This is the transcript for my conversation with Mat McGann on RoamFM.
In this episode, we talk with Matt McGann, who is the founder of Health Horizon. tracking upcoming health technologies as they develop in real time for companies, investors, and innovators as well.
He is also the creator of Roam for Teamwork: covering multiplayer knowledge bases, where everything is connected and nothing is duplicated. How to set up Rome as a place where teamwork naturally builds a coherent repository of the organization's knowledge. So if you are interested in setting up a multi-player Roam for your team, this is a great resource for you. You can find this roamforteamwork.com.
Starting off with a Ph.D. in Theoretical Physics at the Australian National University and getting into entrepreneurship, Mat is now working on Health Horizon, managing a team of four on a shared Roam Graph. So this episode will be all about processes, methods, product management and more, all on detailed Roam-specific processes. We are going to get technical In this episode.
[00:00:19]Norman Chella: In this episode, we talk with Matt McGann, who is the founder of Health Horizon. a health startup, focusing on tracking upcoming health technologies in real time for all sorts of use cases for companies, entities, and innovators as well.
[00:00:35]He is also the creator of Roam for teamwork. An online course, covering multiplayer knowledge bases, where everything is connected and nothing is duplicated. How to set up Rome as a place where a team work naturally builds. A coherent repository of the organization's knowledge. So if you are interested in setting up a multi-player Roam for your team, this is a great resource for you. You can find this roamforteamwork.com .
[00:01:03] Starting off with a PhD in theoretical physics at the Australian National University and getting the itch for entrepreneurship. Matt is now working on health horizon. managing a team of four on a shared roam graph. So this episode will be all about processes, methods, product management, and more all on some amazing detailed processes surrounding queries. So we are going to get quite technical. In this episode,
[00:01:30]We talked about the dark times, how Mat stepped away from academia to focus on something that has much more impact what captures his attention in the world of science, focusing on popular scientific breakthroughs, as opposed to academic papers, the mission behind health horizon, and what is it trying to achieve? .And Matt's workflow both personal and shared a very detailed look at his multiplayer roam. Managing to dos, checklists, sales opportunities, CRMs pipelines, and much more in how he used Roam to replace every other app designed for each of these aspects,
[00:02:04]As well as finding out why he calls a team roam the inverse of his personal roam, which is a sentence I would love to say again, because it's so interesting. the inverse.
[00:02:14] From science communication to artificial general intelligence and optical character recognition. We are going to go through quite a number of things in this episode. So without further ado, let's dive into my chat with Matt McGann of Health Horizon, and Roam for Teamwork.
[00:02:30]Mat McGann: I ran a podcast many moons ago and I really loved it. Yeah. Lots of fun.
[00:02:37] Norman Chella: Oh, you didn't continue it.
[00:02:40] Mat McGann: No, this is actually really long ago. So this is 2009.
[00:02:43] Norman Chella: Oh, wow.
[00:02:44] Mat McGann: 11 years ago. Yeah, we did it at the university and when I was at uni and, um, it's really good. We actually start getting paid for it. The uni started, um, notice that we were making the podcast and podcasts were very new then.
[00:02:57] And, uh, they wanted to get into new areas. So they paid us for a whole bunch of episodes. And then the, the uni just didn't get it. Like the people in charge didn't understand the benefit. Um, you wouldn't believe at one thing they said to us was that, um, they didn't like podcasts because it only reached a local audience. and we said, but they like, these are more global than, than radio. Like it's not even national it's international and they just didn't, they just didn't get it. Um, maybe they're also annoyed that they spent $10,000 and got five episodes, but it was the early days. Like no one knew how much they should cost.
[00:03:40] It was all kind of wild West.
[00:03:43] Norman Chella: Wow. 10 K for five episodes. Oh,
[00:03:45] Mat McGann: Yeah. Yeah. That's what we,
[00:03:46] Norman Chella: That's amazing.
[00:03:47] Mat McGann: we, we didn't know what to charge. So we just said that and then they were like, Oh yeah. Okay, go ahead.
[00:03:54] And then they didn't buy anymore. So.
[00:03:56]Norman Chella: What was the show actually about, was it about like what you were working on or was it in general? Like, Oh, a podcast in the uni and it could be about anything and everything.
[00:04:04] Mat McGann: yeah, it was interviewing academics, but we had done for about a year, a podcast just on our own. And we were, we used it as an excuse to interview the most interesting people at the uni, because while I was at ANU, which is the best uni in Australia, and there are a lot of super researchers there, there's like dark matter Nobel prize winners and.
[00:04:26] Um, a guy doing nuclear non-proliferation, you know, let's just heaps of interesting stuff and we wanted to talk to them. So we just set up an interview. We would interview we'd interview three and then on a topic, and then combine three interviews to cover a topic. It was a lot of editing. It wasn't sort of an interview based thing.
[00:04:48] It was all edited together with voiceover. It was like a Radiolab. We were inspired by radio lab and, um, That's part of the reason it was so expensive, each episode took like 20 hours or something and like, yeah. But, uh, again, no one knew what they were doing back then. And,
[00:05:07] Norman Chella: I need to find this well, I'm really interested in, in, in this thing to this, because I, I actually thought about this exact idea and there was already feedback from a couple of people who are like, I want to listen to it right now. Uh, rather than actually it's a different angle in that rather than each episode is surrounding one topic where maybe you try to explain it or you explain their perspective.
[00:05:29] It's more like, You look at someone's, um, what's the word for it recently published work and they try to defend slash explain it on the interview. So if, if someone has done like a recent article and is, you know, gaining a bit of popularity in, in the academic world, uh, the interviewer is the person who goes on the air to ask them to explain even more after they read the article.
[00:05:55] Of course, cause I mean, how can you, how can you interview someone without knowing what their, their works. And try to make it a little bit more accessible for non-academics to understand its impact, to understand the importance of that experiment or to understand that thesis. So,
[00:06:09] Mat McGann: I liked that I could focus on a particular argument, even it has the opportunity to go quite deep into one path. You don't have to, you have so many interviews because you get one person once for one hour, you got to cover everything and you don't get deep. That's like an opportunity to force yourself into one, one small argument.
[00:06:29] Yeah. Sounds interesting.
[00:06:30] Norman Chella: Yeah. Yeah. You can like go into a huge rabbit hole just from that one thing. So yeah. Well, lots of opportunities there. Uh, of course, uh, in the medium of podcasting where you can really pretty much experiment with any idea. I actually didn't know that you had a podcast, so I'm going to find it. I will, I will try my best to find it.
[00:06:48] Mat McGann: well, uh, you know, one of the biggest annoying regrets ever is we lost half of our episodes to the mist of time. Yeah. Because we were using this online service and then we were coasting them locally, and then we let the server. Like the hosting lapse and so on, on my creative, on the website where, I mean, my, um, the guy did it with, we put all that creative stuff on it.
[00:07:12] Um, there were some episodes there. It, cause it's no longer on iTunes or anything like that. It's, it's just audio files on our website, like a historical artifact.
[00:07:23] Norman Chella: Oh, I love this. Okay. Okay. All right. I'm going to have to, I want to post that in the public graph. Of course, for those who want to. Listen to, I call it narratives because you really do have to Oh, yes. Yeah. Don't worry. I have your website open
[00:07:38] Mat McGann: Oh, right. Okay. So the, um, the one called. We're actually, I'm trying to rejig my online presence, actually give myself a bit of a brand and try to accumulate followers and starting to do that. And so this website needs a good looking at, but, um, the one called, uh, Litmus was the one that we did officially, the five episodes.
[00:08:00] That's not the podcast. Soundproof was the one we did experimentally before that with the interviews. But I think we only have like two or three episodes and the Twain was one where we. There was an opportunity to, um, do you know the website? I fucking love science. I
[00:08:17] Norman Chella: Yeah. Yeah, I know about it. Yeah.
[00:08:19] Mat McGann: Yeah. So there's another one called science alert.
[00:08:21] That guy who made science alert. I think science alight is the second most popular he's in my home city of Canberra. And he was interested in doing podcasts 10 years ago. And so we did a pitch. We pitched the Twain, which was another podcast series where, I mean, we just thought, look, we're a bit. Um, we didn't want to just do another interview sort of thing.
[00:08:41] So we thought we'll only pitch what we want to make. And we thought could we do comedic sketches based on scientific papers, which is a ridiculous idea, but we tried. And so the Twain there, if you take a look at that, that's, I don't know, three episodes or something of, um,
[00:08:58] Just trying to make comedy on, um, on, put on science. Yeah. So th there's a bit of, there's a lot of fun there and I kind missed doing the podcast stuff.
[00:09:09] Norman Chella: Of course. Well, I'll definitely check these out. Oh, it's just a quick, just a quick question. Since we're on this website anyway, I I've always wanted to ask this and I was a bit confused, but maybe you'd have a great answer for this. Why is it called unlamed?
[00:09:21]Mat McGann: We didn't think too hard about it. We just WordPress forced us to give it a blog name. And, uh, we called it unnamed cause we didn't even think of the name. And then around that time we were doing science communication kind of stuff, you know, like the Twain doing comedy, but then also those interviews were really communicating complex topics to people.
[00:09:44] So, uh, we, we thought at the time that most science communication was lame as in not cool, you know, like they would try too hard to be cool or they would, um, Just be boring or, or too normal. And so we didn't take it too seriously. So we just, our named became unlamed. We thought it was clever. Yeah.
[00:10:04] Norman Chella: Oh, I like that. Oh,
[00:10:05] Mat McGann: Our website is us just trying to be clever. So
[00:10:11] Norman Chella: No, I love it.
[00:10:12] Mat McGann: smile, laugh, actually.
[00:10:13] Norman Chella: If there's an origin story that it named. I mean, honestly, even that can be a podcast name, like if unnamed and an each episode is just talking about trying to deconstruct something really complex, uh, like, uh, a scientific topic. I mean, I would totally listen to it because it was an idea that I was thinking of as well.
[00:10:33] I mean, like before and. And it came from the curiosity of trying to really understand whatever the hell this jargon was like. It could range from dark matter, or it can range from, you know, like mechanics or of some form or some term I've read in theoretical physics or something along those lines. Uh, but it felt really inaccessible as if I will just never touch that field or that, that knowledge space at all the limitations.
[00:11:04] In trying to further myself or trying to further my own curiosity are barred because, you know, you have these like huge terms being thrown around only in this one context and it's not accessible for the rest of the world. So that was like a really, really big thing that, um, that I was really annoyed with.
[00:11:18] Uh, so it's, it's great to know that you doing it as well for years. Oh, I love that.
[00:11:23] Mat McGann: Well, it's pretty big in our city, this whole science communication thing. There's a, there's an actual university department at ANU dedicated to communicating science Yeah. So, so I felt the same way. I, I did a physics degree, so I ended up doing science because I was so interested in it. But funnily enough, I found that life, I found that the undergrad degree gave me a lot of information about lots of different types of physics.
[00:11:50] Then I did a master's and that was a bit, um, less interesting, cause it was a bit more constrained than it did a PhD. And that was even more constrained. Then I realized that. I wasn't interested in going that deep because you can go very, very deep. And I realized that most of the value I got out of the science stuff was in the stuff that you might classify as popular science.
[00:12:12] So just like reveling in. What's cool about the cosmos and, you know, trying to imagine the size of stars and you know, all that kind of stuff. That was the fun bit. the actual doing a science too deep, like academic, I realized that was too far. So yeah, I think there's a, there's a, there's a place where you could go where you can go as deep as you want.
[00:12:34] And, um, you don't miss out by not going deep enough because it goes infinitely deep and you get diminishing returns quite quickly.
[00:12:43]Norman Chella: Yeah. Okay. I see. Especially when you know, where you're most interested in, so there's no point going deeper. I feel like that's, that's pretty interesting because that, that sets like you have a very firm or concrete border in which, okay. This is too much. This is not me anymore, or I don't find myself like pursuing anything beyond at PhD level, assuming that each, you know, each thesis is on one specific micro, super specific level of information that maybe not even 1% of the world will ever ever know about.
[00:13:17] Maybe it's even less. I'm
[00:13:19] Mat McGann: Yeah, yeah, totally. Yeah. Much, much, much less. The average number of people who read a PhD thesis, I think is like one, like it seriously, no one reads them. Um, cause most of them are not that good. People are still learning
[00:13:37] Norman Chella: Do you still read them or no? Yeah.
[00:13:40] Mat McGann: Scientific papers. You mean
[00:13:41] Norman Chella: Scientific papers in general, but maybe do you just stop at scientific papers and then you just don't go beyond say some sub sub fields that are just too specific that you don't care.
[00:13:52] Mat McGann: I rarely read scientific papers. Yeah. Cause usually you can read a news article about some science breakthrough and it's useful to have done physics to sort of know what they're talking about. And. What it really allows you to do is kind of understand those things. Without having to read the papers in a way you can, you can read about the experiment and you can go, Oh, this is what they were trying to do.
[00:14:15] I can understand that. Um, so there must've found out that A or B, and then you find the fan B and you're like, okay, good. That's that's good. I'll move on. And you don't really have to go into details because papers are more papers are more about accountability. It's about.
[00:14:31] Norman Chella: Oh,
[00:14:32] Mat McGann: Putting your work forward for criticism by scientific peers.
[00:14:35] And so you don't necessarily, um, it's not about reading to be understood really, or, or about communicating scientific process that other people don't really need to be involved in.
[00:14:49]Norman Chella: Hmm. Uh, now as you put it that way. That's that sets a very interestingperspective in how a PhD or someone who actually contributes to a scientific paper or like a scientific journal would actually want to do their experiment or would actually want to create their paper for the sake of accountability or maybe some formal credentialism because it would mean that, you know, are you doing this for the pursuit of that field, or are you doing this for, to find a place for yourself in that scientific community?
[00:15:22] But that's a whole other rabbit hole to go through
[00:15:24] Mat McGann: Oh, that is a, that is a huge rabbit hole. It's like the problem of science right now. Um, that everyone knows is there. Yeah. You, you, you you've picked a hornet's nest.
[00:15:34] Norman Chella: Yeah. I'm sure that we can probably dive into that, uh, another time, because I'm sure that dilemma may be shared by other Roam users who are listening to this right now, because we're 15minutes in. I haven't even talked about Rome, which is someone's going to kill me for that. So, So since the last few years from creating a podcast surrounding like catering to interviewing academics and really deconstructing what they're working on and really pursuing your, your curiosity, the knowledge you're trying to gain around most popular sciences, which you don't really have to go so deep into. And we are still in the dark times before we even discovered the tool roam. How did you make the jump as a vigilant optimist from being in physics, to being in health? Because I was trying my best to make like a guess or a prediction in my head. Like as to how did you make a jump? I'm going to be honest.
[00:16:28] I could not come up with an answer, but, uh, what's uh, what happened?
[00:16:32]Mat McGann: Well, yeah, well, it wasn't there, there wasn't a, um, continuous line, but between the two yeah. I don't even think it's possible to imagine one, but while I was doing my PhD, I realized I was really wasn't cut out for. Academia more because of the structures that you have to abide by writing papers. It was just a bit too boring for me.
[00:16:56] I just kind of wanted to keep doing the fun stuff. Around that time the university was experimenting with. Innovation and entrepreneurship and trying to instill that in their students, because they kept training STEM kids in maths and physics and science, and then there wasn't enough jobs for them. So they're like, okay, we need to give them entrepreneurial skills so they can make their own jobs, essentially, which is a pretty good strategy.
[00:17:21] They started doing this in 2008. 2008 Oh, seven seven, 2007, my first year there. And I started getting involved in that. So I became more interested in entrepreneurship, gradually peeling away from academia. And then the, someone I met my co-founder, I met during that competition and he was in health and he said, I've got the basic idea for, for a business. Do you want to come in with me and do it? And I was like, yeah. And so that was a move out of academia, which was automatically out of physics. And then that was just a dive into entrepreneurship, which owing to the, the network. Yeah, I had at the time happened to be in health. So I didn't have the health background, but that's been quite interesting.
[00:18:04] Cause I'm, I'm looking at the problem from a data and analysis point of point of view and he brings the. Health and that can provide some good perspectives.
[00:18:17]Norman Chella: Right? Yeah. Okay. I'm bringing, bring a co-founder in from pretty much outside the box, but you've brought your principles in, you've brought your methods in new processes into health. I was going to ask like, if you had any interest in health at all, but I guess it's more like the pursuit of being in that environment or in that context where you are working on something that will make an impact.
[00:18:37] Oh, I love that. Love that so much.
[00:18:40] Mat McGann: Well, the, the, what I discovered by going into health was that it is, the most complicated thing. In the world, I would say it's the most complicated industry because you know, physics is hard cause you're doing experiments on very tiny things. And you're trying to understand what's happening.
[00:18:55] Biology and medicine is harder in many ways, trying to do experiments with the human body and do trials and all that kind of stuff is just so hard trying to get any knowledge out of that process. But then on that you've got the health industry you've got the government's involvement. You've got funding, you've got private industry, you've got all these things.
[00:19:16] And then all of a sudden, this is all being applied to humans. You've got risk, you've got insurance, all of this gets tangled up and the industry is just such a complicated mess. And that's quite interesting. I found that very challenging and made me glad I went into it.
[00:19:36] Norman Chella: yeah, fantastic. Oh, I, I, this is, this is super fascinating to me because it, the complexity of the industry can really scare off a lot of people, especially when. Not only do you have to learn about the technicalities behind it as in what, what needs to work? What needs to, uh, what can't be done or like trials, et cetera, but also the, the actors in it, the players, I call them, you know, entities from, from governments to insurance and all that.
[00:20:01]it's a very hard road to navigate and I've seen a few other startups in health tech or med tech, and they would do their best, but then they're barred by, by certain. Shall we say regulations or laws that prevent them from trying out a new experiment or actually going in a direction that could be beneficial, but they can't because maybe ethics or something, it?
[00:20:23] What's it called? Health horizon. So what is it actually about and what are you working on? Really? Because it says tracking health, innovation technologies, but
[00:20:33] Mat McGann: Yeah. Yeah. Well, we started as a consumer service. The idea was that you, we all watch the news and every now and then. Usually toward the end of the show, they would show a, some medical breakthrough. They'll say scientists at the Australian National University have discovered that this chemical has good results in Parkinson's patients.
[00:20:52] Then they would have this line, like scientists say, this could be a cure in 10 years time, and then you see later the show ends and then you just sit there going, Oh, okay. That'd be cool. But also for 10 years, they've been promising these cures, right? So I had the very simple question of why, why don't we know where all these are up to, you know, why can all these groups just make all these claims all the time?
[00:21:18] And there doesn't seem to be any accountability. So we figured that the progress of any health technology, if it's to be known at all, it's on the internet somewhere. So it should be feasible to. Pulling every claim that's made about health technology, every news feature, patent application, scientific paper, trial registry, bring it all together.
[00:21:40] And then. Join the dots to work out the historical progress of any health technology in the world, and then use that algorithm to track it into the future. So then the hope is as an individual, you see promise like that on the news, you've got a health horizon you sit to type in Parkinson's you just see every technology under development for Parkinson's disease.
[00:22:00] And then. You click follow, follow, follow as if they're social media accounts and then you're following those technologies. And when they enter trials past trials or become available, you get notified. So a new way I've thought of describing this is what if, what if health innovations, were as easy to follow as celebrities were?
[00:22:25] Norman Chella: Yeah. Yeah.
[00:22:26] Mat McGann: so that's kind of what we're
[00:22:27] Norman Chella: I was about to say like, it's like. Twitter for, for innovation or like Twitter for medical innovations or medical breakthroughs. ah okay.
[00:22:36] Mat McGann: Yeah. And then we avoid people having to report their own progress by just having a big machine that generates the updates, essentially. So, so that. We don't, we don't need the owner of that innovation to be reporting their progress because through their interaction with the internet, we kind of pick up their activities and we can form that into something that even consumers can understand.
[00:22:59] And so, as a, as a business, we provide services to, um, other businesses who need to track health technologies for many different reasons.
[00:23:10] Norman Chella: can an individual access it actually.
[00:23:13] Mat McGann: Yeah. Yeah, sure. So you can log in with a free account and follow a bunch of innovations. Yeah.
[00:23:19] Norman Chella: Wow. Okay. That's I didn't, I didn't think that's actually. Oh, okay. Interesting. Okay.
[00:23:26] Mat McGann: If I'm honest, we don't do a very good job of explaining it on the landing page. And it's something I've been thinking about very recently, trying to refresh that. Cause it's, it's a bit tricky where we're providing a service for consumers and professionals and, um, yeah, we could do better with that.
[00:23:41]Norman Chella: But is there a, I take it the ratio of people who would be interested in following medical breakthroughs would. Lean more towards B2B or like larger companies and public entities, more than individuals. Right?
[00:23:54] Mat McGann: Yeah, it's actually pretty, pretty equal based on the, the, I mean the, the kind of outreach we're doing, we don't have a. A pure sample space of who would be interested in the whole world, but we did, we tend to get an equal number roughly of innovators, investors, and the public. Um, yeah. Following these things.
[00:24:16]Norman Chella: Okay. Wow. I'm sure it going to be quite a number of, uh, Romans who are listening, especially if I'm not sure if you're already part of a Roam-medical in the room Slack, but tons of people are interested in finding ways to do multiplayer Rome and, you know, keeping track of. Maybe not specifically medical breakthroughs, but in the case of aggregating all the knowledge around the field of medicine, to make sure that it's constant, that it is accessible by all users, but if they have a way to track like incoming technology or incoming innovations that may affect their field, I think that'd be pretty, pretty fascinating to see how that would marry together.
[00:25:03] Mat McGann: I think that's where we would fit. Yeah. Cause we, we don't, we don't record medical data or patient data or, um, or monitor interventions or improve care or anything like that. It's really at that high level of tracking where technologies are going. So. Medical isn't quite the right area, but certainly health innovation.
[00:25:22] And then those people inside medicine can follow technologies that might matter to them. It'd be very good use of that. We use Rome in a multiplayer mode, uh, to, to basically run the entire company. And, uh, I've been slowly getting off all these online services, building the services inside of Rome. And, uh, it's, it's.
[00:25:47] It's been a lot of fun. I'm very proud of what we, what we made there.
[00:25:51] Norman Chella: So how did you discover the tool? Because I'm assuming that you've had all these, you said you've had all these online apps that you've been using to like hack all these things together. And, uh, when was that amazing moment where you saw the astrolabe? I would love to hear how did you discover it?
[00:26:08] Mat McGann: Uh, last Christmas I was trying to find a new note taking tool because I've been looking for note taking tools, really my whole life gradually finding better and better things as. The world went digital. I was still using a notebook up until last year, really? And now I kind of stopped using a notebook now.
[00:26:30] Uh, but it was, it was Christmas because of the Australian fires were happening and we were locked in my family's. Uh, we were there for Christmas and then we'll sort of lock down there. And so I signed up to about five new note taking apps at once, and that was actually really confusing because, uh, some guy, uh, DMed me on Twitter and said, Oh, I saw you looking for apps.
[00:26:51] Do you want to try my app? And he gave me a link and I was like, yeah, sure. And I clicked this link. I must've got distracted. And then I came back to it and I opened up my browser on my phone and it was the Roam signup page, but it wasn't anyone from Roam who sent me that message. So somehow that link got messed up.
[00:27:09] I, the link closed, or I just opened a different browser window or whatever. So I log into Rome and I'm using it. And I think the guy who messaged me owns it. So I'm using Rome and I messaged the guy and I said, Oh, this is really interesting. Yeah. The sidebars cools and, and this whole thing is built on a graph, is it?
[00:27:24] And this guy is like, no idea what I'm talking about. It's like, that doesn't sound like my app. So that was my first experience. It was a bit confused because I thought the. I thought the guy who suggested it to me
[00:27:38] Norman Chella: Hijacked white by some Roam employee.
[00:27:41] Mat McGann: yeah, but I started developing it. And, um, I was very interested about the fact it was built on a graph because we built a much of health horizon on, on a graph and the way we structure our data, I had been thinking about how to represent knowledge in a graph form for like the last six months. And so then there's this note taking app it's in a graph.
[00:28:08] I was like, Oh, okay. This is, this is exciting. Let's see what, what can happen. So that's what grabbed me about it was the, the graph nature of it. I was skeptical of the UI at first. I thought the UI was, was awful, but now I think it's actually very good because I come to get used to it. So the potential I saw was in the graph underneath.
[00:28:29]Norman Chella: And, and by its graph nature, do you mean it's its ability To connect all these nodes as if it's the graph or do you mean like the graph overview visual.
[00:28:40]Mat McGann: I mean connecting nodes as if they'll graph. The visual, um, I think could be useful. And, and, but that, the one thing I learned looking at graphs was that the visualization of graphs is not all that useful just generally. Like you can get cool visualizations out to get a feel, but you never really get data out of a visualization of a graph.
[00:29:00] But yeah, the, the idea to freely link any concept. Was extremely exciting. And the what helped me was I was always getting stuck. In overly structured table based or, um, directory based note, taking apps. Every time I'd use one, you'd make new folders. And I would just end up getting stuck and jarred and hemmed in by the structure.
[00:29:27] And I realized that yeah, being able to freely link things in a graph just totally removed that problem, replaced it with a different problem, which is structuring it after the fact. But that seemed. To be the right way to go. I think that's, that's the real innovation here that I like most about it.
[00:29:43] Norman Chella: Yeah, I'm sure after tinkering for a few days and then figuring out that Oh right. You can. Consider structure after writing everything down, which is quite innovative in its own way. So if this was all on your personal Roam, my assumption is that you have a personal Roam than you have your multiplayer Roam.
[00:30:00]How did you implement Roam into health horizon?
[00:30:03]Mat McGann: cause I made so many accounts back in the day. I've got all these kinds of accounts lying around. So one account has, has the graph in it, the sort of official health horizon account has a graph. And uh, I go in there with my private accounts and then other people go in there with theirs.
[00:30:19] What I realized is as we were iterating the process, getting the team on was quite a challenge because it's, you know, onboarding is a big challenge as everyone knows, but I really stuck with it. Cause I believe that we could make it work. They get it now, which is good. And it ended up being, which surprised me.
[00:30:40] It ended up being kind of the inverse of my personal graph. So my personal graph it's mostly used for. I may think of things and I see things on the move and I just want to throw it in there in a way that can be easily structured later. That's kind of what personal graph became. so daily notes is how you'd use that.
[00:30:57] You throw everything in daily notes and then you might review it later and give it some structure. I could get found out that the team room, if we did that, it'd be too much of a mess. I found it was much more useful if everyone, because everyone, when they're working, they kind of at the computer, they're not doing this kind of just throw ideas in randomly.
[00:31:16] So they then go to the pages that they need to go to, to add the information in a structured way. And then the daily notes then is the output. So people go to daily notes and they can see here are all the tasks happening today. This day is important because these events that happened, uh, these tests were done today.
[00:31:34] These sales calls were done today. And so the daily notes using queries and things like that, pulls the knowledge graph out and allows you to consume the knowledge graph relevant to today, rather than the notes page, being a place where you record stuff. That's the current stage.
[00:31:52] It keeps evolving, but that's kind of how it looks. Now
[00:31:55]Norman Chella: Wow. Okay. So if, if someone, if one of your members accesses the roam, do they go straight to a specific page or do they immediately go to daily notes page to know what they have to do or to have an overview of the day?
[00:32:09] Mat McGann: we have a. We have an what I call an attention page for each person. So there's a Mat attention page. There's a Coco attention page and that page has four different types of queries. There's there's a query that is, these are the two dues that have been assigned to you officially. These are your notifications, things that you should look at.
[00:32:33] These are your tests that are waiting on you to do. And, uh, I should just open it. Um, I could show it to you, right. We could screen share. Would that help? Um,
[00:32:46] not great for the podcast, but
[00:32:49] Norman Chella: Maybe later on once this a video version, or once I do something with the video, I'll put it up and see, see, uh, any, we can see it. We can follow along as you are screen screen-sharing, but yeah. Yeah. Uh, let's uh, try to audibly describe, uh, what's on the screen for now.
[00:33:06] Mat McGann: Yeah. Yeah. So on the, on the shortcuts on the left, this might be a good way to do it the way, the way it's structured. We have kind of a rough table of contents to our knowledge base. So that's all the documentation, that's all the, um, Sales we've made, all the deliveries we're doing the product development, the sprints or that kind of stuff.
[00:33:26] Then we've got some summary pages which pull things out based on tags. So priorities, milestones deliverables, active sales, active deliveries. These are all things that are tagged. On a page and they get pulled out as a summary. So you can just look at that and see, okay, what are all the current active sales, whether all the current deliveries that are happening, then we've got all our sprints for technical development.
[00:33:51] Uh, the current sprint we put in the shortcut list as well. And then we've got this attention page for every employee. So it Matt's attention, Coco'sattention. And that is broken down. Into the todos, the tests that are waiting for you and then sort of role specific stuff. But the big thing we had to solve, and it's still not really done very well is notifications because it's possible for something to go in and make a change and no one know about it.
[00:34:18] And it's. We would find ourselves sort of making a change in Roam and then going into Slack and saying, Hey, I made this change on this page. And that, that kind of defeats the purpose in, in many ways you may as well just use Slack to inform somebody because what I wanted to do with Rome was that every I wanted to just eliminate that problem whenever in a business or whatever.
[00:34:38] When you think we've done this before, what did we do? You know, that knowledge was lost. That's what I'm trying to eliminate. Can all knowledge that you collect as you run a business, be stored in Roam for later? So if you keep using Slack to tell people things that knowledge gets lost because it goes into your back feed.
[00:34:56] But if you can put the knowledge into Roam and then point the user to the information in Rome, we know it's there for good. And so we've, we've just made a page. So square back at square bracket at symbol Matthew and a little speech bubble emoji, that tag is used as a notification. So if you want anyone to read a block and whatever's underneath it, you pop that name with their name in it @Coco @Matthew speech bubble.
[00:35:27] Then on their attention page when they log in, they see all the things that other people have brought to their attention. So it sort of simulates notifications. It's still a bit manual, but it's working pretty well. Now I know that Roam. I'm pretty sure I got some support feedback that they're going to build this officially.
[00:35:48] Some notification stuff but this works well for now.
[00:35:51] Norman Chella: it sounds like they're going to build that officially. Cause I was thinking either with API or even right before it. That people want to be notified at their changes outside of that context. Uh, especially when, especially like in your case, especially when there's a shared room, the biggest one. If someone has edited a block that other people have worked on it, I mean, there has to be a way to notify other people that, Oh, there has been a change.
[00:36:16] Can you make sure that this is okay, et cetera. I liked that you do it, you did it manually. And it sounds like it works really well,
[00:36:22]Mat McGann: Yeah, it kind of does. I mean, we, we just have the common management problem of people remembering to look at it every day and all that kind of stuff. but even, even if notifications were built, I'm not really sure how it would work because you don't necessarily need to know every block that was changed.
[00:36:40] That, that seems to me to be too much. And so. There is a manual step, which I don't know how you'll get around, which is, you know, I rewrote the structure of this page because now we have better information. I want to bring your attention to it, but you don't need to know every change I've made. You just need to know that this change has occurred.
[00:36:59] So that, that there too, to summarize that for somebody else is a manual step. So you may as well do a tag like this and bring their attention to it. Um, but it'd be exciting to see where it goes.
[00:37:09] Norman Chella: Could you split the types of notifications further, because that sounds like you just need to know that this happened and you don't have to take action on it.
[00:37:17] Mat McGann: Yeah. Yeah. Yep. So we have to get into even more detail. So there are these notifications, which are they at name, speech bubble. And then what they do is they just delete the speech bubble. And that's it read. And that, that means that the query that's on that speech bubble tag, it no longer gets picked up.
[00:37:36] So when they get to the attention page, there's a query of at Matthew speech bubble. I read those, I just go in and I delete the speech bubble. They disappear from that query. So they're essentially read and that's how you kind of work through your notifications. So they're just somebody was saying, Matt, please look at this and maybe answer this question.
[00:37:59] That's what that's used for. But actual to do is if you want someone to do something for you, that is a check box, like a to-do tag, then describe it. And then their name, their full name. Then there's a query on their attention page that picks up to do, and that person's full name. And that is just the current to-do list for that person.
[00:38:23]So that's why there is a separate list of notifications and then a separate list of, to do's for that person. Yeah. Uh, we, we found that we had to split those up and often your to-do might also have a notification tag on it, which is fine. It just means you really can't miss this. You need to see it right now, but, um, I'm starting more generally.
[00:38:44] I'm starting to see Roam more like. A tool like Excel because you can, people use Excel and they, if they use as an Excel system to manage their, their, their sprints or their documentation or whatever, you can use Excel to do anything basically. Cause it's like a drawing machine or whatever you can just put in equations that can do whatever.
[00:39:04] So you can build anything in Excel. And I think you can build anything in Rome in that sense. And so that's what we're trying to do with, with queries, trying to simulate. These different systems. So this, this is kind of the communication management system, where you've got the two dues, but then we also have the sprints and the technical development with testing and all that kind of stuff.
[00:39:25] And then we have the sales, CRM part. And the real thing that benefits me is that all these three things are hooked up. when I'm tagging someone about an idea, I had, I can tag them to our documentation. I can tag them to the user support. That's relevant to that. I can tag them to the sprint that that was written in.
[00:39:41] Then I can tag them to the customers that are going to use it, or the task for it. And all of this is in the same system, which is where I'm trying to get to is, is your, your sales opportunities aren't stuck in Pipedrive your future roadmap isn't stuck in product board, your development isn't in Trello.
[00:40:00] And you're trying to sort of use Slack as this intermediary between all these. It's everything's just in Rome and nothing gets forgotten. That's kind of what I'm goingfor.
[00:40:08] Norman Chella: Basically you've canceled all of these other apps, right? Like now that you built all this
[00:40:12] Mat McGann: yeah. Yeah. We saved some money. doing that actually.
[00:40:16] Norman Chella: Sounds like a lot of money.
[00:40:18] Mat McGann: Well, Roam's kind of expensive. So maybe it's equaled out.
[00:40:23]Running Sprints in Roam
[00:40:23] Norman Chella: I'm really curious about, maybe this is kind of a repeat, because I'm really curious about how you, you plan your sprints via Rome. Normally there's a process to go from step one, to step, the last step and planning out every other aspect that will happen throughout that week or throughout the length of time that you have your sprint.
[00:40:40]if you could walk me through. The moment when you're with your team and you plan your sprint, how does that work? Is that, does it start with a review of every single page or attention page, and then review yand cut unnecessary tasks, etc.?
[00:40:51] Mat McGann: Um, well, we still have quite a small team, so this should all be taken with that caveat because, uh, like I don't, I don't have like 25 man team pulling all this stuff together. There's two developers and then two others. So our whole team is that use Roam at least are four people. So that what we've come up with is, is good for that size so far.
[00:41:18]each sprint has its own page and it contains the usual sprint information. You might want to know about the, the, the goals who's responsible. what aspects of the priorities does it focus on or that kind of stuff Then there are the tests that are getting done. And then each feature we used to use use cases, but I found that didn't work very well with, with Rome.
[00:41:39] Um, so each feature is then listed in its own Kanban board. Each feature is a block. And so that block can be linked with a block reference to the test, So you've got a test, which isn't a block. Then you've got a feature which is in a block, and then you do a block reference. To nest the feature under the test.
[00:41:59] And that now means what you get is when I do this test, I'm testing multiple features. And when I look at a feature, I can see all the tests, that test it. So you get this many to many relationship between features and tests that we found, we just couldn't get in Trello and it kept annoying the shit out of us.
[00:42:18] That have like, you do this test here and that tests this feature, but it also had implications for this feature. And you'd have to keep all that in your head, which tests touch, which feature because each was a separate card. And so this lets you assign essentially any number of features to tests, any number of tests, that, that worked for awhile.
[00:42:39] But when. Tests failed. And then we had to have a kind of fix or a change then introduced that had to become a new block. So we create a new section called fixes.
[00:42:51] Each block there is a modification or a tweak or a bug fix that was brought out by that test. And so now that is attached to the test as well.
[00:43:00] So every test has nested under it, the features that it's developed and then the fixes that got. Raised. And then each of those blocks have their own tags to say the status of them. So when I look at a test right now, I can see this tests two different features.
[00:43:17] One of those features is still building because it has the building tag on it.
[00:43:20] One of the features is built because it has the completed tag on it. The test also triggered three fixes. One of those is. Has been fixed and two of them are waiting to be fixed. So at a glance, I can see if this is ready to test again. All this is then pushed through to my attention page. When I look at that, I can see which tests have all their features done and all their fixes done.
[00:43:46] And then that gets pushed through to me to say, this is now ready to test Hope that makes sense with. Through audio.
[00:43:54]Norman Chella: If, if it's possible, I'm not sure to what extent if you have any like sensitive information, if you could do like a screenshot of like an older test, a feature, if that's allowed, if not that's okay.
[00:44:06] Mat McGann: Yeah. I could probably spend some time and put together a sort of fake sprint and I think that that would help a lot.
[00:44:13] Norman Chella: yeah, like even just a screenshot, just to give an idea, like a visual element to, understanding how. How features get nested under tests and how tests get processed. from there, how does that get pushed on to an attention page? Because you mean you can switch, the labels are on, and I feel like this can, like this system can really fit many other use cases as well.
[00:44:34]Pagination saves time
[00:44:34]Mat McGann: Yeah, yeah, yeah, totally. Uh, the other thing to mention that really saved. Me many, many hours. Was that everything you do because of the nature of Rome, if you do anything more than once it should become its own page. Right? So, and then, then you just link to that one source of truth. So tests that I would find myself writing into Trello for people to do.
[00:44:59]If I write, if I describe how test should be done, I can now make that a page. And that becomes a standard test. And now anytime I spin up a new sprint, I can just literally type like, okay, what tests do I need to do? I start writing. Because one of the functions of Health Horizon is following a technology, right?
[00:45:14] So it's got something to do with following. I can start typing follow, and then in my dropdown list, it says, okay, here are the follow features. Here are the follow tests. Here are the follow documentation and I can go, Oh, Hey, here are the follow tests. I just look at them and I go, all right, we want to do follow test one, two, three, and four.
[00:45:31] And then that's it. I don't need to write the tests again, that whoever's coming in to test just sees the link to the original description of that tests. So you're building up a portfolio of tests that you just link to when you need them. over times saving more and more time as we go on.
[00:45:48]Norman Chella: And in this case, the test could be referred to like a technical document, Like a very specific checklist on like, okay, on doing this or on assigning this test to this context do the following. And you save a lot of hours because you don't have to rewrite everything then at the copy paste it all over again.
[00:46:06] Oh, okay. Okay. Do you have to make any minor modifications or,
[00:46:12]Mat McGann: Yeah. All the time, all the time. And this is really why Roam is valuable because we could try to build this in. Is it air table that, that sort of, um, you can try and build something like this in, in airtable and in, in notion and all those kinds of tools. But what I found using those is because they get you to put the structure in first.
[00:46:35] You always, you always have to do things within that structure and creative work between people. Can't remain in a structure. And so, you know, if I link to a test and this one test just happens to need a new attribute or something, a meta comment, or I just want to link to a picture that no other test has.
[00:46:57] You can just do it in Roam right. You just add another block and you just throw it in. And for that little bit of mess, you avoid having to restructure all your other ones or build a new structure around tests. So. these exceptions are so common that you shouldn't try too hard in my experience to keep to a structure, because that tends to be what reduces motivation and introduces complexity.
[00:47:19] Just leave things a bit messy and move on and you can just push through.
[00:47:24]Norman Chella: I'm seeing that a lot with, um, other members who who've jumped ship to Rome because of the constraints. Due to having a structure first nature, whenever they try to input any note, not from the user side, but rather the demands of the app itself to force you to think, how will this be visualized or how will this be organized first before even considering what is it?
[00:47:48]the versatility of Roam in that way it makes it just so much better, like so much easier for you to test Systems like this, like, Oh, I never would have been able to do something like this, mainly because my lack of knowledge and understanding queries is still a bit limited. So I still want to try out or experiment on what is possible, uh, in terms of integrating, say tasklistsor project management, uh, into Rome.
[00:48:13] But your attention page thing is
[00:48:15] Using attributes to enforce structure and summarizeMat McGann: Um, we, we also use attributes. So all the, um, features and fixes, they all have their own attributes and that attributes a enforce some structure, but I find that the biggest use is if you ever need a summary, if you want a summary of something that's spread across, across multiple pages, that's where I use attributes.
[00:48:37] So for example, we record the customer, the details and the value of each thing we sell in Roam underneath the customer. So you've got a customer and then it's like, we send them this invoice, and this is the information in it. Those that saved the money, the dollar value, the, um, the tax that the description they're all saved as attributes.
[00:48:59] And then I have an invoice page, which has an attribute table. And so I can look at the invoice page and see every, every invoice in a table. The actual source data of those are in the customer pages. And so that's where I found attributes to be most valuable. If you want to pull, summarize together into a table standard data across multiple different pages.
[00:49:25] Another really good example of this for product development is we've got multiple different user types on health horizon. You've got free consumers, you've got professionals, you've got internal users of different tiers. And each of them have different limits. And this was a nightmare to keep track of, you know, how many things can this group follow what visibility rules applied to what people.
[00:49:47]We used to keep a table of that information, but it could never be kept up to date whenever we'd write different documents. So now what we have is a page dedicated to each user type and each user type has an attribute, which might be, this is the number of searches they can form. This is the number of innovations they can follow because they're attributes.
[00:50:05] I also have a general user types page that pulls in those attributes. And so if you look at an individual user page, user type. You can see all their limits, or if you look at the general user type page, you can see a summary table of every user type and all their limits. And that's really the power of attribute tables because now when I go in and I go to the internal user page and I changed the number of searches, that's now updated in that table too, because there's a single source of truth about what that user limit is.
[00:50:36] So this is where I've really found attributes to be useful to.
[00:50:39]Norman Chella: And, and just to clarify, maybe you've already answered this, uh, when you want to introduce someone new or you are adding a new customer to your general CRM, and you're starting to add these attributes in it's directly on their page right.
[00:50:52] So would your workflow be, let's just say a guy named Bob started, signed up for health horizon and he's a premium user, or he's a, you know, from a B2B company or something like that. Is it immediately? Oh, because your daily notes, page pages only output. So is it that, so you immediately create a page, Bob immediately put the attributes in and then just leave it be because it's already connected to everything else.
[00:51:15]Tracking in new sales opportunities through tagging
[00:51:15]Mat McGann: Yep. Yep. I would, I would write if I had a new person I was talking to, So this is changing and this is another advantage of Rome because I can just change structure whenever I want. If I meet someone new, they'll go down as a block inside of a sales opportunities page. So they're just a block to start with.
[00:51:35] Like, that's them a little gamete or whatever, zygote, like that's the starting point. They just one little block. And then if that relationship has more opportunity or becomes more serious. We want to take things up a notch, then I give them their own page. And then once they become their own page, I just mentioned, I say, um, let's say it's an individual, Joe blogs.
[00:51:57] The first line I'll just write is a bracket bracket. I forget what I use actually active sale close bracket. That's it, then I leave it. But cause I know once I've used that active sales tag, I know I'll find it later when I look at my sales list. So now I know that page is there and then any information, anything I've done, I just keep as a list.
[00:52:20] And so using Joe blogs is page. There is now a block called. I think sale or progress and indented in that block is a whole list of to-dos. And that's literally the timeline of everything I've done and everything I need to do. So yeah, email him about this tick. Then every time I do a task, I put a task into the future as well.
[00:52:43] Follow up three days later, you know, follow up three days later, if for nothing. So that whole block can be collapsed into nothing, but if you expand it, it's basically a historical timeline of your interactions with them. And then if we make the sale, I do a new block, which is delivery. And it's exactly the same.
[00:52:59] It's a big, long timeline, which is we need to deliver this to them. We need to do this here. When you do this here. And anytime I refer to a task, I can link to the documentation or I can notify other team members. When they need to come in and help me through each of these things. Uh, and so don't be afraid to let one of those to-do lists become its own monster because you can just indent.
[00:53:21] So this is task might be send them an invoice. And then I'm like, okay, what does this involve? So then in that to do, I assign a task to say, can you, uh, scope out this what's involved and then. I put a feature in there as well and everything that needs to be done in that to do. And once all those are done, it gets wrapped up and then that to-do is done.
[00:53:42] Now each to do also has a date on it. And that's why when you go to the daily notes page, it's automatically pulling in through the link, mentions every sales to do. Every day and that's really why it's super useful. And if I need to delay something, I just changed the date and then it disappears. And it'll come back on that next day. That to do.
[00:54:07] Norman Chella: Wow. This is. Wow. I feel like this is months of testing before you came to the system. This is,
[00:54:14] Mat McGann: Yeah, it was, it took, it has, it's been pretty disruptive to come up with the system, but, um, I'm certainly not going back. That's for sure.
[00:54:25]Norman Chella: When you want to formalize the relationship with the individual from block to page level, which sounds like to me, like a deepening of friendship for some reason. Uh, but, um, is there an immediate checklist that you just copy paste to get right into time to do all the sales, like time to initiate all the sales processes?
[00:54:47] Or is it more of, depending on what happened in the initial call or the initial interaction? These are the following checklist and then immediately do it manually.
[00:54:56] Mat McGann: For sales. I don't have a template. I'm not sure why. I feel like sales are just so case by case that it hasn't really been useful to have standard structure to them. But I do have templates. For sprints templates for tests, templates for features, templates for fixes. so we, we use templates extensively.
[00:55:21] Yeah. Cause you, especially when you have multiple people, it's very hard to keep track of every attribute that is needed for every type of object. Object for every type of block. And so. We used to keep that as a template, we still do. There's a sort of a template page, which just lists every template we use.
[00:55:38] But then there's a page that explains how to move all those templates into something like texts, text expander. We use one called Lintalist, which is a bit buggy, but, uh, it works. Okay. And so yeah, people are using templates all the time, all the time.
[00:55:53]Norman Chella: Well, this is probably the most integrated startup system I've ever heard being applied in Rome because previously when I would try to visualize how do people organize information or how they do their SOPs for most things, it's normally nested under a group of apps that cater to specific things like CRM or even marketing, or even, sales and all that and hearing it in Rome from how you're explaining it, it sounds so exciting.
[00:56:21]And it also sounds a little bit overwhelming at the same time because the technical knowledge needed to implement it in Roam requires a little bit more than just, Oh, now that Rome is not a note taking app anymore. It's not just a note taking app anymore It's now possible with all of these things
[00:56:38]Recreating Team Excel in Roam
[00:56:38]Mat McGann: It is. It is a bit crazy to do this because it isn't built for this, but, that's why that, metaphor with excel works. I think, cause you can make it do anything. And um, if, if Rome decided to break some of these features or they could decide to stop something because they want to focus on just being really good for note-taking and something might break here and our whole system will go down.
[00:57:06] So that's obviously not very good. Maybe it's a bit reckless on my half, but, um, yeah, things are, things are working really well and you don't actually need that many. Nothing here is too complicated. The most complicated thing here is an attribute table and, um, and templates, I would say, and couple of semi-complicated queries.
[00:57:27] No, queries
[00:57:28] Norman Chella: for a number of users are already hard to maybe not even implement, but hard to understand because sometimes it caters for. You know, w what do you put in between? What are the, uh, what's the list of logics that they can put in a query to make it work to your own system? So, Oh, this is, this is super exciting.
[00:57:45] I think, I feel like I have to like, create a system for myself after this cause, or at least edit my current one. Cause I'm not the power of queries it makes it really fascinating that Rome allows you to do all these things like these modular pages. Okay. And with that, Is there anything that you want to see come out of Roam in the future? Or is it mainly the notifications bit, otherwise you're set.
[00:58:06]Mat McGann: Notifications
[00:58:08] are a huge deal. Uh, but as I mentioned before, I'm not really sure how they will, um, how they'll help when you have lots of users making lots of small changes to lots of pages. So yeah, that, that I can see that happening very easily with a multiplayer, Roam but they've been thinking about it for longer than I have.
[00:58:27] So yeah, I, I really do want to see that
[00:58:29]Norman Chella: yeah, it's probably to do with tracking each block's version history and finding a way to. That the combination of that one and finding a way to either restore or at least view deleted blocks, because that's one of the things that a user on Rome can do. If you were on a roam graph in quite a while. And all of a sudden, you know, the five blocks that you were working on are deleted by somebody else.
[00:58:51] You want to know what happened, right? Like you want to know why and why all the connections are no lost.
[00:58:58] Mat McGann: I have another, um, yeah, this idea of put it on the Slack a couple of times, which is, there's an or component curly bracket, curly bracket or colon, and then you can put a list of things. Then they come in a drop down. That's super useful for tags, where they have sort of discreet stages, so doing to do doing done or something.
[00:59:23] When you know they're exclusive and you only want to use one at a time, it's really useful to have a dropdown because then you can just pick which one, which status you want. And you don't have to remember what all the other statuses are because they're in that list.
[00:59:36] So imagine that you have to do doing done in a dropdown, in a block. And then you've got lots of blocks with different either to do doing or done. If you go and make a query on doing currently, it matches every block because the, or, component. Literally has the text of each three, the three of those tags. So it matches every one of them. If it didn't, if it ignored, if it only used the selected one in the, or then you would have almost a programming, you'd almost be able to program things.
[01:00:12] So you'd be able to on a block, change the tag from, to do, to doing, and then all your queries will update automatically. Because the, um, you you've essentially swapped. So, I mean, it's no different to deleting the tag and typing a new one, but it means you can do that with a dropdown menu. And now you have sort of sets of tags, which describe, exclusive progressive statuses for example, and just swap between them really easily.
[01:00:40] That's another feature I have tried to ask for a few times.
[01:00:43]Norman Chella: Oh, that that'd be really useful, especially when you can edit that dropdown because some people might not even consider. Like even, just even just inclusion of the ability to add a dropdown to a block and turn that into a template because some people might have SOPs where they use the drop down for other things as well.
[01:01:00] And then if that can be searchable via query. Oof. The amount of angles that you can tackle it this Oh, okay.
[01:01:08] Mat McGann: I have thought about documenting what we've done so far, but I just wasn't really sure if anyone was interested. So if people sort of contact me on Twitter or whatever and ask for it, I'd be happy to pull this together and
[01:01:21]Norman Chella: I honestly thought that you would already have an article detailing all of this stuff. Um, I I've re I've messaged a few people who are using Roam for their own startups or for their own team. Uh, team-based use cases. So not just, not just as a, like a shared knowledge graph, yours is much more intuitive in that you have intended output.
[01:01:44] It's not just like a gathering of knowledge and you just stick it into graph, see what happens. Right. But you actually have tasks, you have to do you have people that do follow up, et cetera. You have so many different angles in terms of trying to achieve something.
[01:01:55] The resultant output or the collective output of that is a startup, you know, which is what Health Horizon is doing right now that can be applied to many other teams or a many other, you know, group cases. So, I think I've seen a few people ask questions maybe in the Slack, like, Oh, how do you do a proper CRM? Or how do you do a pipeline or how do you emulate Asana into Rome?
[01:02:19] Mat McGann: Yeah, that's kind of what we're trying to do.
[01:02:21] Norman Chella: yeah, I feel like you've already done like an all-in-one package for that. So if there's a way. For us to see like a visual element to this. There'll be fantastic. other than that, we have this show to give you a primer and, uh, if anyone has any questions, uh, I'm sure Matt will answer.
[01:02:39]Actually, no there's one question. If you're using Roam for your personal self as well. Is there anything that you want to try to create using Roam? That is like in your interest.
[01:02:50] Mat McGann: For the personal one. uh, yeah. Yeah. I still like reading physical books and I make all my notes in the margins and highlights, and then I want them to go into Rome. And so I was looking into lots of, optical character recognition, automated systems to scan a book or scan your notes. And. Get them in text and whatever. I couldn't find any that would work mostly because, um, vertical lines that are highlights down the side.
[01:03:18] It's hard to tell the computer to sort of grab the sentence contained by the line. But then also in my handwriting in the notes was just awful. So no, one's only a human can do that. Um, the other, the other thing we do in health horizon is a lot of outsourcing. So I, I did experiment with getting a transcriptionist to.
[01:03:35]Read my notes. So now I just read this book. For example, it's got notes sort of scrolled in the margins like this, and I just take it, literally take a photo of every page with a highlight in it and then put it into a Dropbox document. And I have a couple of people on Upwork who then transcribe it manually.
[01:03:59] And then there's a certain rules. They follow. They put it into Dropbox paper. I think they type it straight into Dropbox paper because I can export that as Mark down. And then I paste that into Rome. And so then in my private Rome, I can search for a topic and then I just have all my physical notes coming up as well.
[01:04:17] So it's like, you're searching this book is about creativity. If I searched for the word creativity in it, I can see every note I hand wrote in margins of all my books that had the word creativity in it in the unlink mentioned unlinked references. So I've kind of built a sort of search engine for my handwritten notes for the personal Roam.
[01:04:39] And I was thinking about how to make that even more efficient, because I think I've got it down to roughly. Two or $3 per thousand transcribed words using these outsource people. I know other people have this issue as well, so that's something I'm trying to build in the, for personal use.
[01:04:59]Norman Chella: yeah, there, there are a lot of others who, who have had years of. Physical notes and they want it to be put into Rome just to see what possibilities come on. Outfits. I've seen that actually on #roamcult called Twitter where they, they wish they had like a physical version and then just upload the page and hopefully it will, it will transcribe and pull out all the text. That's something way further in the future.
[01:05:23] Mat McGann: I thought the same thing and I'm sitting there going, Oh, this, if only we had AGIs that got automate this process, right. It was smart enough to do this. And it took you so long. And then one day I just realized, okay. There are AGIs. They're called people like they're everywhere and they would be very happy to type out your notes if you pay them.
[01:05:44] So rather than paying $0 for an AGI, can you, could you just pay $2 and give it to someone who really needs the money and then they can transcribe it all for you and you get your notes. You want, you just have to pay a little bit of money instead of zero money. I think that that was the kind of the breakthrough there.
[01:06:02] So, um, Yeah, we, we have AGI the call people and you can just ask them to do stuff and they'll do it if you pay them. It's great.
[01:06:10] Norman Chella: Fantastic. Now, if there's a way to have that implemented in, uh, into Rome directly. So, uh, Roam Research team, if you, uh, are considering, optical character recognition,
[01:06:23] Mat McGann: Oh, that, that, that makes me think of the other. Thing would be easy. It's easy to sharing individual pages publicly from a, from a shared one because often I'll, I'll write practice stuff in Rome and then have to sort of paste it somewhere else and then share that. And, but just to be able to share individual page. That'd be
[01:06:41]Norman Chella: Yeah. And I've heard there were some issues concerning security or, uh, the possibility that sharing just one page will result in the entire graph being public. Uh, And maybe it's just the nature of the graph in that once you have maybe one link that one link messes with the logic, thinking that, Oh, okay.
[01:07:03] If, if they're okay with one page being public, then they're okay with it being public with everything else. So that's, uh, that's also, if they can handle that, that'd be great as well.
[01:07:10] Cause I, I do that, um, especially through notion since notion can do like individual page share links and it, it can handle direct copy text from Roam. So that's like a, like a little hacky way to share things that I've written in Rome to people without sharing with them my graph. So, okay, cool. If we can get the OCR, I think that'd be great. Cause I also write a lot of margins in my books and I forget after a while, like I can only refer to those notes by going into that context, which means just actually picking up the book again.
[01:07:42] But. I'm already here. I want to search it. Like it would be great if I could search the
[01:07:46] Yeah. that all the notes I have written
[01:07:51] Mat McGann: Well, if you want to get a, if you want to grab one, if you want to try it out and take photos of them, um, take photos of the pages. It takes like five minutes. I could get my, um, transcriptionists to give it a go. And you can see if that, if it works for you, I'd be interested to see who else wants something like that.
[01:08:09] Norman Chella: Yeah. Yeah. I'll okay. Sure. Yeah. I'll, I'll write a few pages in, Oh, that's a to-do for me, uh, in the public Roam graph And, uh, on that note, uh, we are coming up on time, but, uh, are a few questions right at the end that I would love to hear your take on it. So, uh, Matt, the first question is how would you describe Roam to someone who hasn't it yet?
[01:08:32]Mat McGann: Oh, so I've done that many times. Um, yeah. Oh, the last time I did it, that was at the pub. And I said, um, I guess I did the kind of marketing story thing. It was more like it's, it's bit too complicated to explain how it works. So I just said to my friend, I've been, I spent 18, 20 years trying to find a note taking system that works for me, that works with my brain. And, I found it a few months ago. That's how I introduce it to my friends. Cause a lot of them have the same issue. Then that note taking stuff doesn't. No note taking system has been sufficient up to today and, um, So that's how I usually introduce it.
[01:09:22] And then I will say that it's based on a graph so that you, uh, you can just throw ideas in and structure it later, rather than having a structure and putting ideas into it. Probably don't go any deeper than that when describing it to somebody.
[01:09:38] Norman Chella: So is this on the assumption that, that they would know how, how a graph and its connections work, like the concepts behind it.
[01:09:47] Mat McGann: No.
[01:09:48]Yeah, it's more the structure aspect of it. The bottom up structure, rather than the top-down structure.
[01:09:53] Norman Chella: Okay. Yeah, because I'm having trouble there as well, because at least when I tell some of my friends about Rome and I would say, yeah, it's based on the graphs, I can connect all these things. They would immediately think of, you know, the usual, just X and Y graph, like the most simplistic empty graph, um, something you'd see in maths homework in high school or something.
[01:10:13] And it's not really as useful in terms of trying to create some sort of visual element to your explanation. Uh, so I was just curious, like, how would you handle that? But, okay.
[01:10:22] Mat McGann: Yeah, I should avoid the word graph. Actually. I think that's fine. But, but structure emerging from your notes, I think is quite, that's quite exciting as well, because you know, you don't know what you don't know and your notes contain more knowledge than they have just individually when you combine it up, there's more value there.
[01:10:41] I think everyone can understand that.
[01:10:43] Norman Chella: fantastic. And of course, final question. What does Roam mean to you?
[01:10:50] Mat McGann: ha uh, well, to me, it's the, uh, when, when I mentioned I'd been looking for the right note taking system for 20 years, that's literally true. I do feel what it means to me is I feel like this is the way to do it. This is the real breakthrough. You know when notes systems became electronic, that kind of sold a bunch of problems, but didn't solve all of them.
[01:11:16] It introduced new problems and it never felt like the flexibility that digital should give you was realized because people were still putting just notebooks in computers. They were still just digitizing notebooks. And I say, this is the true innovation was finally made in 2019, or whenever happened when the, uh, I guess what did happen?
[01:11:41] A notebook was splinted and there's no such thing as a notebook anymore. It's notes. It's a whole mess of notes that just gets. Uh, they have relationships between them and that's how note taking should be done on a computer. And so that's the genuine innovation. So it means a lot to me that that's been discovered. And I can use that.
[01:12:03] Norman Chella: And then you can build that entire startup on just Roam, which is still blowing my mind. Like I'm still trying to comprehend looking back at my. On the fly notes. I'm like, Oh my goodness. Oh, I have to read through it again. on that note,
[01:12:16] Mat McGann: I should give you a tour then at some stage, maybe that will be an easy way.
[01:12:20] Norman Chella: yeah, there also a couple of other people who would be eager To see you do a Roam tour. So if we want to contact you about how you use Roam for your startup or for health horizon, or pretty much how you interact with your own graph or anything that we talked about in this conversation, like, for example, like the Twain or your where can we find you and how do we contact you?
[01:12:43] Mat McGann: Twitter is the best. It links to that website. And, um, I've got a little bit of Roam stuff on there, but I'll certainly do more if, if people want it, but the Twitter handle is Matt McGann. Uh, one, one T. So mat, M C G A N N.
[01:12:57] Norman Chella: All right. And I will put that in the public roam graph right below. So Matt, thank you so much. And I will see you on Twitter.