My 4 types of Logseq Notes

The core finding is that note-taking works best when different kinds of notes follow different workflows—some should be captured fast and left unprocessed, while others must be organized immediately so they’re actually retrievable later. In a Logseq-based setup (and intended to carry over to a future Logseq DB migration), four note types—journal, collections, research, and reference—are treated as distinct jobs with different timing, linking habits, and expectations.

Journal notes are for “get it out of my head” writing and for time-traveling context, not for future tasks. These entries are intentionally not tied to projects or to-do lists. The point is either to record something the future self might want to reread without needing to act on it, or to preserve a snapshot of past feelings and circumstances. The workflow stays lightweight: free-flow writing in a journal page, often unlinked, so the emphasis remains on capturing mood, thoughts, and small observations (like looking back at what was happening “last year” or “two years ago”) rather than building a system.

Collections shift from raw journaling to curated subsets using hashtags. The method groups small items—ideas, quotes, and other reusable fragments—so they can be filtered quickly later. Two frequent examples are #idea for video ideas and #quote for quotes from specific people. A key refinement is treating collection items as “inspiration with structure”: for #idea, adding a task marker helps hide already-used ideas from an overview. Collections aren’t meant to be finished like tasks; they’re meant to be skimmed, filtered, and mined. Adding properties makes collections more powerful, such as storing a quote’s source or, in a desk-chair research example, attaching price and brand so the system can generate a table with images and support later decision-making.

Research notes resemble collections but attach to a topic page rather than a hashtag. They’re gathered “just in time”: links to related URLs, videos, blog posts, and supporting notes are collected with minimal processing—sometimes only a quick scan—to avoid falling behind when information arrives faster than time allows. Later, when a concrete deliverable is needed (a video, article, or work document), the system is revisited and the scattered inputs are processed into a coherent package. A processing date marker helps separate already-processed material from new incoming information.

Reference notes look similar to research at first—notes linked to a place for later use—but they’re handled differently because they’re checked repeatedly in day-to-day work. A mistake described from earlier practice: dumping reference notes into a journal and linking them to a database page led to forgetting and difficulty finding items quickly, undermining the whole purpose. The fix is simple but strict: spend extra seconds up front to route reference notes into the correct location and summarize them immediately. The distinction is timing and inevitability—reference is needed now and often, while research is needed later on a planned schedule.

Overall, the workflow philosophy is pragmatic: decide whether a note is research (process later) or reference (process immediately), then design capture and organization steps accordingly. The same four-part approach is presented as largely tool-agnostic, with the most change expected around how collections are implemented in a database-oriented future setup.