Record the research objective
Every log entry should connect to a question or evidence gap. Otherwise a later researcher cannot tell why the search mattered.
Record exactly where you searched
Include repository, database or collection, jurisdiction, date range, filters, names and variants, browse path where relevant, and access date.
Record the result even when it is negative
A failed search can prevent duplicated work and may become meaningful negative evidence if the expected record set is sufficiently complete.
Capture useful citations immediately
Do not wait until report writing to reconstruct where a record came from. Save enough source detail to find the item again.
Write the next action
A good log turns a result into a decision: follow a witness, search a sibling, inspect neighboring pages, change jurisdiction, test a competing identity, or stop pursuing a weak lead.