The database that runs my channels: Notion as a video production tracker
The least glamorous tool in my stack is the one that makes the rest of it smarter. Every video across my channels lives as a row in a Notion database before, during, and after production, and that boring habit is where all the receipts on this site come from.
What one row holds
Each planned video gets a row with:
- Status: planned → in production → uploaded (or shelved). This drives the workflow: my production tooling picks up “planned” rows and moves them forward.
- Title pattern and hook type tags, chosen from a fixed list. These tags are what made the hook experiment earlier in this series possible: no tags, no comparison, no verdicts.
- Evidence fields: why this keyword, what trend it rides on. Future me can audit past me’s reasoning.
- Results written back after upload: views, retention, a performance note. The loop closes or the data is worthless.
The receipts
| Item | Number | Source |
|---|---|---|
| Uploads tracked in the database | 40 | Notion query |
| Rows with view counts recorded back | 13 | Notion query |
| Average views on recorded rows | 676 | Notion query |
| Best recorded video | 2,318 views | Notion query |
Notice the honest gap: 40 uploads tracked, only 13 with results written back. Writing results back is the habit that keeps slipping, because it has no deadline and no immediate reward. I’m publishing that gap on purpose; if your tracker has one, you’re normal, and the fix is making result-entry part of the upload ritual, not a someday task.
Why a database beats a folder of scripts
The difference shows up the day you ask a question. “Which title style holds retention?” was answerable for me in one query because the tags existed. Every experiment verdict that now lives in my scripting rulebooks started as a query against this database. A folder of scripts can’t answer anything; it can only be reread.
Small print: mine is Notion because the pipeline reads and writes it programmatically, but nothing here is Notion-specific. A spreadsheet with the same columns delivers the same power.
What you can use today
- Make a tracker with five columns: title, status, title-style tag, views, one-line result note. That’s the whole system.
- Backfill your last ten uploads today. Ten rows is enough to see your first pattern.
- Add result-entry to your upload checklist so it happens the same minute you publish, not “later.”
This tracker is also where this site’s own numbers will live. Whatever the blog earns or fails to earn, the receipt gets published here.
Here are the receipts.