Every productivity tool has workarounds. You learn them in the first week, refine them over months, and eventually forget they are workarounds at all. They become "just how I do things."

I am a certified Notion consultant. I spend my days helping companies in Japan set up their workspaces. About six months into an implementation, clients always say the same thing: "I've gotten used to it." The database design might be a little off, the property order might slow them down, but they say they have gotten used to it. I used to hear that and feel relieved. If they have gotten used to it, things must be fine.

In March 2026, Notion shipped a major update with AI integration and dashboard customization. But on Reddit, the feature that got the loudest reaction was none of those. It was Header 4. One extra level of document hierarchy. That was it. And that small addition made me question every workaround I had ever accepted.

On March 26, 2026, a post appeared on Reddit's r/Notion.

"Notion Has Header 4 Now!!!"

Three exclamation marks. Forty-six upvotes. The comments were filled with celebration. One person wrote: "tiny feature, but this matters more than any AI feature they've added."

Header 4. You could nest your document structure one level deeper. That was the whole thing.

That same day, Notion released Version 3.4. AI that understands data within your pages and makes suggestions. Dashboard customization. A presentation mode that displays pages like slides. These were all major changes. You could tell at a glance how much the development team had put in.

But what Reddit buzzed about most was Header 4.

Not AI. Not a new database feature. One quiet formatting option. Why was everyone making such a fuss?

I have been using Notion professionally for over three years without Header 4.

When building documents with deep structure, I would place bold text below Header 3. Header 3, then bold. Bold below bold. Before long, even I could not tell whether something was a heading or just emphasis. Readers had it worse.

I also nested toggles -- the collapsible sections you can click open and shut -- to create extra layers of hierarchy. Toggles inside toggles, faking visual structure. Someone on Reddit commented about "overusing toggles just to fake structure." That was exactly what I had been doing. I did not even realize I was faking it. At the time, I thought of it as a clever workaround.

In Obsidian, a note-taking app, the file name functions as Header 1, so you start the body at Header 2. When writing documents with Claude Code, an AI coding tool, I would catch myself starting at Header 2 without thinking. My brain had been operating on the assumption that Header 1 simply was not used. In neither case did the absence of Header 4 bother me. To be precise, I had adjusted myself so that it would not bother me.

The day Header 4 arrived, I realized something.

I had not gotten used to it. I had been enduring it.

Getting used to inconvenience accrues interest. The more you get used to something, the weaker your ability to recognize inconvenience becomes. Something you initially thought was "a little lacking" turns into "I guess that's just how it is" after six months. After a year, you forget it was lacking at all. By the time you have forgotten, the lack has become a premise.

In software, the debt that accumulates from deferred engineering decisions is called technical debt. The absence of Header 4 was, I think, Notion's design debt. A deferred design decision. But the debt had not only been piling up on the product side.

It had been piling up on my side too. Every time I faked hierarchy with toggles, substituted bold for headings, and told myself "this is good enough," my debt grew. Users had been shouldering the product's design debt.

This quiet debt that accumulates on the user's side. I want to call it habituation debt.

Writing this, something came back to me.

When explaining things to clients, I used to say: "Header 3 is plenty." I would add: "You can use bold text and it's basically a fourth level." What was I doing?

As a certified consultant, I was helping people shoulder inconvenience. No. More precisely, I was teaching them to shoulder it as "operational best practice." How to lay carpet neatly over a hole in the floor. Not how to fix the hole, but how to forget it was there.

I can picture the faces of the people I kept telling "that's plenty." What do they feel now that Header 4 exists? Probably something like "oh, that's handy." But feeling that way is itself the interest on habituation debt. Three years of endurance, settled by a minor convenience.

I had been on the side that increased my clients' habituation debt.

The ones struggling were the clients. But they probably said "I'm struggling" only once or twice. By the third time, they stopped. I took their silence as "they've gotten used to it." Maybe they had not gotten used to it at all. Maybe they had given up. Or maybe, because I said "that's plenty," they started thinking the problem was their own dissatisfaction.

When a consultant says "that's plenty," clients begin to doubt their own frustration. This was adaptation to a constraint, not optimal design. But nobody said so. Myself included.

Here is the thing. I have a feeling this is not just about tools.

I once reordered the properties in a Notion database. All I did was move the frequently used ones to the top. My workflow sped up by three seconds.

Three seconds.

But I had been paying those three seconds every day, dozens of times, feeling nothing. At thirty times a day, that is ninety seconds. Over a month, forty-five minutes. Over a year, nine hours. The numbers add up, but the issue is not the quantity. The issue is that I never felt those three seconds as inconvenience. If I had, I would have fixed it. Because I did not feel it, I kept paying for nine hours.

Too small to notice. Because you do not notice, it accumulates. Because it accumulates, you are startled when it is finally fixed: "Wait, I was paying that much?"

In quality management, there is a framework called the Kano model. It divides product quality into "expected quality" and "attractive quality." Expected quality is what should be there as a matter of course. Its absence causes dissatisfaction, but its presence earns no gratitude.

Think of it as a building. If there is a hole in the floor, people complain. But when the hole gets fixed, nobody says thank you. Of course the floor should be intact. Living on a floor without holes is not a delight. It is a baseline assumption. On the other hand, if someone installs a beautiful ornament on the ceiling, people are delighted. They might post a photo on social media. A hole in the floor and an ornament on the ceiling get entirely different kinds of attention.

Header 4 was a hole in the floor. Not an ornament on the ceiling. AI integration, dashboards, presentation mode: those are ceiling ornaments. They are useful and they look good. But what Reddit buzzed about most was one quiet formatting option being added. I think that is why Header 4 drew the strongest reaction in Notion 3.4's major update. The people on Reddit were not celebrating. They were relieved. The hole was finally patched.

You might want to reach for the boiling frog metaphor here. But that one is actually a myth. Frogs do escape. When the water temperature rises, they jump out. Humans are far more oblivious. The reason: humans have the ability to explain to themselves, "This is within acceptable range." Frogs lack that ability. That is why they can escape.

We cannot tell the difference between "got used to" and "endured." That boundary only becomes visible the moment something is fixed. Just as I realized I had been enduring, on the day Header 4 arrived.

Why can't you see it, even when it's yours? I looked into it. There is a mechanism that keeps you from noticing, and there is a way to find it anyway.