PKM #2

Categorization systems and things

Mark Bao
The Personal Knowledge Management Saga
7 min readMar 7, 2015


There are a few ways to approach this project. After talking with Q today, he brought up a great point: one idea is to look at those companies who are facing knowledge management issues, specifically CEOs and that sort of group, and see what problems they have and what pain points you could solve. Essentially gathering issues and thinking about solutions.

Another thing that came up was that this is not too useful for 99% of people. Most people, it seems, are happy with Evernote (probably). A vocal minority seems to want something more powerful. But I still think there is such a desire for a tool, especially among knowledge workers, like academics, journalists, writers, managers — and maybe there’s a pain point that Evernote is not solving that such a product could eventually address that could expand the addressable userbase.

But I’m not building this hoping it becomes a product and a company. It’s a tool for me and whoever finds it useful. One that could be useful for others, but priority number one is whether it works for me.

There’s the worry that I’m too restless with these kinds of products. That because other products aren’t perfect, I don’t even want to use them. Maybe that’s an artifact of being someone who can build something. People who can’t build will just put up with the crappy tools out there and use them. I feel like something better can be made — so I don’t use what we have now, and instead think about what an ideal system might look like.

So back to the structure of this exploration. Maybe it’s a good idea to explore all the problems with current products. An ideal PKM might not or maybe even should not solve all of these problems. So the set of problems with current products is not the only thing to think about when it comes to designing an ideal system, since if we just think about correcting the errors of current systems, we’re only making a less sucky product, in a reactionary way, instead of approaching it from a holistic, more proactive, more idealistic view, which would consider the context it is in and design something that fits into that context.

Yet, it’s still a useful practice, I think, to use current tools and see why they don’t work. I’m thinking I start with the most popular one, Evernote. But that’s not the only choice. To pick something to start with, I think I’ll have to consider something that is probably the most important factor of this whole system: what kind of categorization system does it use?

Categorization systems

Filesystem folder categorization

A highly structured format that people are used to, but a bit wary of. Having folders within folders may increase drilldown fatigue, and the overhead that comes from having to maintain a clean folder structure is often overwhelming. Just think about how messy your hard drive is, and how crappy it is to reorganize it. Yet, when we try to replace the filesystem with something “smarter” — say, iCloud’s app-based categorization — we also lose a lot of flexiblity. Is there a middle ground?

To test this, flat files in Dropbox is probably the way to go. Some mobile/desktop apps are built for this as well.

Notebook categorization

Evernote-style. Only 1 hierarchy that separates you from your notes: the categorization of the Notebook. Inside notebooks, you have notes, and that’s it. Unfortunately, this seems inflexible, since there’s no latitude in how much you can “cluster” similar information.

If I have a notebook for my classwork, I can’t group my work for one class together — I’d have to create a separate notebook for that. And if I have a project in that class, I can’t group information like that together at all, either. There’s very little leeway in the clustering, and as a result each notebook becomes a disjointed collection of notes. Yet, that also makes it much easier to use and dramatically reduces organizational overhead. So what’s the right balance for the power–ease of use tradeoff?

Is there an ideal level of levels? Two, perhaps?


Tag-centric systems like Simplenote and Vesper use tags to organize everything. However, this means you just have a big jumble of notes, without any hierarchy of notes, and you can’t really use the system as a “personal wiki” — more like a big collection of notes. While useful, the lack of organization seems to be where it fails.

Other systems

  • Hybrid notebook/tag system, like Evernote’s: each document can be categorized into a notebook, and also have tags. Might be useful, though two systems can either greatly increase its flexibility or make it suck at both.
  • Categorized tags. Tags can be applied to any document, but tags can exist inside a hierarchy. So I can tag something Nutrition and it could show up in the Personal > Health > Nutrition hierarchy. Other tags can be freeform; existing inside a hierarchy is not required. If a document has multiple tags, it will show up in multiple places inside the hierarchy. Issues exist with name-binding.
  • Folders as documents. Similar to Gingko’s Russian doll system, in which each note is also a folder for other notes in the hierarchy (great example here), but a bit more radical. Dissolve the different classes of items: there is only one Item primitive, and it can contain other Items. I can have an Item named Nutrition; I can write stuff in there; I can also add another Item inside that Item such that Nutrition contains a list of the healthiest vegetables and a link to a nutrition guide on While this may be cumbersome, this is the sort of post-filesystem, post-notebook rethinking and reframing that I want to pursue.

Today’s brainstorming

Most of the ideas here have been incorporated into the above, but there are a few things to expand upon here:

  • Terminology. Do we call loose notes different than wiki notes? Do we call the loose notes just “notes” and wiki notes “documents”? Can documents contain loose notes, as a sort of reference?
  • Tying behaviors to each section of the site. If we look at this UI/UX as rooted in user behavior, we can try to understand what things need to be prioritized to match user behavior. Search always being present seems important. What about needing to know where in the hierarchy someone is? That might also be pretty essential.
  • Layout of the home screen. Would it be a good idea to have loose notes on the side? Or should they be given higher-class treatment?


Right now, I’m feeling really strongly about a structured, categorized note management system. Although I was going to set up a filesystem on Dropbox to handle this, there are pretty much no good apps that integrate a Dropbox browser and a solid WYSIWYG editor. Atom and Sublime Text have good Markdown editors, but nothing WYSIWYG. I believe that Markdown is a great format for clean, structured texts, but for all practical purposes, WYSIWYG is just more functional.

I’ve settled on two apps that are dinosaurs in the knowledge management space: DevonThink and Together.


Both suck in their own respective ways, but they do get the hierarchy thing down, and that’s respectable. So, I’ll try these out and see how they do.

But my initial feeling: categorization systems are already showing its foibles. I wanted to add meeting notes for a dinner I had the other day with two people — but a categorization system would make me choose between adding that note to one person or the other. A tagging system would allow me to tag it to two people.

Is there a middle ground? Do we have single notes that can be put under a category (e.g. Meeting Notes) but then the tags inside that category are what are considered ‘folders’?

No easy answer here.



Mark Bao
The Personal Knowledge Management Saga

Applying technology, strategy, and behavioral science to benefit the greater good. I like systems, mindfulness, coding, travel, AI, and evidence. @Columbia.