How to organize your notes in Obsidian // The LATCH method

TL;DR

Organize notes around retrieval: decide how you want to find a note again before choosing a storage method.

Briefing Cornell Notes

Briefing

Note organization in Obsidian boils down to one practical question: how should a person reliably find a note again later? The approach starts with choosing among four core organization mechanisms—folders, links, tags, and metadata—each suited to different kinds of “finding.” Folders work best when a note truly belongs in only one place, such as separating highlights imported from other sources from the author’s own interpretations. Links, created with Obsidian’s [[note name]] syntax, are better for semantic connections that don’t fit neatly into categories; a note about “Application Performance” can be linked to “Productivity” even if they don’t share an obvious folder. Tags, added on the fly with #, handle system-level processing needs when notes don’t share meaning but still require the same workflow treatment. Metadata—specifically Dataview parameters in YAML front matter or inline—adds structure for combinations of those methods and enables database-like views and visualizations across a vault.

From there, the organizing logic shifts from “where to store notes” to “what to record inside notes so retrieval stays fast.” The LATCH system, attributed to Richard Saul Wurman’s Information Anxiety and popularized through Obsidian plugin developer Zsolt Viczián’s work, defines five retrieval anchors: Location, Alphabet, Time, Category, and Hierarchy. Location is usually optional, but it becomes useful in role-playing games where a “place” in a fantasy world can be searched via Dataview queries and supported with both metadata and links for redundancy. Alphabet covers filename-based recall—knowing the first letters of a note—so folders and Obsidian’s Quick Switcher become key tools. Quick Switcher (a core plugin) lets users jump to notes by typing partial names quickly, using a shortcut like ⌘O.

Time adds another retrieval axis. By storing date parameters in YAML and using calendar plugins (Fantasy Calender for game sessions and the Calender community plugin for daily notes), a person can find what happened on a specific day or session number. Category combines semantic grouping and workflow tagging: manual “maps of content” (MOCs) summarize related notes for tabletop role-playing games, while tags like TVZ mark items that have been synced but not yet processed. Hierarchy addresses the “I forgot the name” problem by searching upward through parent-child relationships. With ExcaliBrain, a note can list parent concepts in front matter; even if “JMeter” is forgotten, starting from “Load Testing Tool” can reveal the child node. The practical takeaway is that perfect organization isn’t required—people can pick what works now, then rely on Quick Switcher and Obsidian Search for cleanup later when notes were tagged or linked imperfectly.

Cornell Notes

Obsidian organization works best when it’s designed around retrieval: folders, links, tags, and metadata each support different kinds of searching. The LATCH method adds a second layer by specifying what information to capture in notes so they can be found later—Location, Alphabet, Time, Category, and Hierarchy. Location and Time rely on metadata and Dataview/calendar queries (especially useful for role-playing games). Alphabet is handled through filename-based structures and Obsidian’s Quick Switcher. Category uses semantic MOCs plus workflow tags, while Hierarchy uses parent-child relationships—often via ExcaliBrain—so users can find a note even when they’ve forgotten its exact name.

Why do folders work well for some notes but not for others?

Folders are ideal when a note truly belongs in one category, because a file can’t live in two folders at once. That constraint is useful for maintaining uniqueness—for example, separating imported highlights from other people’s work (stored in a Readwise folder) from the author’s own interpretations (stored in a Book folder). If two versions of the same title exist, placing one in Readwise and one in Book allows both to coexist, while a single-folder approach would force a conflict.

How do links differ from folders when it comes to retrieval?

Links connect notes based on meaning rather than category. A folder is mainly for categorizing; links let a person create a semantic bridge even when two topics don’t naturally share a folder. For instance, a note about “Application Performance” can link to “Productivity” so that visiting one topic can trigger recall of how it applies in another context.

What role do tags and metadata play when notes need workflow handling or database-style views?

Tags handle system-level processing when notes don’t share semantic meaning but still need the same treatment. The TVZ tag marks notes that still need processing. Metadata—specifically Dataview parameters in YAML front matter or inline—supports combinations of folders/links/tags and enables database-like queries and visualizations across the vault.

How does the LATCH system help someone find a note when they don’t remember its exact name?

LATCH provides multiple retrieval paths. Alphabet uses filename-based recall (folders and Quick Switcher). Time uses date parameters and calendar views to locate notes by day or session. Category uses MOCs and tags to group related knowledge. Hierarchy uses parent-child relationships so a search can start from a broader concept (like “Load Testing Tool”) and still surface the specific note (like “JMeter”).

How does ExcaliBrain support Hierarchy-based searching?

ExcaliBrain can infer relationships from links, but it can also use explicit front matter. In the “JMeter” example, the note lists “Load Testing Tool” as a parent. When the user searches from the parent concept, ExcaliBrain can display the child node; hiding inferred nodes helps reduce clutter so the intended result appears clearly.

Why is “good enough now, fix later” a core part of the organization strategy?

Organization is treated as iterative. People may not have time to set everything up perfectly, so they choose a method that makes sense immediately and accept later upkeep. Retrieval tools like Quick Switcher and Obsidian Search can compensate when a note is missing the right metadata or links, making cleanup manageable rather than blocking.

Review Questions

Which retrieval anchor in LATCH would you use if you only remember the day a note was created, and how would you implement it in Obsidian?
Give one example of when a folder-based approach would fail and a link-based approach would succeed.
How does Hierarchy in LATCH differ from Alphabet, and what plugin feature helps make Hierarchy work in practice?

Key Points

1
Organize notes around retrieval: decide how you want to find a note again before choosing a storage method.
2
Use folders when a note belongs in exactly one place, such as separating imported highlights (Readwise) from your own writing (Book).
3
Use links for semantic connections that don’t fit cleanly into categories, so one topic can trigger recall of another.
4
Use tags for workflow/system states (like TVZ for items still needing processing) and use metadata/Dataview parameters for queryable structure.
5
Apply LATCH as a retrieval checklist—Location, Alphabet, Time, Category, Hierarchy—so notes carry multiple findable anchors.
6
For “forgot the name” situations, rely on Hierarchy with parent-child relationships (e.g., via ExcaliBrain) rather than only filename recall.
7
Treat organization as iterative: pick what works now, then use Quick Switcher or Obsidian Search to correct gaps later.

Highlights

Folders enforce a single home for each note, which is useful for maintaining uniqueness—like keeping Readwise highlights separate from Book notes.

Links are for meaning, not categories; they let unrelated-looking topics (e.g., Application Performance and Productivity) still inform each other.

Time-based retrieval becomes practical when notes store date metadata and calendar views (Fantasy Calender or Calender) surface the right entries.

Hierarchy-based search can recover forgotten note names by starting from a parent concept and surfacing the child note (JMeter from Load Testing Tool) via ExcaliBrain.

Perfect organization isn’t required; retrieval tools make later cleanup feasible when metadata or links are imperfect.

Topics

Obsidian Note Organization
LATCH Method
Dataview Metadata
Quick Switcher
ExcaliBrain Hierarchy