Planning by Date: 9 Years of Lessons from Marian Sorca

I didn't set out to build a productivity app.

In 2016, I had a pen-and-paper system that already worked. A calendar for future events. An agenda for my main daily tasks. A notebook for quick notes and the small to-dos I didn't want to forget. Three tools, three layers, one honest answer every morning to the question "what am I doing today?"

When I tried to move that system into digital apps, nothing fit.

I tried the popular ones. Trello-style boards. Project-first tools with long lists and column views. Calendar apps that wanted every task to become an event. Each one was impressive in its own way. None of them worked for me.

So I started building my own version. I called it Self-Manager.net. It was a side project for seven years. Two years ago I went full-time on it and rebranded to SelfManager.ai when AI finally made the last missing piece work.

Today the app has over 3,000 commits and more than 10,000 tasks I've personally run through it. Nine years of daily use. One mental model that survived every version.

This is what I learned.

The paper system that came first

Before I get into the digital part, I want to spend a moment on the paper system. Because the whole idea comes from there, and the reasons it worked still hold up.

Three tools, three layers. That's all it was.

The calendar

For events that must happen on a specific date. Meetings, travel, deadlines, commitments.

The calendar made time real. You can't scroll forever on a paper calendar. You see the month. You see the week. You see what's coming. There's no infinite backlog, no hidden tab, no filter between you and reality.

That's a feature, not a limitation.

The agenda

For the day-to-day. Not every task in the universe, just the ones that mattered today.

A short list I could glance at several times a day without scrolling. If I finished everything on the list, the day was a win. If I didn't, I carried things forward to tomorrow and the list reset.

The agenda created focus by constraint. There was only so much space on a page.

The notebook

The messy capture layer. Random to-dos. Ideas. Notes from calls. Small things that didn't deserve the main list but still mattered.

The notebook is the part most people underrate. Without a fast capture layer, your brain becomes the inbox. That drains energy all day. You end up thinking about things you should be writing down.

The paper system worked because the three layers protected each other. The calendar held the future. The agenda held today. The notebook caught everything else before it fell through the cracks.

And all three had one thing in common. Time. Every page had a date at the top.

What I noticed when I went digital

When I tried to move this system into apps around 2016, I hit the same wall everyone hits.

Most tools asked me to think in a way that didn't match how I actually thought.

They were built around projects and boards first. You create a project. You create a board. You drag tasks between columns. The board becomes your home base.

That's fine if your work naturally breaks into projects. But most of my work didn't. Some of it was client work, some was admin, some was personal, some was half-ideas that hadn't become anything yet. Forcing all of it into boards felt like paperwork.

A few specific patterns kept showing up.

Time was a filter, not a container. If I wanted to know what I did last Tuesday, I had to guess. Open a project, filter by date, check another project, repeat. My history was scattered across views and almost impossible to revisit quickly.

Planning and time tracking lived in different tools. I ended up doing the same work twice, or giving up tracking entirely. Neither felt right.

Reviews were something I had to set up manually. They weren't a natural outcome of how the system was organized. So I kept skipping them, even though I knew they mattered.

Calendar-style tools had the opposite problem. They wanted every task to become a timed event. But tasks don't fit in 30-minute blocks. Real days don't run on schedules. Things move, priorities shift, interruptions happen. Turning a task list into an event list just adds friction.

None of this makes project tools or calendar tools wrong. They're built for specific jobs. Project management software solves real problems for people running large teams with dependencies and stakeholder reporting.

I just kept realizing my job wasn't that job.

So I built my own.

From Self-Manager.net to SelfManager.ai

The first version was simple. Dates at the top, tables of tasks underneath, the same mental model I had on paper, now digital and searchable.

I didn't start with features. I started with the structure. Every task belonged to a day. Every day could hold multiple tables. Every table could hold unlimited tasks. That was it.

For seven years it stayed a side project. I used it daily, fixed what bothered me, and added features I actually needed. Not features I thought other people might want. Features that came from real friction in my own work.

Time tracking came in because I wanted to know where my hours actually went. Comments came in because some tasks needed context that didn't fit in a title. Images and files came in because real work involves screenshots, sketches, and reference material. Notes attached to real days came in because sometimes I just needed to write something down without it being a task.

Every change came from real use, not from a roadmap.

For most of those seven years, it was invisible. No marketing. No launch. No pressure. Just a tool I was building for myself that happened to be online.

Two years ago I went full-time and rebranded to SelfManager.ai, because AI finally made the missing piece work.

Weekly and monthly reviews had always been the weakest part of my system. On paper, flipping back through pages was natural. Digitally, I had the data, but looking back was still work. I would sit down to review a week and end up just staring at tables.

AI changed that. When you combine a date-first structure with modern language models, something interesting happens. You can ask "what did I work on last week?" and get a real answer. You can ask "what's been sitting too long?" and the system can actually tell you. You can generate weekly, monthly, and quarterly summaries across your own real data.

This only works because the structure underneath was already date-first. AI on top of messy data gives you generic summaries. AI on top of structured days gives you answers.

The rebrand wasn't just a new name. It was the moment the system finally did what my paper system had always done naturally, which is support real review without effort.

Nine years of daily use - what actually stuck

Some lessons only show up after years. These are the ones that stayed with me, in no particular order.

Days are the real unit of life

This is the one I come back to the most.

Meetings happen on days. Deep work happens on days. Interruptions happen on days. Progress and setbacks both happen on days. When your tool matches that, everything feels more natural. You stop translating your life into an artificial structure and just work.

It sounds obvious. It isn't. Most tools ignore it.

A permanent backlog breaks trust in the system

I've had backlogs of 200+ items in various tools over the years. Every one of those backlogs eventually became useless.

The problem isn't the backlog itself. The problem is that items sitting on a list forever become noise. You stop reading them. You stop trusting that the list reflects what actually matters. And once you stop trusting the list, you stop trusting the system.

Dates keep work honest. If a task has been sitting on Monday for three weeks, that's real information. It's telling you something about your priorities, your energy, or the task itself.

Capture has to be fast

If capture is slow, you stop capturing. Then your brain becomes the inbox again.

This is why the paper notebook worked. You open it, you write. No tags, no categories, no project selector. Just the idea, on the page, with a date at the top.

A good digital system has to protect that same speed. Any friction at the capture step kills the whole habit.

If review is not easy, it does not happen

This is probably the most underrated lesson of the nine years.

Most productivity systems fall apart at the review step. People talk a lot about capture, about planning, about prioritization. Almost no one talks about review, because it's the part that usually doesn't happen.

If your data is scattered across tools and views, looking back is work. When it's work, you skip it. When you skip it, you stop learning from your own life.

The whole point of a structured system is to make review effortless. If you can't sit down at the end of the week and see what actually happened, your system is doing only half the job.

Structure makes AI useful

This one only became clear in the last two years.

AI can summarize a week, a month, or a quarter well, but only if the underlying data has real shape. Date-first structure gives AI something to work with. It can compare weeks. It can surface patterns. It can answer follow-up questions with real context.

Without that structure, AI is just a guess generator that sounds confident. You get summaries that read well and mean nothing.

The AI features in SelfManager.ai only work because the structure underneath them has been date-first for nine years. The intelligence comes from the organization, not from the model.

Context belongs with the task

A task title alone almost never tells you the full story.

You need the note you wrote when you first captured it. You need the comment from when priorities shifted. You need the screenshot from the call. You need the link to the reference material. Real work is messy, and the tool should handle that mess.

On paper, context lived on the page. In a good digital system, it should live with the task.

Pinned work matters as much as dated work

Not everything fits neatly into a single day. Some work is ongoing. A content calendar. A long-running project. A pinned list of priorities you check every morning.

Date-first doesn't mean date-only. It means the day is the center of gravity. But you also need a way to hold persistent work without forcing it into a specific date.

This was one of the later additions to SelfManager.ai, and it changed how I used the product.

The principle that survived from paper

One idea carried through every version of the system. Paper in 2016. Self-Manager.net through the side-project years. SelfManager.ai today.

Everything belongs to a date.

Not a project. Not a board. A date.

Tasks, notes, time logs, context - they all live inside a day. Projects still exist, but they sit inside time, not above it. That one shift changes how the whole system feels.

You stop asking "what's in this project?" and start asking "what happened this week?" Those are very different questions. The second one is easier to answer, and usually more useful.

It also changes what the system remembers. A project-first tool remembers projects. A date-first tool remembers your actual life.

That's a different kind of memory, and once you have it, you don't want to go back.

Who planning by date is actually for

This mental model isn't for everyone, and that's fine.

Most apps are built for project managers. Not everyone is managing projects. Some of us just want to get through the day with a clear head and know what we did at the end of the week.

If that's you, a date-first system will feel obvious the moment you try it. You'll stop fighting the tool and start using it. The structure will match how you already think.

If you run large cross-functional projects with dependencies, stakeholder reporting, and long delivery timelines, you probably want something built for that. Different job, different tool. No shame in either direction.

Planning by date is for the people whose lives are mostly just days that need handling. Freelancers balancing multiple clients. Solo founders juggling product, sales, and content. Small teams where everyone wears many hats. Agency owners tracking work across projects without wanting a project management bureaucracy. Anyone who opens their laptop in the morning and wants one honest answer to "what am I doing today, and what did I do last week?"

If that description fits, the nine years of lessons in this article were written for you.

Takeaways

• Days are the real unit of most people's lives, not projects
• Paper worked because it respected time as a container, not a filter
• A capture layer for messy ideas is as important as the main list
• Permanent backlogs break trust in the system
• Capture has to be fast or your brain stays full
• Review only happens when structure makes it easy
• AI becomes useful only after your data has real shape
• Context belongs with the task, not in a separate tool
• Pinned work handles what doesn't fit in a single day
• Project-first tools are great for project-first jobs. Date-first tools are great for day-first lives

Closing

Nine years of planning by date taught me one simple thing. The best system is the one that matches how you actually think about time. If you think in days, your tool should reflect that.

That's the whole idea behind SelfManager.ai. It started as pen and paper in 2016, became Self-Manager.net as a side project for seven years, and is now the full-time thing I work on. The mental model never changed. Just the tools around it.

Not because project tools are wrong, but because not every workflow is a project. Some workflows are just days, one after another, getting handled. Meetings. Deep work. Interruptions. Progress. A weekly review at the end that actually tells you what happened.

If that sounds like your life, SelfManager.ai was built for you.