Engineering
Features Will Not Save You. With Chris Ladd, CEO of Better Notes
By
Admin
on •
Nov 10, 2022
“Are you making something that is solving a real problem for people?” That’s the question on Chris Ladd’s mind whenever he’s tempted to design a new app feature. Chris’s iOS app ChordBank has been downloaded millions of times; it gives guitar players tools to expand on and improve their craft. With a chord library, lessons and games that use monophonic and polyphonic pitch detection, the helps players hone their skills and become better players.
Chris built the first version of ChordBank 10 years ago, back when he was working as a freelance journalist while getting into programming. With years of experience as an independent iOS developer under his belt and experience in product teams at The New York Times, Chris has honed his craft and built ChordBank into his full time job.
Here’s Chris Ladd on Crafted, Artium’s new podcast.
Listen and subscribe to Crafted: Apple Podcasts | Spotify | All Podcast Apps
Chris Ladd: Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? Like I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
Dan Blumberg: That's Chris Ladd, the founder of ChordBank, a mobile app for guitar players. The app features thousands of chord voicings up and down the fret board, as well as games and lessons. Chord Bank has been downloaded millions of times since Chris built the first version 10 years ago. On this episode, we'll learn how Chris's experience as a developer on the iOS team at the New York Times informs his work as the solo entrepreneur behind ChordBank. We'll zoom in to understand how and why Chris built ChordBank, including how he paired programs he wrote with a professional guitarist trained ear to create ChordBank's unique collection of chord voicings. And we'll zoom out as Chris details his need to better empathize and communicate with the players using the app.
Welcome to Crafted, a show about great products and the people who make them. I'm your host, Dan Blumberg. I'm a product and engagement lead at Artium where my colleagues and I help companies build fantastic software and dynamic teams.
So Chris, you've managed to combine two great passions of yours, playing guitar and developing apps into a business. And before we get into the ins and outs of ChordBank, I'd love for you to just share how'd you get into guitar?
Chris Ladd: Well, my dad always played guitar when I was growing up. And I must have started when I was 11 or 12 or something like that. When I had the dexterity to hold one, I would always be messing around with my dad's guitar. But when I started seriously trying to teach myself how to do this, probably in that preteen age, and this was back in the day of Usenet, and so I would be on these Usenet archives looking for transcriptions of songs that I could be learning, like Jackson Brown songs or Beetles songs. And I would go from there. So I'd have on the one hand a old chord book and on the other hand I would have these dot matrix printed askie tabs in a accordion binder. And those would essentially be my afternoons as a kid just trying to learn these songs.
Dan Blumberg: It's fascinating that even in those early days of pre-web internet, you were using it as a resource to learn. How did you get into coding? Are you also a self-taught coder?
Chris Ladd: Yeah, unfortunately self-taught coder as well. So I can learn how to do anything kind of good basically is where I'm at. I wouldn't call it programming, but I started doing like HTML stuff, making stuff for the web, making stuff where you type something and it makes something happen. In college, towards the end of my college years, I made an online magazine. And then after college, I was more focused on writing prose, like journalism, that kind of thing. And I remember very clearly I was at some bar in Cambridge with a friend and he was saying like, "Hey, you're creative, why don't you make iPhone apps?" This was like 2008, so it was just possible to make iPhone apps. And I remember thinking that's ridiculous. I didn't think of programming as a creative skill, and I just thought it's a commodity. Even if I learned how to do it, somebody else would learn how to do it more cheaply and that would be that.
So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
Dan Blumberg: But things changed for Chris when he built his first program. Around 2005 to 2011, Chris was working as a freelance journalist pitching stories to the Boston Globe, the New York Times, NPR and others. He was having trouble keeping track of which outlets he'd pitched stories to, who had said yes, who had said no, who had said no nicely, and when to follow up with them. So Chris decided to write a program to solve his own problem.
Chris Ladd: Getting into actually programming well took about two years, I think. And also, well is a relative term, but I think the first year that I was making stuff, you spend an hour looking for the problem and it's like a semicolon in the wrong place. It took about a year to get to the point where I could reliably type something, compile it, feel like it was going to work, have it actually work without crashing. And once I got to that phase and then I think it took another year to learn how to put the pieces together in a way that I could feel pretty confident about how things were going. So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
It wasn't about Steve Jobs or Apple or that stuff, it was more just being in a whole room full of people that did exactly what I did and feeling like I belonged there. And so that sense of belonging I feel like was a very important turning point for me.
Dan Blumberg: Thanks to connections Chris made at that Apple Developer's Conference, he landed a full-time job as a software engineer at the New York Times, where he worked on their early iPhone and iPad apps. This is how Chris and I first met. I was also at the Times as a mobile product manager and later director of product. Working as part of a team for the first time, Chris was pushed in new ways.
Chris Ladd: The timing was very important in me getting a job so soon after just learning how to do this stuff. And then the consequence of that, as you know, is working with a bunch of people, you get much better. So however good you think you are, if you join a company, you are being asked to fix problems that you never would've created in the first place, or you're being asked to work on features in directions that you never would've gone in. Advertising or some data storage things or whatever. And I think I'm sure the same thing is true in other disciplines too.
So I feel like the best thing about having a job is the colleagues that you're around you learn from. That's very important. But then also just being asked to do other stuff. It's like summer camp, there's always another activity that you show up and you have to do versus having your own things, you have to both make up the activities and do them. When you're making up the activities, the bias is towards things you actually are capable of doing versus that your employer has no idea what you're capable of doing, so they just ask you to do whatever they need done and you have to figure it out.
Dan Blumberg: One of Chris's first assignments was to get the iPad app to stop crashing. The app was built quickly under cloak and dagger conditions to be ready when Apple unveiled this new thing called the iPad.
Chris Ladd: There was a mysterious phone call that arrived at The Times and said, "Hey, we're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people? Sign this."
But it was written fast and it was written to get it done. And it was written to show it to Steve Jobs when he was in the room and not get yelled at. It's tough to build. This probably is applicable to the kind of work that you all do working with companies. It's like, there's a difference between working fast to figure out what the thing is and get it out the door and then building towards something. And the answer is not always to build sustainably. I don't know how many rewrites have happened since I joined the New York Times.
I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch.
Dan Blumberg: I'm a firm believer, we're firm believers at Artium, that sometimes you need to get something out the door to get user feedback and other times you need to shore up the foundation so you can scale. It was around the time that Chris started at The Times that he developed the earliest versions of ChordBank. Chris worked on the app alongside his full-time job. It took off and after six years he left the Times and turned his attention to ChordBank full-time.
Take us back to ChordBank one, or even if it's the predecessor to ChordBank one, what was the problem you were trying to solve? What was the first release? What did it look like?
Chris Ladd: So ChordBank, you grew up playing guitar, so you probably had some kind of a chord book at some point. And everybody grows up with a chord book. It's usually got 12 or 15 different kinds of chords in it. Major, minor, seventh, the half diminished, whatever. ChordBank was essentially always designed to be that, but in your pocket. The very first version of ChordBank was actually a part of another app that was all about guitar tabs and being able to have a library of those on your phone. And as part of that, there was a little button and when you clicked it you would pop up a thing and you could look up the chords in the songs that you were trying to play and it would try to figure out what you were playing and that kind of stuff. But people really liked that feature and it made sense to pull it out. So that ended up being its own app and that is the start of ChordBank and that was 2010.
And over the years that's grown from what was, I don't even remember how many chords there were, but there probably weren't more than 20 types of different chords in there. To today, there are 89 different types of chords, every single voicing has been actually played by an awesome guitarist who is not me. I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch. So over the years it's evolved to add more features like adding a scale library.
Today I think I have the opposite problem, which is I find it difficult to talk about what ChordBank is without using the word. And I find that to be a problem and I think my users probably find that to be a problem when they're talking about it with themselves. So a lot of my focus these days is what is this thing? What is a succinct version of what this thing is that I can create? What's missing? What's not useful in trying to evolve the product in the direction of the player?
Dan Blumberg: Could you give us a sense of scale, how many users are there? And then what are the different types of users that you're building for? And do you have one user type? Is it a novice, is it someone who's getting back into guitar after a layoff? Who is the person you think of most when you're answering questions like you were just asking?
Chris Ladd: Yeah, sure. I don't have a number for total number of users. In 10 years, it's been millions of downloads, but I think in terms of monthlies, it's probably honestly under a hundred thousand at this point. In terms of the kind of user I'm building for, when I think of who the app is for, I really think of somebody... the picture in my mind is a 62 year old dude who played when he was younger, loved playing, but it fell by the wayside and really, really wants to dedicate time to playing now.
That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I hope not to become that 62 year old you just described because I am that person, but a little bit younger. But also love playing, don't do it enough kids, et cetera. It has fallen by the wayside. Yeah, as I was playing with ChordBank, I was enjoying it. And my background is I played a lot of jazz, so I know the theory.
Chris Ladd: Oh cool.
Dan Blumberg: I know ninth chords and diminished and all that, but it's really rusty. And frankly, the stuff I should be working on today is really more sing along songs for the young kids that I have in the house.
Chris Ladd: Oh sure. That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I want to get into the details of the collaboration you mentioned with the Berkeley music professor, you talked about there are good chords and bad chords or easy chords to play and hard chords to play. And I'd love to understand a little more of how, I know you built program to figure out every single possible way to play like an F sharp minor seven chord, but then you didn't trust a machine to just suggest that so you work with this professor. Can you just dig into that collaboration a little bit more?
Chris Ladd: When I try to think about what is unique about ChordBank, I think it is that canonical collection. I think that is the core of it. So there's apps that go, I don't know if I'd say they go deep, but they have some level of curation, but it's over a smaller subset of cords. And then there's apps that go very broad that can mash together intervals in any possible way, but there's no human level of curation there. And you get some very funky shapes that are just uncomfortable or maybe not musically sound. So you can combine 12 notes in any way you want, but there's some combinations of those notes that sound good and are musically applicable.
Dan Blumberg: There are millions of combinations, but only some make sense musically. So Colin would define some rules like that a chord can't stretch the player's hands so far that it hurts, or that the root note has to be in a certain place. And then Chris would write code to codify those rules.
Chris Ladd: And so it creates all these and then it generates chords and diagrams for every single kind. And then there's a little private little app for the two of us, which puts them all in a big grid and he would go through and look at them and have his guitar in hand and play through them and say like, "Well this one doesn't work." And then we drill in on why not. And the cool thing is you end up with three things. You end up with a pile of chords that are, like what we called our spotlight voicings, that are great hand curated, officially okayed like an actual guitarist has played those voicings. You end up with a pile of chords that should not be played, the blacklist. And then, by the process of creating this stuff, you also generate a machine that does a pretty good job of creating cord voicings. The core of ChordBank is now something that I can honestly say is I would challenge somebody to find a better cannon of chords somewhere.
Dan Blumberg: So through this process of writing code, playing the chords and rewriting the code, Chris and Colin identified 89 chords and 40,000 voicings that guitarists could rely on, at least when their guitar is tuned the traditional way. But there are many, many ways to tune a guitar, and ChordBank's users had frequently requested help with chords in alternate tunings. Here's where Chris's programming really paid off. Instead of repeating the entire process with Colin to generate cords for dozens of alternate tunings, he let the machine do the work. Satisfied that his program would recommend musically sound cords, Chris launched a new app called Alt Tunings.
One of the other features that I love in the app is that you can play a chord and the app will recognize the chord and then you've built some cool games on top of that. So there's a word that as a missing letter and if you play an A chord, it finishes the word and confetti falls and you'll feel very proud of yourself.
Chris Ladd: Confetti was a big day, making the confetti.
Dan Blumberg: With that feature, how much of that is, in air quotes, out of the box iOS stuff that any app can do with using the microphone? And how much of that is custom work? Can you talk through some of those games and how you built them?
Chris Ladd: Sure. So the core of those games... So there's two types of things going on. They're building a guitar tune where it listens to one string and tells you whether it's sharp or flat. That's called monophonic pitch detection. And so that means you're looking for the pitch of a single note, and that's pretty well established and that's something that it's not out of the box on iOS, but there's a common enough problem that there are ways to solve it. I wouldn't say it's easy, but it's commonplace, let's say. Polyphonic where you're trying to get a whole chord to be recognized as true or not, I think that's, at least the way it's done in ChordBank, it's a little more art than science. It does both an FFT analysis where it takes all the sound that it hears and it tries to translate it from what's called the time domain to the frequency domain.
So when you have a sound that's recorded like the sound in this microphone, what actually happens is that it's measuring the pressure level that hits the diaphragm in a number between negative one and one. And it gets that number and at some number of times a second, that's the sample rate. So I didn't actually look at what you guys are recording, but I imagine it's either 48,000 or 44,100 times. Every second, you're getting another one of those numbers that comes in. And the way that you can get frequency from that is that you... Well, there's a few different algorithms that do. You can count the peaks and find the time between the peaks and then that time is your frequency. So that's how a lot of the monophonic things work. And you can also take a whole bunch of those samples, usually about like 8,000 or something like that, at a time. You run this analysis on it, it's called a four year transform, and then you turn it into, it looks like a bar chart and you can get the peaks in of where which frequencies are being played.
Dan Blumberg: Chris says the polyphonic pitch detection is still a work in progress and that he gives players the benefit of the doubt preferring to shower digital confetti when they're playing roughly the right chord rather than have them get mad at the app if it tells them they're playing something wrong when they're actually getting it right.
I want to go back to the early days of ChordBank, or what became ChordBank? What was the moment when you knew that you were onto something?
Chris Ladd: People would download it. That was a big deal to have people all over the world making something that you made in your apartment. I think that I would get emails that people loved the app, that was really good. I would also get emails about things that didn't go well, but I don't mind that either. I guess the difference between thinking about ChordBank as a side project or a business card and thinking about it as a thing that would essentially consume most of my working time, probably happened in 2012, maybe. It was during the switch to in-app purchases in the premium world, which we've seen a couple of these shifts over the app ecosystem. First, everything was a paid app and then there was free apps with advertising and then you'd have a free app and a paid app.
But when I switched everything to a freemium thing where you could try it first, the number of people... And because this was still when people were getting their first iPhones, so every Christmas was a big Christmas because what happens at Christmas? Everybody gets new guitars, everybody gets new iPhones, it's like a bonanza.
So just seeing that, the numbers were starting to be real numbers. When I started to see that I was making the same at home as I was at work at a good job in New York City, then I sort of started to feel like this is something I could make happen. So then it was like that was a good push.
Dan Blumberg: What is it like working on your own now? You mentioned the benefits before of working as part of a team, you called it summer camp. What's it like now where you're working independently?
Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists?"
Chris Ladd: It's awful. It's awful. No, it depends on which phase I'm in. So if I'm in a phase where I have a clear direction, I know what I'm producing, what I'm building, and I can just build, man, that is just delight. Especially if you're in that phase of development where, I'm sure you're familiar, 90% of any project takes half the time. Doing the first 90% of a project, whew, that's the best. If you can get to the sense where, you know were just making headway every single day, that is awesome. And then you get to that last 10% and it's like this took two weeks to fix this alignment issue, what is going on with the world? And you feel like it'll never be over. So anyway, that first 90% is great, even the last 10% is great.
But having to both choose direction and implement direction are tough things to do because they're totally different skills. It's a totally different vantage point of being up above and trying to see what things are and picture what the most important thing to do is. And by the way, how does that important thing that you're going to do tie into what you're going to do in a month in two months and tie into what you did six months ago?
Dan Blumberg: Is there something that you wish you had known earlier in your career that is helping you accelerate today?Chris Ladd: Yeah, I think that... and I know this and I have known this for a while, but I don't remember it until it's too late. It's just that features will not save you. It's very satisfying to make something new and make a new piece of the app, new feature a new thing, design it, it looks great, it works really well, it's not going to save you. I feel like somebody needs to give me a call, if you're available and could do this, that would be great. Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
And it's just this concept of, I think communication I feel is one of the things I do the worst job at, but recognize that it's so important and I feel like building is so easy and satisfying, but I just wish I could just cut it out and do a better job of communicating about the stuff that's already there.
Dan Blumberg: I love it. And I think you just gave us the title of this episode, Features will not save you.
Chris Ladd: Features will not save you.
Dan Blumberg: Yeah. Chris, thank you so much.
Chris Ladd: Yeah, of course. Thanks for having me.
Dan Blumberg: Chris Ladd is the CEO of Better Notes and the creator of ChordBank. And this is Crafted from Artium. At Artium, we help organizations build best in class products, industry leading software and collaborative cross-functional teams at scale. We artisans love partnering with creative people to build their visions of the future. If you've got an opportunity you'd like to discuss or just want to learn more about us, check us out at, thisisartium.com or email us at hello@thisisartium.com. Please subscribe to Crafted, and join us as we explore more great products and the craft of software development. I'm Dan Blumberg. See you next time.
“Are you making something that is solving a real problem for people?” That’s the question on Chris Ladd’s mind whenever he’s tempted to design a new app feature. Chris’s iOS app ChordBank has been downloaded millions of times; it gives guitar players tools to expand on and improve their craft. With a chord library, lessons and games that use monophonic and polyphonic pitch detection, the helps players hone their skills and become better players.
Chris built the first version of ChordBank 10 years ago, back when he was working as a freelance journalist while getting into programming. With years of experience as an independent iOS developer under his belt and experience in product teams at The New York Times, Chris has honed his craft and built ChordBank into his full time job.
Here’s Chris Ladd on Crafted, Artium’s new podcast.
Listen and subscribe to Crafted: Apple Podcasts | Spotify | All Podcast Apps
Chris Ladd: Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? Like I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
Dan Blumberg: That's Chris Ladd, the founder of ChordBank, a mobile app for guitar players. The app features thousands of chord voicings up and down the fret board, as well as games and lessons. Chord Bank has been downloaded millions of times since Chris built the first version 10 years ago. On this episode, we'll learn how Chris's experience as a developer on the iOS team at the New York Times informs his work as the solo entrepreneur behind ChordBank. We'll zoom in to understand how and why Chris built ChordBank, including how he paired programs he wrote with a professional guitarist trained ear to create ChordBank's unique collection of chord voicings. And we'll zoom out as Chris details his need to better empathize and communicate with the players using the app.
Welcome to Crafted, a show about great products and the people who make them. I'm your host, Dan Blumberg. I'm a product and engagement lead at Artium where my colleagues and I help companies build fantastic software and dynamic teams.
So Chris, you've managed to combine two great passions of yours, playing guitar and developing apps into a business. And before we get into the ins and outs of ChordBank, I'd love for you to just share how'd you get into guitar?
Chris Ladd: Well, my dad always played guitar when I was growing up. And I must have started when I was 11 or 12 or something like that. When I had the dexterity to hold one, I would always be messing around with my dad's guitar. But when I started seriously trying to teach myself how to do this, probably in that preteen age, and this was back in the day of Usenet, and so I would be on these Usenet archives looking for transcriptions of songs that I could be learning, like Jackson Brown songs or Beetles songs. And I would go from there. So I'd have on the one hand a old chord book and on the other hand I would have these dot matrix printed askie tabs in a accordion binder. And those would essentially be my afternoons as a kid just trying to learn these songs.
Dan Blumberg: It's fascinating that even in those early days of pre-web internet, you were using it as a resource to learn. How did you get into coding? Are you also a self-taught coder?
Chris Ladd: Yeah, unfortunately self-taught coder as well. So I can learn how to do anything kind of good basically is where I'm at. I wouldn't call it programming, but I started doing like HTML stuff, making stuff for the web, making stuff where you type something and it makes something happen. In college, towards the end of my college years, I made an online magazine. And then after college, I was more focused on writing prose, like journalism, that kind of thing. And I remember very clearly I was at some bar in Cambridge with a friend and he was saying like, "Hey, you're creative, why don't you make iPhone apps?" This was like 2008, so it was just possible to make iPhone apps. And I remember thinking that's ridiculous. I didn't think of programming as a creative skill, and I just thought it's a commodity. Even if I learned how to do it, somebody else would learn how to do it more cheaply and that would be that.
So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
Dan Blumberg: But things changed for Chris when he built his first program. Around 2005 to 2011, Chris was working as a freelance journalist pitching stories to the Boston Globe, the New York Times, NPR and others. He was having trouble keeping track of which outlets he'd pitched stories to, who had said yes, who had said no, who had said no nicely, and when to follow up with them. So Chris decided to write a program to solve his own problem.
Chris Ladd: Getting into actually programming well took about two years, I think. And also, well is a relative term, but I think the first year that I was making stuff, you spend an hour looking for the problem and it's like a semicolon in the wrong place. It took about a year to get to the point where I could reliably type something, compile it, feel like it was going to work, have it actually work without crashing. And once I got to that phase and then I think it took another year to learn how to put the pieces together in a way that I could feel pretty confident about how things were going. So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
It wasn't about Steve Jobs or Apple or that stuff, it was more just being in a whole room full of people that did exactly what I did and feeling like I belonged there. And so that sense of belonging I feel like was a very important turning point for me.
Dan Blumberg: Thanks to connections Chris made at that Apple Developer's Conference, he landed a full-time job as a software engineer at the New York Times, where he worked on their early iPhone and iPad apps. This is how Chris and I first met. I was also at the Times as a mobile product manager and later director of product. Working as part of a team for the first time, Chris was pushed in new ways.
Chris Ladd: The timing was very important in me getting a job so soon after just learning how to do this stuff. And then the consequence of that, as you know, is working with a bunch of people, you get much better. So however good you think you are, if you join a company, you are being asked to fix problems that you never would've created in the first place, or you're being asked to work on features in directions that you never would've gone in. Advertising or some data storage things or whatever. And I think I'm sure the same thing is true in other disciplines too.
So I feel like the best thing about having a job is the colleagues that you're around you learn from. That's very important. But then also just being asked to do other stuff. It's like summer camp, there's always another activity that you show up and you have to do versus having your own things, you have to both make up the activities and do them. When you're making up the activities, the bias is towards things you actually are capable of doing versus that your employer has no idea what you're capable of doing, so they just ask you to do whatever they need done and you have to figure it out.
Dan Blumberg: One of Chris's first assignments was to get the iPad app to stop crashing. The app was built quickly under cloak and dagger conditions to be ready when Apple unveiled this new thing called the iPad.
Chris Ladd: There was a mysterious phone call that arrived at The Times and said, "Hey, we're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people? Sign this."
But it was written fast and it was written to get it done. And it was written to show it to Steve Jobs when he was in the room and not get yelled at. It's tough to build. This probably is applicable to the kind of work that you all do working with companies. It's like, there's a difference between working fast to figure out what the thing is and get it out the door and then building towards something. And the answer is not always to build sustainably. I don't know how many rewrites have happened since I joined the New York Times.
I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch.
Dan Blumberg: I'm a firm believer, we're firm believers at Artium, that sometimes you need to get something out the door to get user feedback and other times you need to shore up the foundation so you can scale. It was around the time that Chris started at The Times that he developed the earliest versions of ChordBank. Chris worked on the app alongside his full-time job. It took off and after six years he left the Times and turned his attention to ChordBank full-time.
Take us back to ChordBank one, or even if it's the predecessor to ChordBank one, what was the problem you were trying to solve? What was the first release? What did it look like?
Chris Ladd: So ChordBank, you grew up playing guitar, so you probably had some kind of a chord book at some point. And everybody grows up with a chord book. It's usually got 12 or 15 different kinds of chords in it. Major, minor, seventh, the half diminished, whatever. ChordBank was essentially always designed to be that, but in your pocket. The very first version of ChordBank was actually a part of another app that was all about guitar tabs and being able to have a library of those on your phone. And as part of that, there was a little button and when you clicked it you would pop up a thing and you could look up the chords in the songs that you were trying to play and it would try to figure out what you were playing and that kind of stuff. But people really liked that feature and it made sense to pull it out. So that ended up being its own app and that is the start of ChordBank and that was 2010.
And over the years that's grown from what was, I don't even remember how many chords there were, but there probably weren't more than 20 types of different chords in there. To today, there are 89 different types of chords, every single voicing has been actually played by an awesome guitarist who is not me. I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch. So over the years it's evolved to add more features like adding a scale library.
Today I think I have the opposite problem, which is I find it difficult to talk about what ChordBank is without using the word. And I find that to be a problem and I think my users probably find that to be a problem when they're talking about it with themselves. So a lot of my focus these days is what is this thing? What is a succinct version of what this thing is that I can create? What's missing? What's not useful in trying to evolve the product in the direction of the player?
Dan Blumberg: Could you give us a sense of scale, how many users are there? And then what are the different types of users that you're building for? And do you have one user type? Is it a novice, is it someone who's getting back into guitar after a layoff? Who is the person you think of most when you're answering questions like you were just asking?
Chris Ladd: Yeah, sure. I don't have a number for total number of users. In 10 years, it's been millions of downloads, but I think in terms of monthlies, it's probably honestly under a hundred thousand at this point. In terms of the kind of user I'm building for, when I think of who the app is for, I really think of somebody... the picture in my mind is a 62 year old dude who played when he was younger, loved playing, but it fell by the wayside and really, really wants to dedicate time to playing now.
That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I hope not to become that 62 year old you just described because I am that person, but a little bit younger. But also love playing, don't do it enough kids, et cetera. It has fallen by the wayside. Yeah, as I was playing with ChordBank, I was enjoying it. And my background is I played a lot of jazz, so I know the theory.
Chris Ladd: Oh cool.
Dan Blumberg: I know ninth chords and diminished and all that, but it's really rusty. And frankly, the stuff I should be working on today is really more sing along songs for the young kids that I have in the house.
Chris Ladd: Oh sure. That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I want to get into the details of the collaboration you mentioned with the Berkeley music professor, you talked about there are good chords and bad chords or easy chords to play and hard chords to play. And I'd love to understand a little more of how, I know you built program to figure out every single possible way to play like an F sharp minor seven chord, but then you didn't trust a machine to just suggest that so you work with this professor. Can you just dig into that collaboration a little bit more?
Chris Ladd: When I try to think about what is unique about ChordBank, I think it is that canonical collection. I think that is the core of it. So there's apps that go, I don't know if I'd say they go deep, but they have some level of curation, but it's over a smaller subset of cords. And then there's apps that go very broad that can mash together intervals in any possible way, but there's no human level of curation there. And you get some very funky shapes that are just uncomfortable or maybe not musically sound. So you can combine 12 notes in any way you want, but there's some combinations of those notes that sound good and are musically applicable.
Dan Blumberg: There are millions of combinations, but only some make sense musically. So Colin would define some rules like that a chord can't stretch the player's hands so far that it hurts, or that the root note has to be in a certain place. And then Chris would write code to codify those rules.
Chris Ladd: And so it creates all these and then it generates chords and diagrams for every single kind. And then there's a little private little app for the two of us, which puts them all in a big grid and he would go through and look at them and have his guitar in hand and play through them and say like, "Well this one doesn't work." And then we drill in on why not. And the cool thing is you end up with three things. You end up with a pile of chords that are, like what we called our spotlight voicings, that are great hand curated, officially okayed like an actual guitarist has played those voicings. You end up with a pile of chords that should not be played, the blacklist. And then, by the process of creating this stuff, you also generate a machine that does a pretty good job of creating cord voicings. The core of ChordBank is now something that I can honestly say is I would challenge somebody to find a better cannon of chords somewhere.
Dan Blumberg: So through this process of writing code, playing the chords and rewriting the code, Chris and Colin identified 89 chords and 40,000 voicings that guitarists could rely on, at least when their guitar is tuned the traditional way. But there are many, many ways to tune a guitar, and ChordBank's users had frequently requested help with chords in alternate tunings. Here's where Chris's programming really paid off. Instead of repeating the entire process with Colin to generate cords for dozens of alternate tunings, he let the machine do the work. Satisfied that his program would recommend musically sound cords, Chris launched a new app called Alt Tunings.
One of the other features that I love in the app is that you can play a chord and the app will recognize the chord and then you've built some cool games on top of that. So there's a word that as a missing letter and if you play an A chord, it finishes the word and confetti falls and you'll feel very proud of yourself.
Chris Ladd: Confetti was a big day, making the confetti.
Dan Blumberg: With that feature, how much of that is, in air quotes, out of the box iOS stuff that any app can do with using the microphone? And how much of that is custom work? Can you talk through some of those games and how you built them?
Chris Ladd: Sure. So the core of those games... So there's two types of things going on. They're building a guitar tune where it listens to one string and tells you whether it's sharp or flat. That's called monophonic pitch detection. And so that means you're looking for the pitch of a single note, and that's pretty well established and that's something that it's not out of the box on iOS, but there's a common enough problem that there are ways to solve it. I wouldn't say it's easy, but it's commonplace, let's say. Polyphonic where you're trying to get a whole chord to be recognized as true or not, I think that's, at least the way it's done in ChordBank, it's a little more art than science. It does both an FFT analysis where it takes all the sound that it hears and it tries to translate it from what's called the time domain to the frequency domain.
So when you have a sound that's recorded like the sound in this microphone, what actually happens is that it's measuring the pressure level that hits the diaphragm in a number between negative one and one. And it gets that number and at some number of times a second, that's the sample rate. So I didn't actually look at what you guys are recording, but I imagine it's either 48,000 or 44,100 times. Every second, you're getting another one of those numbers that comes in. And the way that you can get frequency from that is that you... Well, there's a few different algorithms that do. You can count the peaks and find the time between the peaks and then that time is your frequency. So that's how a lot of the monophonic things work. And you can also take a whole bunch of those samples, usually about like 8,000 or something like that, at a time. You run this analysis on it, it's called a four year transform, and then you turn it into, it looks like a bar chart and you can get the peaks in of where which frequencies are being played.
Dan Blumberg: Chris says the polyphonic pitch detection is still a work in progress and that he gives players the benefit of the doubt preferring to shower digital confetti when they're playing roughly the right chord rather than have them get mad at the app if it tells them they're playing something wrong when they're actually getting it right.
I want to go back to the early days of ChordBank, or what became ChordBank? What was the moment when you knew that you were onto something?
Chris Ladd: People would download it. That was a big deal to have people all over the world making something that you made in your apartment. I think that I would get emails that people loved the app, that was really good. I would also get emails about things that didn't go well, but I don't mind that either. I guess the difference between thinking about ChordBank as a side project or a business card and thinking about it as a thing that would essentially consume most of my working time, probably happened in 2012, maybe. It was during the switch to in-app purchases in the premium world, which we've seen a couple of these shifts over the app ecosystem. First, everything was a paid app and then there was free apps with advertising and then you'd have a free app and a paid app.
But when I switched everything to a freemium thing where you could try it first, the number of people... And because this was still when people were getting their first iPhones, so every Christmas was a big Christmas because what happens at Christmas? Everybody gets new guitars, everybody gets new iPhones, it's like a bonanza.
So just seeing that, the numbers were starting to be real numbers. When I started to see that I was making the same at home as I was at work at a good job in New York City, then I sort of started to feel like this is something I could make happen. So then it was like that was a good push.
Dan Blumberg: What is it like working on your own now? You mentioned the benefits before of working as part of a team, you called it summer camp. What's it like now where you're working independently?
Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists?"
Chris Ladd: It's awful. It's awful. No, it depends on which phase I'm in. So if I'm in a phase where I have a clear direction, I know what I'm producing, what I'm building, and I can just build, man, that is just delight. Especially if you're in that phase of development where, I'm sure you're familiar, 90% of any project takes half the time. Doing the first 90% of a project, whew, that's the best. If you can get to the sense where, you know were just making headway every single day, that is awesome. And then you get to that last 10% and it's like this took two weeks to fix this alignment issue, what is going on with the world? And you feel like it'll never be over. So anyway, that first 90% is great, even the last 10% is great.
But having to both choose direction and implement direction are tough things to do because they're totally different skills. It's a totally different vantage point of being up above and trying to see what things are and picture what the most important thing to do is. And by the way, how does that important thing that you're going to do tie into what you're going to do in a month in two months and tie into what you did six months ago?
Dan Blumberg: Is there something that you wish you had known earlier in your career that is helping you accelerate today?Chris Ladd: Yeah, I think that... and I know this and I have known this for a while, but I don't remember it until it's too late. It's just that features will not save you. It's very satisfying to make something new and make a new piece of the app, new feature a new thing, design it, it looks great, it works really well, it's not going to save you. I feel like somebody needs to give me a call, if you're available and could do this, that would be great. Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
And it's just this concept of, I think communication I feel is one of the things I do the worst job at, but recognize that it's so important and I feel like building is so easy and satisfying, but I just wish I could just cut it out and do a better job of communicating about the stuff that's already there.
Dan Blumberg: I love it. And I think you just gave us the title of this episode, Features will not save you.
Chris Ladd: Features will not save you.
Dan Blumberg: Yeah. Chris, thank you so much.
Chris Ladd: Yeah, of course. Thanks for having me.
Dan Blumberg: Chris Ladd is the CEO of Better Notes and the creator of ChordBank. And this is Crafted from Artium. At Artium, we help organizations build best in class products, industry leading software and collaborative cross-functional teams at scale. We artisans love partnering with creative people to build their visions of the future. If you've got an opportunity you'd like to discuss or just want to learn more about us, check us out at, thisisartium.com or email us at hello@thisisartium.com. Please subscribe to Crafted, and join us as we explore more great products and the craft of software development. I'm Dan Blumberg. See you next time.
“Are you making something that is solving a real problem for people?” That’s the question on Chris Ladd’s mind whenever he’s tempted to design a new app feature. Chris’s iOS app ChordBank has been downloaded millions of times; it gives guitar players tools to expand on and improve their craft. With a chord library, lessons and games that use monophonic and polyphonic pitch detection, the helps players hone their skills and become better players.
Chris built the first version of ChordBank 10 years ago, back when he was working as a freelance journalist while getting into programming. With years of experience as an independent iOS developer under his belt and experience in product teams at The New York Times, Chris has honed his craft and built ChordBank into his full time job.
Here’s Chris Ladd on Crafted, Artium’s new podcast.
Listen and subscribe to Crafted: Apple Podcasts | Spotify | All Podcast Apps
Chris Ladd: Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? Like I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
Dan Blumberg: That's Chris Ladd, the founder of ChordBank, a mobile app for guitar players. The app features thousands of chord voicings up and down the fret board, as well as games and lessons. Chord Bank has been downloaded millions of times since Chris built the first version 10 years ago. On this episode, we'll learn how Chris's experience as a developer on the iOS team at the New York Times informs his work as the solo entrepreneur behind ChordBank. We'll zoom in to understand how and why Chris built ChordBank, including how he paired programs he wrote with a professional guitarist trained ear to create ChordBank's unique collection of chord voicings. And we'll zoom out as Chris details his need to better empathize and communicate with the players using the app.
Welcome to Crafted, a show about great products and the people who make them. I'm your host, Dan Blumberg. I'm a product and engagement lead at Artium where my colleagues and I help companies build fantastic software and dynamic teams.
So Chris, you've managed to combine two great passions of yours, playing guitar and developing apps into a business. And before we get into the ins and outs of ChordBank, I'd love for you to just share how'd you get into guitar?
Chris Ladd: Well, my dad always played guitar when I was growing up. And I must have started when I was 11 or 12 or something like that. When I had the dexterity to hold one, I would always be messing around with my dad's guitar. But when I started seriously trying to teach myself how to do this, probably in that preteen age, and this was back in the day of Usenet, and so I would be on these Usenet archives looking for transcriptions of songs that I could be learning, like Jackson Brown songs or Beetles songs. And I would go from there. So I'd have on the one hand a old chord book and on the other hand I would have these dot matrix printed askie tabs in a accordion binder. And those would essentially be my afternoons as a kid just trying to learn these songs.
Dan Blumberg: It's fascinating that even in those early days of pre-web internet, you were using it as a resource to learn. How did you get into coding? Are you also a self-taught coder?
Chris Ladd: Yeah, unfortunately self-taught coder as well. So I can learn how to do anything kind of good basically is where I'm at. I wouldn't call it programming, but I started doing like HTML stuff, making stuff for the web, making stuff where you type something and it makes something happen. In college, towards the end of my college years, I made an online magazine. And then after college, I was more focused on writing prose, like journalism, that kind of thing. And I remember very clearly I was at some bar in Cambridge with a friend and he was saying like, "Hey, you're creative, why don't you make iPhone apps?" This was like 2008, so it was just possible to make iPhone apps. And I remember thinking that's ridiculous. I didn't think of programming as a creative skill, and I just thought it's a commodity. Even if I learned how to do it, somebody else would learn how to do it more cheaply and that would be that.
So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
Dan Blumberg: But things changed for Chris when he built his first program. Around 2005 to 2011, Chris was working as a freelance journalist pitching stories to the Boston Globe, the New York Times, NPR and others. He was having trouble keeping track of which outlets he'd pitched stories to, who had said yes, who had said no, who had said no nicely, and when to follow up with them. So Chris decided to write a program to solve his own problem.
Chris Ladd: Getting into actually programming well took about two years, I think. And also, well is a relative term, but I think the first year that I was making stuff, you spend an hour looking for the problem and it's like a semicolon in the wrong place. It took about a year to get to the point where I could reliably type something, compile it, feel like it was going to work, have it actually work without crashing. And once I got to that phase and then I think it took another year to learn how to put the pieces together in a way that I could feel pretty confident about how things were going. So a year to not be terrible, and then a year to move from terrible to good. That was a two year span from early 2009 to 2011. And in 2011, I made a big gamble on myself at the time, which was to spend $1,600 plus airfare plus hotel plus plus plus and go to the Apple conference in California to WWDC. And that was an awakening.
It wasn't about Steve Jobs or Apple or that stuff, it was more just being in a whole room full of people that did exactly what I did and feeling like I belonged there. And so that sense of belonging I feel like was a very important turning point for me.
Dan Blumberg: Thanks to connections Chris made at that Apple Developer's Conference, he landed a full-time job as a software engineer at the New York Times, where he worked on their early iPhone and iPad apps. This is how Chris and I first met. I was also at the Times as a mobile product manager and later director of product. Working as part of a team for the first time, Chris was pushed in new ways.
Chris Ladd: The timing was very important in me getting a job so soon after just learning how to do this stuff. And then the consequence of that, as you know, is working with a bunch of people, you get much better. So however good you think you are, if you join a company, you are being asked to fix problems that you never would've created in the first place, or you're being asked to work on features in directions that you never would've gone in. Advertising or some data storage things or whatever. And I think I'm sure the same thing is true in other disciplines too.
So I feel like the best thing about having a job is the colleagues that you're around you learn from. That's very important. But then also just being asked to do other stuff. It's like summer camp, there's always another activity that you show up and you have to do versus having your own things, you have to both make up the activities and do them. When you're making up the activities, the bias is towards things you actually are capable of doing versus that your employer has no idea what you're capable of doing, so they just ask you to do whatever they need done and you have to figure it out.
Dan Blumberg: One of Chris's first assignments was to get the iPad app to stop crashing. The app was built quickly under cloak and dagger conditions to be ready when Apple unveiled this new thing called the iPad.
Chris Ladd: There was a mysterious phone call that arrived at The Times and said, "Hey, we're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people?" And they said, "What are you working on?" And they said, "We're working on something, can you send out some people? Sign this."
But it was written fast and it was written to get it done. And it was written to show it to Steve Jobs when he was in the room and not get yelled at. It's tough to build. This probably is applicable to the kind of work that you all do working with companies. It's like, there's a difference between working fast to figure out what the thing is and get it out the door and then building towards something. And the answer is not always to build sustainably. I don't know how many rewrites have happened since I joined the New York Times.
I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch.
Dan Blumberg: I'm a firm believer, we're firm believers at Artium, that sometimes you need to get something out the door to get user feedback and other times you need to shore up the foundation so you can scale. It was around the time that Chris started at The Times that he developed the earliest versions of ChordBank. Chris worked on the app alongside his full-time job. It took off and after six years he left the Times and turned his attention to ChordBank full-time.
Take us back to ChordBank one, or even if it's the predecessor to ChordBank one, what was the problem you were trying to solve? What was the first release? What did it look like?
Chris Ladd: So ChordBank, you grew up playing guitar, so you probably had some kind of a chord book at some point. And everybody grows up with a chord book. It's usually got 12 or 15 different kinds of chords in it. Major, minor, seventh, the half diminished, whatever. ChordBank was essentially always designed to be that, but in your pocket. The very first version of ChordBank was actually a part of another app that was all about guitar tabs and being able to have a library of those on your phone. And as part of that, there was a little button and when you clicked it you would pop up a thing and you could look up the chords in the songs that you were trying to play and it would try to figure out what you were playing and that kind of stuff. But people really liked that feature and it made sense to pull it out. So that ended up being its own app and that is the start of ChordBank and that was 2010.
And over the years that's grown from what was, I don't even remember how many chords there were, but there probably weren't more than 20 types of different chords in there. To today, there are 89 different types of chords, every single voicing has been actually played by an awesome guitarist who is not me. I worked with a wonderful professor at Berkeley College of Music named Colin Sapp to really curate this library took a year and a half. So the total number, once you transpose them and do all the things you want to do, it's over 40,000 different voicings of guitar chords that are all top notch. So over the years it's evolved to add more features like adding a scale library.
Today I think I have the opposite problem, which is I find it difficult to talk about what ChordBank is without using the word. And I find that to be a problem and I think my users probably find that to be a problem when they're talking about it with themselves. So a lot of my focus these days is what is this thing? What is a succinct version of what this thing is that I can create? What's missing? What's not useful in trying to evolve the product in the direction of the player?
Dan Blumberg: Could you give us a sense of scale, how many users are there? And then what are the different types of users that you're building for? And do you have one user type? Is it a novice, is it someone who's getting back into guitar after a layoff? Who is the person you think of most when you're answering questions like you were just asking?
Chris Ladd: Yeah, sure. I don't have a number for total number of users. In 10 years, it's been millions of downloads, but I think in terms of monthlies, it's probably honestly under a hundred thousand at this point. In terms of the kind of user I'm building for, when I think of who the app is for, I really think of somebody... the picture in my mind is a 62 year old dude who played when he was younger, loved playing, but it fell by the wayside and really, really wants to dedicate time to playing now.
That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I hope not to become that 62 year old you just described because I am that person, but a little bit younger. But also love playing, don't do it enough kids, et cetera. It has fallen by the wayside. Yeah, as I was playing with ChordBank, I was enjoying it. And my background is I played a lot of jazz, so I know the theory.
Chris Ladd: Oh cool.
Dan Blumberg: I know ninth chords and diminished and all that, but it's really rusty. And frankly, the stuff I should be working on today is really more sing along songs for the young kids that I have in the house.
Chris Ladd: Oh sure. That's the underlying tension right now is that I have to fight building new stuff. So I'm always having ideas about new features to build new stuff like that and I really have to pull it back and say like, "No, you need to focus on copy." And it's just awful.
Dan Blumberg: I want to get into the details of the collaboration you mentioned with the Berkeley music professor, you talked about there are good chords and bad chords or easy chords to play and hard chords to play. And I'd love to understand a little more of how, I know you built program to figure out every single possible way to play like an F sharp minor seven chord, but then you didn't trust a machine to just suggest that so you work with this professor. Can you just dig into that collaboration a little bit more?
Chris Ladd: When I try to think about what is unique about ChordBank, I think it is that canonical collection. I think that is the core of it. So there's apps that go, I don't know if I'd say they go deep, but they have some level of curation, but it's over a smaller subset of cords. And then there's apps that go very broad that can mash together intervals in any possible way, but there's no human level of curation there. And you get some very funky shapes that are just uncomfortable or maybe not musically sound. So you can combine 12 notes in any way you want, but there's some combinations of those notes that sound good and are musically applicable.
Dan Blumberg: There are millions of combinations, but only some make sense musically. So Colin would define some rules like that a chord can't stretch the player's hands so far that it hurts, or that the root note has to be in a certain place. And then Chris would write code to codify those rules.
Chris Ladd: And so it creates all these and then it generates chords and diagrams for every single kind. And then there's a little private little app for the two of us, which puts them all in a big grid and he would go through and look at them and have his guitar in hand and play through them and say like, "Well this one doesn't work." And then we drill in on why not. And the cool thing is you end up with three things. You end up with a pile of chords that are, like what we called our spotlight voicings, that are great hand curated, officially okayed like an actual guitarist has played those voicings. You end up with a pile of chords that should not be played, the blacklist. And then, by the process of creating this stuff, you also generate a machine that does a pretty good job of creating cord voicings. The core of ChordBank is now something that I can honestly say is I would challenge somebody to find a better cannon of chords somewhere.
Dan Blumberg: So through this process of writing code, playing the chords and rewriting the code, Chris and Colin identified 89 chords and 40,000 voicings that guitarists could rely on, at least when their guitar is tuned the traditional way. But there are many, many ways to tune a guitar, and ChordBank's users had frequently requested help with chords in alternate tunings. Here's where Chris's programming really paid off. Instead of repeating the entire process with Colin to generate cords for dozens of alternate tunings, he let the machine do the work. Satisfied that his program would recommend musically sound cords, Chris launched a new app called Alt Tunings.
One of the other features that I love in the app is that you can play a chord and the app will recognize the chord and then you've built some cool games on top of that. So there's a word that as a missing letter and if you play an A chord, it finishes the word and confetti falls and you'll feel very proud of yourself.
Chris Ladd: Confetti was a big day, making the confetti.
Dan Blumberg: With that feature, how much of that is, in air quotes, out of the box iOS stuff that any app can do with using the microphone? And how much of that is custom work? Can you talk through some of those games and how you built them?
Chris Ladd: Sure. So the core of those games... So there's two types of things going on. They're building a guitar tune where it listens to one string and tells you whether it's sharp or flat. That's called monophonic pitch detection. And so that means you're looking for the pitch of a single note, and that's pretty well established and that's something that it's not out of the box on iOS, but there's a common enough problem that there are ways to solve it. I wouldn't say it's easy, but it's commonplace, let's say. Polyphonic where you're trying to get a whole chord to be recognized as true or not, I think that's, at least the way it's done in ChordBank, it's a little more art than science. It does both an FFT analysis where it takes all the sound that it hears and it tries to translate it from what's called the time domain to the frequency domain.
So when you have a sound that's recorded like the sound in this microphone, what actually happens is that it's measuring the pressure level that hits the diaphragm in a number between negative one and one. And it gets that number and at some number of times a second, that's the sample rate. So I didn't actually look at what you guys are recording, but I imagine it's either 48,000 or 44,100 times. Every second, you're getting another one of those numbers that comes in. And the way that you can get frequency from that is that you... Well, there's a few different algorithms that do. You can count the peaks and find the time between the peaks and then that time is your frequency. So that's how a lot of the monophonic things work. And you can also take a whole bunch of those samples, usually about like 8,000 or something like that, at a time. You run this analysis on it, it's called a four year transform, and then you turn it into, it looks like a bar chart and you can get the peaks in of where which frequencies are being played.
Dan Blumberg: Chris says the polyphonic pitch detection is still a work in progress and that he gives players the benefit of the doubt preferring to shower digital confetti when they're playing roughly the right chord rather than have them get mad at the app if it tells them they're playing something wrong when they're actually getting it right.
I want to go back to the early days of ChordBank, or what became ChordBank? What was the moment when you knew that you were onto something?
Chris Ladd: People would download it. That was a big deal to have people all over the world making something that you made in your apartment. I think that I would get emails that people loved the app, that was really good. I would also get emails about things that didn't go well, but I don't mind that either. I guess the difference between thinking about ChordBank as a side project or a business card and thinking about it as a thing that would essentially consume most of my working time, probably happened in 2012, maybe. It was during the switch to in-app purchases in the premium world, which we've seen a couple of these shifts over the app ecosystem. First, everything was a paid app and then there was free apps with advertising and then you'd have a free app and a paid app.
But when I switched everything to a freemium thing where you could try it first, the number of people... And because this was still when people were getting their first iPhones, so every Christmas was a big Christmas because what happens at Christmas? Everybody gets new guitars, everybody gets new iPhones, it's like a bonanza.
So just seeing that, the numbers were starting to be real numbers. When I started to see that I was making the same at home as I was at work at a good job in New York City, then I sort of started to feel like this is something I could make happen. So then it was like that was a good push.
Dan Blumberg: What is it like working on your own now? You mentioned the benefits before of working as part of a team, you called it summer camp. What's it like now where you're working independently?
Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists?"
Chris Ladd: It's awful. It's awful. No, it depends on which phase I'm in. So if I'm in a phase where I have a clear direction, I know what I'm producing, what I'm building, and I can just build, man, that is just delight. Especially if you're in that phase of development where, I'm sure you're familiar, 90% of any project takes half the time. Doing the first 90% of a project, whew, that's the best. If you can get to the sense where, you know were just making headway every single day, that is awesome. And then you get to that last 10% and it's like this took two weeks to fix this alignment issue, what is going on with the world? And you feel like it'll never be over. So anyway, that first 90% is great, even the last 10% is great.
But having to both choose direction and implement direction are tough things to do because they're totally different skills. It's a totally different vantage point of being up above and trying to see what things are and picture what the most important thing to do is. And by the way, how does that important thing that you're going to do tie into what you're going to do in a month in two months and tie into what you did six months ago?
Dan Blumberg: Is there something that you wish you had known earlier in your career that is helping you accelerate today?Chris Ladd: Yeah, I think that... and I know this and I have known this for a while, but I don't remember it until it's too late. It's just that features will not save you. It's very satisfying to make something new and make a new piece of the app, new feature a new thing, design it, it looks great, it works really well, it's not going to save you. I feel like somebody needs to give me a call, if you're available and could do this, that would be great. Somebody needs to call me monthly and just tell me, "Features will not save you." Are you making something that is solving a real problem for people? Are you telling them that it exists? I have people that email me after having the app for five years and say like, "Thank you for adding this feature. I've wanted it forever. It changes the way I use the app. I use it every day." The feature has been there for eight years.
And it's just this concept of, I think communication I feel is one of the things I do the worst job at, but recognize that it's so important and I feel like building is so easy and satisfying, but I just wish I could just cut it out and do a better job of communicating about the stuff that's already there.
Dan Blumberg: I love it. And I think you just gave us the title of this episode, Features will not save you.
Chris Ladd: Features will not save you.
Dan Blumberg: Yeah. Chris, thank you so much.
Chris Ladd: Yeah, of course. Thanks for having me.
Dan Blumberg: Chris Ladd is the CEO of Better Notes and the creator of ChordBank. And this is Crafted from Artium. At Artium, we help organizations build best in class products, industry leading software and collaborative cross-functional teams at scale. We artisans love partnering with creative people to build their visions of the future. If you've got an opportunity you'd like to discuss or just want to learn more about us, check us out at, thisisartium.com or email us at hello@thisisartium.com. Please subscribe to Crafted, and join us as we explore more great products and the craft of software development. I'm Dan Blumberg. See you next time.