Keeping track of your ideas- creating a blog idea pipeline through queries and processing in Roam

A practical Roam workflow turns scattered ideas into a manageable “idea pipeline” by using queries to separate unprocessed notes from items worth developing and work currently in progress. The core problem driving the system is simple: tagging ideas automatically creates huge linked-reference lists (the creator reports 80+), which makes it hard to find the next writing prompt and forces constant re-evaluation.

The solution is to treat Roam like an inbox plus a progressive processing queue. As ideas arise during other work, they get tagged (for example, with a “blog” tag). But instead of letting everything accumulate in one place, the workflow relies on query logic to show only what hasn’t been processed yet. The key mechanism is combining an OR condition for multiple equivalent page names with a NOT condition that filters out anything already marked as processed.

In the transcript, the OR operator is used so that different page-name references—such as “blog content” and “video”—are treated as the same category for query purposes. Then a NOT-wrapped OR clause excludes blocks that already contain any of the “processed” markers. The result is an “unprocessed” view that functions like a to-do inbox: it lists items that match the idea tag/category but do not yet appear in the processed states.

From there, the workflow splits into focused views. A “to develop” (or “develop”) view shows ideas judged worth expanding—often those with enough context to be actionable. For example, blog ideas related to onboarding frequently include a page name like a user goal (“features rarely make sense to new users without the context of user goals”), because those contexts recur in product work. A separate “in progress” view tracks what’s actively being worked on, so the linked references for a chosen idea become a starting point for outlining rather than a sprawling list to sift through.

The system also mirrors a personal content pipeline built earlier for stand-up comedy: joke premises move to jokes to develop, then to jokes to perform. The Roam version uses similar stages—premise, develop, published (mapped to “perform”), plus an “archive” bucket for blocks that can’t simply have the blog tag removed but are no longer relevant. This keeps the pipeline clean without losing historical material.

A major benefit is progressive processing. Instead of fully processing every idea before moving on, the unprocessed query lets the user process items in batches and return later, picking up where they left off. That matters most when ideas keep flowing during ongoing product research; the pipeline prevents a backlog of unprocessed items from swallowing writing time once actual blog work begins.

Underneath the mechanics is a workflow philosophy: processing notes forces review, which keeps ideas usable over time. By repeatedly filtering, deciding what’s good, and moving items through stages, the notes become a durable asset rather than content written once and forgotten. The transcript ends with a recommendation to adopt the same “unprocessed inbox” approach and to share how others handle note processing in Roam.