The tool I write code with is now the tool that keeps me organized

For a long time my organization system was five systems, none of which talked to each other. Feature ideas in Apple Notes, technical decisions in code comments that only made sense to the version of me who wrote them, posting plans in my head, and whatever was actually urgent living in a chat scrollback I’d have to search to find again.

The problem was never storage. Notes apps are fine at storage. The problem is that a decision made on a Tuesday evening, with all the context loaded and every tradeoff still fresh, is basically gone by the following month, and I’d sit down and make the same call again from scratch, usually worse, because the reasoning that produced the first version never got written anywhere.

Why the previous Notion attempts died

I’d tried Notion before, more than once, and it always failed the same way. Setting up a Notion workspace properly is its own small project. Databases, relations, views, properties, deciding whether a thing is a page or a row. It’s a few hours of work before it holds anything useful, and when you’re already building apps solo alongside a full-time job, that setup afternoon gets pushed until the whole idea quietly dies.

What changed is that I stopped building the structure and started describing the problem. The structure came out of the conversation, as a side effect of explaining what I actually needed to keep track of, and by the end of it there was a workspace I hadn’t sat down to design.

The Marketing Command Center

That’s the name it ended up with, and it’s five databases doing five jobs.

Calendar holds what’s going out and when. Assets holds the generated images and video pieces so I’m not regenerating something I already made three weeks ago. Captions holds finished copy, already in my voice, ready to paste. Tracking Links holds the campaign links so I can actually tell which post moved installs instead of guessing. And Decisions holds the one thing that used to have no home at all.

Day to day the flow is unremarkable, which is the point. I open the Calendar, see what slot is next, check whether the Assets are already there, and if they are the work for that post is mostly done. When they aren’t, generating them starts from prompt patterns already saved in a Content Prompts page rather than from a blank box, because a prompt that worked once is a reusable thing and rewriting it from memory every time is how you end up with content that drifts off brand without anyone deciding it should.

Where the decisions live now

The Decisions database is the part that changed the most for me.

Here’s a real one. A Glaner video opened on a standalone text card and lost 23% of viewers at the first second. That’s a specific, expensive lesson, and the old version of me would have absorbed it, felt bad about it, and then made a video opening on a text card again in February because the number wasn’t written down anywhere. Instead it became a rule: frame 0 is the strongest visual plus the hook, never a text card on its own, payoff inside two or three seconds.

That rule now lives in two places, the Decisions database and the skill I use to research content ideas, which means it isn’t a thing I have to remember. It’s a constraint that shows up in the work automatically, and any idea that violates it gets caught before I’ve spent an evening rendering it.

The same applies to the technical side. Caption sits 480 pixels from the bottom to clear the platform safe zone. On-screen text stays between 12% and 75% of the frame height because the bottom quarter is covered by interface. Those numbers came from a session of trial and error, and the value of writing them down is that the trial and error happened exactly once.

The difference between a productivity tool and a tool that thinks with you

Notion never asked me a question. It’s an excellent filing cabinet and it will hold whatever shape I give it, but the shape is entirely my problem, and the reason my earlier attempts collapsed is that I was bad at deciding the shape while also being the person doing all the work that needed filing.

The difference now is that the organizing happens inside a conversation I was already having. I describe a mess, get asked what actually needs to be findable later, argue with the answer a bit, and what comes out of that is both the decision and the place it goes. The tool that helps me write Swift is the same tool that notices I’ve explained the same posting rule three times and suggests it should be written down somewhere permanent.

What it actually gets used for

Feature specs, mostly, before any code exists. Writing the spec in conversation surfaces the questions I’d otherwise hit two days into building, and the ones I can’t answer yet get written down as open questions instead of quietly becoming assumptions.

App Store work is the other big one. Keyword research, subtitle and description drafts, screenshot captions and their ordering, all of it produced against the actual positioning of the app rather than generic advice, and all of it landing in Notion where I can compare what a listing said before a change against what it says now. Campaign tracking closes that loop, because a listing change with no link tracking behind it is just a vibe.

And the blog, including this post. Themes get logged and planned in Notion, then drafted through a skill that holds my writing rules so I don’t have to restate them every time I sit down to write.

The loop

It’s the same four steps every time. A conversation where something gets figured out, a decision that comes out of it, a row in Notion so the decision survives the week, then execution against that row.

The part that took me longest to understand is that the loop only works if I close it. A decision made in a conversation and never written down is a decision I will make again, and I’ll make it worse the second time because the context that produced it is gone. The Notion step isn’t bureaucracy. It’s the only reason any of the rest of it compounds.