🧠 Roam Research - Let me Introduce you to my Second Brain
Overlooked #19
Hi, it’s Alexandre from Idinvest. Overlooked is a weekly newsletter about underrated trends in the European tech industry. Today, I’m talking about my new favorite productivity tool called Roam Research.
I discovered Roam Research (a “note-taking tool for networked-thought” founded by Conor White-Sullivan) on Twitter two weeks ago and I started to use it one week ago. I immediately fell in love with the product. I'm still in the early days in my usage. I'll make an in-depth review of how I use it next month but I could not resist sharing with you why I'm so excited about this project.
I think it’s the best note taking tool ever invented and it has the potential to go way beyond this use case to compete directly with Google.
Roam wants to be the best note taking tool to mimic how our brain naturally works.
Most note taking tools like GoogleDoc, Evernote or Notion are failing to do the job properly. I'm sure you have already felt the friction of starting to write down a note on one of these tools because you were afraid to lose the note, you did not know where to put it etc. I'm sure you have also spent hours searching an idea in your notes - lost in the hierarchical structure of the tools formed by pages and folders.
When your brain thinks about a topic, it builds connections between different ideas in an anarchical process that you need to structure. Your new ideas emerge when you collude existing ideas from other people or from the past, present or future yourself.
Your brain works hierarchically but also through associations. Another way to say this is that new ideas can emerge from a top down hierarchical structure but also bottom up by mixing pre-existing ideas within your notes. Roam is able to capture both phenomena when other tools are only able to capture the hierarchal top down approach.
Roam is based on a graph database when other note taking tools are based on a relational database.
In a graph database, the relationships between records are stored at the individual record level, while in a relational database, the structure between records is defined at a higher level.
When you make a query in a graph database, it takes more time than when you make a query on a relational database because each record has to be analyzed individually and you don't have a hierarchical structure to accelerate the process.
As a result, relational database are operating more quickly but you miss the strong variation in the relationships that you have in complex structures. It’s easy to understand that having a graph database note taking tool is more powerful to keep the granularity and the infinite variations that you can have with your ideas.
Example of a Roam’s knowledge graph database: each point is a page and each line is a link between two pages
The key innovation behind Roam is the introduction of bi-directional linking thanks to its graph database design which means that when you make a link on a page A to a page B, the link also appears on page B and redirects you to page A.
Roam can be used to fulfill numerous use cases ranging from daily journaling, to meeting note taking, thesis research tool, personal CRM, habit tracker etc.
I love the horizontality of the tool because people are using it for use cases you will never have thought of and can customize the tool to their image.
Airtable has also this horizontality. Boring users like VCs use it to screen startups but more creative ones have been also using Airtable for their recruiting processes, to plan their weddings, to build websites, etc.
Roam is a "low floor, high ceiling tool".
The role model of Roam is Excel. Used by 750m individuals worldwide, Excel is really easy to use for beginners. You can easily make computations and build charts. But Excel has also a steep learning curve and has almost endless possibilities for power users. Power users can run their whole business on Excel making it a high ceiling tool.
In investment banking, I became geeky about learning all the keyboard shortcuts to format and navigate through the tool. It took me 2-3 months to master all of them but this feeling is super powerful. I'm starting to have the same behavior with Roam. I want to crack the usage of the tool and be super strong at using it.
Like in gaming, I believe that the best consumer products are the ones which have an easy entry point but that are hard to master perfectly.
Sarah Tavel (GP at Benchmark) talks about retaining users by creating “accruing benefits” and “mounting loses” within the product and Roam is the best example to illustrate this idea.
She often takes the example of Evernote to illustrate this idea. Evernote's core action is to create a note. The more notes you put in Evernote, the more value you get from the tool (accrual benefits) and the harder it is to leave the tool (mounting loses).
I believe that Roam is a more appropriate example. Evernote does not really have strong mounting loses when you leave the tool because the exportation of your notes from one note taking tool to another has become super easy and the Evernote search functionality is not sufficiently powerful by itself to retain users. With Roam, you have both accrual benefits and strong mounting losses and because of this unique graph knowledge structure.
When you are a power user, you cannot afford to lose all the relationships between your ideas that you have spent months to build. Conor talks about having compounding interests on your notes. Roam smoothes the relationships between your past, present and future ideas as well as the relationships between ideas about very different topics.
Roam is not an overnight success.
It's the result of more than one decade spent by the founder on thinking about how to build collective intelligence.
Conor founded a first company called Localocracy at the age of 19 that was already around this topic of collective intelligence. The idea was to build a tool to help citizens crowdsource local policies for their cities, vote for them and help the mayors have a more collaborative way to exercise their political power. The startup was a small success - being acquired by AOL.
Roam started by targeting high expectation customers who knew that their personal knowledge management system was completely inefficient. It started to work with researchers from the AI community who were ready to pay for the tool from day one.
These researchers knew that the tool was super early in its development but they were also aware that Roam had the potential to solve their pressing issues with their existing system. In fact, researchers are exactly the perfect population caring about structuring thoughts, gathering and connecting existing ideas, being able to generate new ones etc. Roam was able to refine the tool with these alpha users and released an open beta in Q4-19.
I love the fact that Conor is passionate about this issue of building the best personal knowledge management system, to bring it to a collective dimension and that he is ready to dedicate his whole life to this project.
The upside in Roam is almost infinite with this vision of building collective intelligence
In fact, the end vision is to go way beyond your personal knowledge management system to build the best tool to generate collective intelligence. Nicolas Wittenborn discussed about the importance of looking for optionality in VC-backed businesses. He illustrates the idea with a lovely balloon metaphor. To evaluate the potential upside of a deal, you should try to understand what is beyond the natural addressable market (the bottom of the balloon) and understand what can be the next potential opportunities the companies can go after in the medium term (the top of the balloon). In the case of Roam, the upside from the starting note taking tool market is almost infinite. Below are some ideas on the future of Roam to make you feel this upside.
Imagine a world in which you can access to the personal management knowledge system of anyone else to learn what he has been working on but also to make a digital copy of it that you will be able to compose on to push the frontiers of knowledge one step further.
Imagine a Roam based version of Wikipedia in which ideas from very different angles on a certain topic could be confronted.
Imagine a web browser powered by Roam. When you want to dig a specific topic, you will be able to access the personal knowledge management system of eminent people who spent months digging a certain topic. You will be able to access all their sources, to follow their thought processes, etc. In my opinion, Roam is not going after Notion or Evernote. Roam is going after something much bigger a.k.a. Google.
Imagine a graph knowledge sufficiently powerful to replace the education system. Conor wants a tool that you can hand to your children to learn anything from it without having to go to school. Conor wants a tool that allows anyone from anywhere to learn all the things he wants from Roam and get a Nobel prize because of his contribution to the global knowledge graph.
Roam has also built an engaged and fervent community around its product which is quietly becoming a cult. Just look at the hashtag #roamcult on Twitter to grasp what is going on.
I will give you two examples of this engagement behind the product:
People are already building products on top of Roam. For instance, Anne Le Cunff has built a Chrome Extension to copy past in a nice way content from other websites to Roam. I can’t wait the opening to third parties of the Roam’s API.
Nat Eliason has built a $99 course on Teachable on how to master Roam and has been able to earn more than $50k in 4 months. This is incredible. Knowing the steep learning curve inherent to the tool and wanting to avoid heavy customer support functions, Roam committed itself to offer $100 worth of Roam subscription for all the people that are taking the course. Of course, I subscribed to the course. I will start it in June when I'm done with my HEC master thesis and I will debrief it during my next post about Roam. (It's also a great way to try the Teachable platform)
Last week, Roam closed its doors to new users with this almost religious Tweet. Roam was struggling to keep pace with the strong user growth. 10k people had signed the weekend before the Tweet and the servers had been overwhelmed by users requests. Now, the goal is to increase the user experience and to start charging the customers on board $15 per month - a high pricing point similar to what we see in all the tools unbundling the Google / Microsoft Suite.
I’m confident in Roam’s ability to charge users at this high pricing point - at least for power users. Conor is talking about following the Tesla approach. From day one, Roam started by building a product for people who had identified pain points and wanted them solved and who were ready to pay and exorbitant amount for it. Roam did this by monetizing the beta product through enterprise deals with AI researchers.
Then, Roam opened the beta through referrals in August 2019 and opened the beta for everyone in Oct. 2019 without charging anything to use the product. Coming back to a paying product is the natural continuation in Roam’s initial go to market strategy. A mass market adoption with a lower pricing tier will come later.
Besides this event, I tried to find data points to confirm Roam’s traction which was not an easy task:
In a video dating back in December 2019, Conor was mentioning a 20% WoW growth in signups over the past 20 weeks (which is outstanding).
On Similarweb, the monthly visits on Roam increased by 110% between March and April 2020.
Similarweb Data on Roam Research
When you look at vanity metrics like the number of Twitter’s followers for Conor and Roam, the growth is also impressive.
Social Blade Graph Measuring Twitter’s Followers Growth
I hope that you now understand why I’m excited about this product. It’s the third time in the past two months that I’ve experienced the feeling of having butterflies in my stomach while meeting with entrepreneurs or digging into their products.
As an investor, it’s the feeling you have when you see the light burning strongly within a project (“the lightning strikes” Benchmark is looking for) and you know almost instinctively that you should pour money into the project to make the flywheel spinning.
You want to try Roam? Sorry, the door is closed but we will keep you updated.
Enter The Void Vector GIF By Jacqueline Jing Lin
🗞 Chasing Balloons (Nicolas Wittenborn, Aug. 2017)
🗞 Why Roam is Cool (Divinations, Apr. 2020)
🗞 Investing in Roam Research (Jeff Moris Jr., Apr. 2020)
🎧 Creating Tools For Networked Thought with Conor White-Sullivan of Roam Research (Venture Stories, Apr. 2020)
🎧 20VC - Roam Research CEO (Harry Stebbings, May. 2020)
🎧 Invest Like the Best w. Sarah Tavel (Invest Like the Best, Apr. 2020)
📽 What's So Great About Roam? (Nat Eliason, Jan. 2020)
Thanks to Julia for the feedback! 🙏
See you next week for another issue! 👋