DP-900 is easiest to study when you separate “data ideas” from “Azure product names.” Start with the concepts: structured, semi-structured, and unstructured data; transactional versus analytical workloads; batch versus streaming patterns; and the difference between relational and non-relational stores. Once those ideas are clear, Azure SQL, Cosmos DB, Synapse, Data Factory, and Power BI become easier to place.
The official Microsoft outline for DP-900 is organized around core data concepts, relational data on Azure, non-relational data on Azure, and analytics workloads. Use those groups as your notebook sections. Do not make one long list of services; write down why each service exists and which problem it solves.
How to study without getting lost
Begin with the DP-900 exam objectives and mark every service you do not recognize. Then study the service by workload: operational databases, document or key-value data, data lakes, pipelines, warehousing, reporting, and visualization. That structure is much more useful than memorizing names alphabetically.
After each topic, run the DP-900 Practice Test. If your mistakes come from confusing relational and non-relational choices, return to the objective map. If your mistakes come from rushing, read the DP-900 exam format and passing score page before your next simulator session.
For review, write one short sentence for every missed question: what was the workload, which Azure service family was involved, and which clue in the wording mattered. That turns practice into a feedback loop instead of a memory game.
Keep your final notes short enough to use before a session: one page for concepts, one for relational services, one for non-relational choices, and one for analytics flow. If a note does not help you choose between two answers, rewrite it.
Verified against the official vendor reference: learn.microsoft.com.