1.3 Blueprint Domains and Weighting

1.3 Blueprint Domains and Weighting

The official content outline is the exam map. It does not expose live questions, but it defines exactly which tasks and knowledge areas Snowflake's item writers are allowed to test. Your study plan should mirror it: heavier domains get more hours and more practice questions.

The five domains and weights

These five weights sum to 100%. Note that the top three domains — Architecture, Performance, and Governance — together make up roughly 72% of the exam. If you master those three, you are close to the passing line before even touching loading and collaboration.

Why Architecture dominates

At 31%, Architecture is the single largest domain and underpins everything else. Snowflake's defining design is its separation of storage and compute across three layers: a centralized storage layer that holds data in compressed, columnar micro-partitions; a compute layer of independently scalable virtual warehouses; and a cloud services layer that coordinates metadata, security, query optimization, and transaction management. Almost every other domain assumes you understand this model — caching behavior, scaling, governance objects, and sharing all trace back to it. Invest here first.

Translating weights into hours

A simple, defensible allocation is to make your study hours proportional to the weights, then bias toward your diagnosed weak domains:

• Architecture (31%) — most hours; build the three-layer model cold.
• Performance (21%) — second; learn warehouse sizing, caching tiers, and reading the query profile.
• Governance (20%) — third; master the role hierarchy and policy objects.
• Loading (18%) — fourth; know stages, COPY options, and Snowpipe.
• Collaboration (10%) — least hours, but do not skip — easy points here can decide a borderline pass.

Do not neglect a low-weight domain just because it is small. Ten percent is ten potential questions; missing all of them can be the difference between 740 and 760.

Reading the outline like an item writer

The content outline lists not just topics but the task verbs Snowflake's writers use — "identify," "describe," "summarize," "compare," "manage." Those verbs tell you the depth expected. "Identify the layers of the Snowflake architecture" signals recognition-level recall; "manage virtual warehouses" signals you must know configuration actions, not just the definition. When you study a topic, map it to its verb and rehearse at that depth. A topic you can only define will fail a "manage" or "compare" question.

Also watch for topics that span multiple domains. Caching, for example, is an architecture concept but is tested through a performance lens; roles are governance objects but appear in loading and sharing scenarios. Cross-domain topics are high-yield because one solid concept earns points in several places, so prioritize the connective concepts — architecture layers, micro-partitions, warehouses, roles, and caching — that recur throughout the blueprint.

Blueprint tracker

Keep a one-page tracker. For each domain, mark four readiness levels: understand the concept, can apply it to a scenario, can decide under time pressure, and can explain why each distractor is wrong. You are exam-ready in a domain only when all four are checked.

Exam-ready mental model

Apply cue–authority–action–evidence–risk to each domain. The cue is the scenario (e.g., "queries are slow"); the authority is the governing Snowflake behavior (caching, clustering, warehouse size); the action is the configuration step; the evidence is the query profile or object setting; the risk is over-provisioning cost or a security misconfiguration. Tagging every practice question with its domain shows you, statistically, where your points are leaking.

Error-log rule

Tag each miss with its domain and subtopic. If your error log shows a cluster in one high-weight domain, that is your highest-leverage fix. "I missed this because" plus "Next time I will look for" turns each miss into a targeted remediation item rather than a vague feeling of weakness. Revisit the weighting whenever you review the log: a cluster of misses in the 31% architecture domain is a far more urgent fix than the same number of misses in the 10% collaboration domain, because the high-weight domain simply puts more points at stake on test day. Let the blueprint, not your comfort level, decide where the next study block goes.