I’ve had a love-hate relationship with note-taking applications for years, at one point spending years on Evernote before abandoning it for a paper journal and the eloquently simple Apple Notes. Apple Notes is present on all my devices, it’s fast, and it’s simple. But organizing and finding notes becomes an impossible chore at some point, and notes in my paper journal go there to never been read again. I try to use notes for everything—self-therapy, articles or book summaries, insights, personal organization, research, and more—so having a good and easy system is important.

In the desire to find the “perfect” tool, I often end up using rather esoteric tools. When I do software development, I usually use Vim, a complicated but efficient text-based editor, despite my seeming inability to ever remember more than a dozen of its endless keyboard shortcuts and commands. Even though it’s not as flashy as many of the more graphical user interfaces, Vim allows me to efficiently program and edit, freeing my mind to focus on the big picture.

However, because no tool is truly perfect, it’s easy to wonder if the grass really is greener. So, it should not have been a surprise that when during the pandemic I found myself stuck at home in an NYC apartment with young kids that I would decide to reach for a new note-taking tool. I was trying to organize research notes I was taking about the pandemic and related topics and was very dissatisfied with what I was using.

I couldn’t help but continue to see tweets in my timeline about Roam Research, which describes itself as “a note-taking tool for networked thought.” At its core, this means that it is easy to link and connect not just different note pages, but paragraphs inside those note pages. Roam seemed to be much loved by its cult-like supporters on Twitter, who argue that connecting notes this way is more natural and may allow better organization than a top-down categorization system. Roam would make it easy to trace your thoughts on topics and enable serendipitous discovery of past insights and ideas. I found this idea of a relatively simple notes tool that promised to let me forget about top-down categorization and just start taking notes very appealing.

Having now used Roam for more than half a year, I can only say that I’m moderately happy with it. It’s good enough for me to use almost daily, which is certainly something, but not nearly great enough for me to fall in love with it—I’d switch to something better if it were offered to me. In a sense, Roam does remove some of the mental overhead of attempting to categorize everything, but it doesn’t eliminate it. My impression is that not having a plan or strategy for your notes on Roam will ultimately result in unfindable notes again, and I haven’t yet really experienced the serendipitous note discovery I was hoping for from the app. My experiences with the app and my nagging worry about the future direction of the application have kept me from being willing to drink the Roamcult Kool-Aid.

How I use Roam Research

Everyone seems to use Roam differently, which is both a blessing and a curse—it’s difficult to figure out what works best and to learn from others’ experiences. I’m sure that my usage of Roam is no exception, which makes it difficult to really recommend for or against Roam for a particular person.

When I first load my Roam “graph,” I am sent to the Daily Notes section, a thoughtful and somewhat unique way to prompt my note-writing and allow a natural organization to emerge. Some Roam experts recommend starting all notes through the Daily Notes section.

In the Daily Notes section, I immediately type ;; and select a Daily notes template that I have previously created in Roam. This automatically inserts several sections like “Daily Journal” and “Daily Todo List”, which I use to make sure that nothing is slipping through the cracks while avoiding any sort of strongly structured TODO system. Roam probably isn’t a great comprehensive TODO application replacement. On the other hand, I’ve never actually managed to use a TODO app or system long-term. They always become unmaintainable and taunt me with everything undone, while making me feel misleadingly productive because I’m checking off TODOs rather than focusing on the important big picture. Casually throwing TODOs into my notes and occasionally onto project pages seems to work better for me.

I also initially created a separate “Meal Planning” and “Week Planning” page, which I fill, making links to the upcoming days of the week. When I fill in details under the date for Wednesday on each of these pages, those details will show up when I get to Wednesday on the Daily Notes page as a “Linked Reference.” These references are a large part of the power of Roam. I put important events I want to make sure I don’t miss, even though they’re on my calendar, and I put any meals I am planning to make.

Most importantly, I use Roam to take notes about things I’m thinking about and notes about decisions I’m trying to make. Roam has replaced my paper journal thinking. I also use Roam for taking notes on articles, books, and research papers. Although I like taking notes in Roam, I haven’t yet come back to these notes in a way that makes me think “Aha! This is why I’m using Roam” — most notes sit there as notes do in most note apps: forgotten. I’m not sure that’s bad, part of the value of note-taking is the act of note-taking, but I haven’t yet figured out how to leverage my notes in a more useful way. I think (or imagine, anyway) that if I were doing deep research on a topic, it would be easier to make the notes useful, so this may just reflect where I am at the current moment: mostly jumping back and forth between unrelated ideas and topics.

Browsing around Twitter, YouTube, and Google, I have found that people have invented all sorts of amazing ways to use Roam, but unfortunately, many of them are complicated to implement, very rigid to use appropriately, and seem very individualized to the particular person showing off the technique. I sometimes wonder if these people actually get anything important done other than inventing the perfect productivity system.

What’s great about Roam Research

In the big picture, I find the ease and simplicity (if you want) of Roam, combined with the block/paragraph level linking and referencing, to be great. The fact that Roam also creates implicit “unlinked references” for words and phrases that match creates a lot of interesting possibilities. Roam seems great for certain types of research and thought and (maybe) good for those of us who crave organization but are bad at organizing. Roam’s approach seems to help solve some of the problems of perfectionists like myself who get stymied by figuring out categories, hierarchies, and tags. It’s clear that if you can master Roam’s linking functionality, the possibilities become great, perhaps even allowing Roam to become a mostly all in one solution to Build a Second Brain.

Roam is also extremely customizable. It’s relatively easy to add your own custom CSS and JavaScript and there are countless themes and extensions for Roam. I’ve only used these minimally; I added some custom CSS and JavaScript to strike out an entire line when a TODO is crossed out and also slightly tweak a couple of visual parts of Roam to make it more pleasant. The value so far has been minimal, but I could see it being useful to really make the app fit me in the future.

Finally, Roam has a great depth of tools that are not immediately obvious. It has Pomodoro timers, block-level version control, kanban boards, calculations, tables, word counts, diagrams, latex syntax, and many other (poorly documented and practically hidden) features. It’s easy to geek out on some of these, but I’ve found that most of them aren’t useful (or usable) enough.

What’s not so great

I started taking notes on everything that annoyed and frustrated me (using Roam, of course), and the list of things that are “not so great” could fill their own blog post. I think that the vast majority of these things are of minor consequence, but I sometimes feel that Roam is killing me with a thousand little cuts and annoyances. Everything feels unpolished and clunky and as a result, I can’t imagine asking a team of people to use it with me or using it for something like a company wiki.

Loading time

When you first load up a Roam page, the loading time can be infuriatingly long. Sometimes I click on someone else’s publicly available Roam document and I begin wondering if the page has broken as the seconds tick by with just the obnoxious Roam loading animation mocking me. Fans of Roam encourage users to “keep a tab open for Roam” because of how universally bad the loading time is, but I find it impossible to stick to this habit, let alone that leaving a tab open doesn’t help when I want to open a second Roam window or want to check something quickly on my phone.

Access on the go

Speaking of phones, there’s no phone app, just a web page. I find it to be difficult to capture anything quickly using this page and especially difficult to find something quickly. Multiple times I’ve significantly regretted putting something that I need to access on the go into Roam. Enough that if I know I might want to see something when I’m out, I avoid putting it into Roam. Because I find it annoying and frustrating to take notes on my phone, I end up just never taking notes if I don’t have my computer near me. Some people have workflows where they take notes in another app on their phone and then later transfer these into Roam, but I can never get around to that. This is a significant issue that continually makes me doubt my decision to keep using Roam.

Steep learning curve

Many features are confusing, poorly documented, and difficult to remember how to use. I’m not very enthusiastic about having to find a ten minute YouTube video to try to remind myself how to do something or to understand a new feature. Does everything have to be a ten minute YouTube video, often made by some third party?

While Roam sort of allows you to write in a “freeform” way, much of the benefit of Roam also comes about through carefully planned structures. People have made countless different systems to make Roam work for them, and it seems like those systems do work for them. But, these systems are complicated and once you commit to a data structure, it’s nearly impossible to ever change.

Data security

I don’t have high confidence yet about security or data backup, and it’s easy to accidentally mess up a page with no good way to revert to the way it was an hour ago. I try to remember to export and backup locally, but this is a bit cumbersome. Neither of these things is atypical for a startup, but part of why I rage quit Evernote years ago was that they permanently lost some of my data through a bad sync.

Is Roam Research the next big thing?

I have very mixed feelings about Roam long term. The team behind Roam is obviously smart, talented, and motivated. At the same time, I don’t have the feeling that the team behind Roam has a clearly defined product roadmap. While I don’t wish to ignorantly criticize from my armchair, I’m often surprised by some random new niche feature while there are so many things that need improvement. I’m worried that they’ve coded themselves into a tough spot based on how little fundamental improvement I’ve seen since I’ve started using Roam. I worry that the Roam team built the application so quickly that now it’s difficult to make further changes, a problem familiar to any software developer.

Roam also seems to be putting a lot of hope into third party customization to fix up Roam’s issues, with custom CSS, custom Javascript, chrome extensions, or the always coming soon API. It’s nice to have community customization of an application and this is a real strength of a web-based product compared to a desktop application. At the same time, I find that solutions that rely on heavy community customization ultimately become overwhelming and less valuable — maintainers stop maintaining a project, people’s HOWTOs and other online information begin to depend on particular customizations, and different customizations conflict or break when the software is updated. Customizations are part of why some software is incredible, but I want a solid core product too.

And I keep going back to asking, What am I really getting out of this? I do like the ease of creating new pages and the ease of linking information together, but I’m worried that the lack of categorization and hierarchy is going to haunt me as the number of my notes increase. I’m considering building my own simple hierarchy to help with this, but doing so feels daunting and non-obvious, and it wouldn’t be easy to change. This far into Roam, I haven’t really experienced the promised serendipity from “networked thought,” and I’m not confident that this will ever truly emerge naturally from Roam usage. The lack of structure and best practices may ultimately hinder Roam.

I expect I’ll keep paying for Roam and using it almost daily for now, so I clearly like the product. At the same time, I keep wondering about something better. Most people are probably better off with something different, although that doesn’t mean there’s not a place for Roam Research. There’s a temptation to imagine that the next note-taking app will revolutionize your thinking and organization, which is why so many people try to create new ones, but I always find that the promises aren’t delivered. I hope the Roam team can ultimately deliver on the promise of Roam.