Research done in the wrong order produces confident wrong answers.
Frame the question. Jurisdiction, the precise issue, and the procedural context. Most research failures are framing failures.
Start with a secondary source. A treatise, a practice guide or a jurisdiction-specific manual gives the framework, the vocabulary and the leading authorities in an hour. Beginning with a database search produces a pile of cases without a structure.
Statutes and rules first, then the annotations, then the cases.
Read the leading case in full, not the headnote. Headnotes are editorial and are wrong often enough to matter.
Update. Citator every authority relied on. A decision reversed, superseded or heavily criticised is worse than useless.
Check the other jurisdiction. Where authority is thin, decisions elsewhere are persuasive, and identifying whether your jurisdiction follows the majority or minority position is part of the answer.
Confirm currency. Statutes are amended; regulations are withdrawn; guidance is rescinded. Note the date of the version relied on.
Record the research. The sources consulted, the searches run, and the conclusion with its limits. The memorandum is the evidence of the analysis when the position is questioned later.