When To Use Tags and Links

Mem’s core distinction is that notes function as nodes in a network, not entries in a hierarchy. That framing determines when to use bi-directional links versus tags: bi-directional links are for connecting ideas to show how one thought leads to another, while tags are for grouping nodes by category so they’re easy to retrieve later.

Bi-directional links work best when they’re embedded in full sentences. The emphasis is on making the link meaningful in context, which is why note titles matter—titles should be usable inside other sentences when building new writing. In practice, bi-directional links let a person capture ideas as they occur and later trace the chain of thinking: if a note was sparked by something earlier, the linked note reveals what triggered it. The transcript illustrates this with an example note (“personal network of knowledge”) where completed sentences include links to other mems, sometimes pointing to empty placeholders and sometimes to notes with actual content. The payoff is recoverability—when revisiting an idea later, the linked trail helps reconstruct the original reasoning.

That same linking approach scales to content creation. When drafting an article, the workflow is to embed relevant mems into sentences throughout the draft, sometimes by combining entire notes into new text using Mem Spotlight. The result is an “atomic” creative process: individual notes become building blocks that can be assembled into a blog post, with many bi-directional links throughout the piece—each link placed inside the sentence where it contributes meaning.

Tags enter when the goal shifts from idea-to-idea connections to category-to-category organization. The transcript treats tags as a way to associate nodes with a finite set of contexts. A key example is email production: individual emails for a product launch get a tag (e.g., “launch sequence” / “campaign email”), and then a search for that tag returns all related launch sequences or all emails in a specific sequence. The method supports context-based retrieval: one tag can represent a group of campaign emails, while another can represent a single email item.

An explicit guiding principle comes from Thiago Forte: tag by context rather than topic, because topics expand endlessly while contexts stay bounded. The transcript also notes a practical exception for projects. Instead of relying on links to connect ideas, projects can be treated as containers that gather related resources—so linked resources under a project are acceptable when the aim is collection rather than conceptual linkage.

In short: use bi-directional links to map the relationships between ideas (and embed them in sentences for clarity), use tags to group nodes by context for fast retrieval, and treat projects as a special case where resource collections may rely more on linking than on idea-chaining.