — and make avoidable errors on — the last 10.

• Not accounting for unscored questions psychologically. A question that feels unusually vague or oddly worded might be one of the five unscored ones. Candidates who get rattled trying to “figure out” an oddly phrased question sometimes lose composure for the rest of the section.

Salesforce  Study Timeline by Starting Experience Level

New to Salesforce architecture roles (strong admin/dev background, no formal architecture experience): Budget 8–10 weeks. Spend the first few weeks building genuine hands-on familiarity with big objects, Salesforce Connect, and bulk data tools in a developer org — reading about these isn’t a substitute for configuring them. Save the last two weeks for scenario-based practice.

Has related data architecture experience (just not on Salesforce, or has worked adjacent to a Salesforce data architect): 4–6 weeks is typically enough. Your gap is usually Salesforce-specific mechanics (license-type-to-object mapping, big object query limitations) rather than architectural thinking itself, so prep can be more targeted.

Retaking the exam: Before anything else, figure out which domain(s) pulled your score down — Salesforce’s score report breaks results down by section. A retake candidate who scored low specifically in Large Data Volume Considerations or Master Data Management doesn’t need to re-study Data Modeling from scratch; narrow, targeted review of the weak domain plus a couple of full-length timed practice runs is usually enough within 2–3 weeks.

What Changes Most Often on Salesforce  Exam

This credential has gone through a naming change worth knowing about: it was previously administered as “Salesforce Certified Data Architecture and Management Designer,” which Salesforce has since retired in favor of the current “Salesforce Certified Platform Data Architect” exam. If you see older prep content referencing the “Designer” title, treat it as historical — the current exam guide and the retired one are not identical, even though the core domains overlap heavily.

Salesforce periodically  exam questions to a specific platform release (the current published guide aligns questions to the Spring ‘23 release cycle), which mainly affects edge-case questions tied to newer platform features rather than the core domain weightings, which have been stable. Because Salesforce doesn’t publish a public changelog for exam content the way some vendors do, the most reliable way to catch a shift is to recheck the official exam guide shortly before you book, rather than relying on prep content that could be a release or two behind.