Rails Business
Brendan Buckingham and Ryan Frisch talk about developing with Ruby on Rails and how to leverage it to build a business.
Rails Business
Effective Project Management and Handling Tech Debt
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this episode, the hosts explore the challenges of technical debt (tech debt) and delve deeply into effective project management strategies. They highlight the importance of categorizing tasks using a backlog system and explain how they utilize tools like Trello for tracking. Additionally, they explore the PARA method for organizing digital life and managing projects effectively.
00:00 Introduction to Tech Debt
00:11 Defining Tech Debt
00:20 Performance Issues as Tech Debt
02:36 Bugs vs. Design Defects
03:27 Prioritizing and Managing Tech Debt
06:20 Organizing the Backlog
07:05 Using Trello for Project Management
15:19 The PARA Method
19:09 Daily Workflow and Prioritization
23:01 Final Thoughts on Tech Debt
I was thinking we could talk about some, tech debt, I don't really have a better term, but it always feels like a, kind of a weird term. But anyways, tech debt's kind of what's on my mind lately.
Ryan:what does it mean to you?
Brendan:Well it, it means that we've got a lot of things that either need to be fixed or maintained or essentially it's, it's not new features, right? It's like all the things that need to be done to keep the app running efficiently or, you know, I guess let's just start with what, one of the things I've been thinking of a lot, which is performance of the app, mainly like Active Record queries and how. When you initially write some of those, they might be fine because of the volume of customers or the size of the app, all those different things. But as the app grows, those can become tech debt because now they're maybe not efficient enough. The database has grown a lot more, you know, a lot of different variables that play there, but. That's one form at least that I consider tech debt that I see more often than not. Granted, there's a lot of different, you know, flavors of tech debt, if you will.
Ryan:That's interesting because.
Brendan:definition of it?
Ryan:Is wrought with confusion because it can mean so many different things. And I, I don't think there's a right or wrong way to define it. I think it's just a way of saying, Hey, there's stuff that we know is not great that we need to improve. I like to break it down into different buckets though, right? Like, I forget who said it and it's from a famous. Book about TDD, but it's like, make it work, make it right, make it fast. you know, make it fast in terms of performance is like the last step. Like you don't wanna over optimize it at the beginning. So like, it depends on,
Brendan:Yep.
Ryan:I would sometimes call that tech debt and sometimes I would just call that growth. Right. And it's semantics which one you wanna use, but like. Performance optimizations. Right, but is important is you're measuring it against, you know, some sort of. You know, an average response time or an average load time, or an average queue time. And do you measure against that to determine, you know, is it performant enough for our user base or is it really a problem? Or is it just like, Hey, I would like it to be faster because, you know, as a developer we always wanna make our code better. I think that's a, just a natural progression of things. but I think of it as. I also make a distinction between bugs and what I call design defects. So like bugs are like, there's like a error page showing, right? Like it crashes the app and blocks flow. design defect is much more nuanced. It could be a bug, but it could also just be, Hey, this isn't working the way we thought it was gonna work with our customer base. or, you know, another flavor of design defect is. It actually is working the way we designed it, but it's not fitting the need of the customer, and because of that we have to. Rethink how we position this product. And to do that we have to rearchitect the domain model or change the database model or something. Or sometimes it's just as simple as like, I wrote this feature fast in an afternoon. It's not well written. It's hard to change. There's no tests. Like that's technical debt too, in a different way. But I would call that a refactor opportunity, and I would, you know, it's under the technical debt umbrella for me, but I try to make a distinction about the different types so that I can sort of rank the importance, right? Like a bug that breaks the app, completely for the user is more important to address, in most cases than a design defect or a refactor, assuming that the thing is working. But then if you're in the bucket of bugs, there's varying levels of priority there, right? Like I like to prioritize any bug that is gonna affect the initial perception of a user using our app, right? So if there's a bug during the signup process or shortly thereafter, that's much more important to me to fix than something that only happens once every six months. because I don't want the first experience they have. With us to be a block. Right. you know, so like technical debt is like, all of that is technical debt because at some point I made a choice to take a shortcut in order to ship something. And at the end of the day, you know, we, we can, that even applies to upgrades, right? You don't upgrade because you need to ship a feature. I call that technical debt,
Brendan:Yeah. Well, I think we, you just talked about there makes me not like the term even more because there are so many different variations of it and I, I would agree with you in general, like my example of I. Slow queries isn't necessarily technical debt. Now, when you've got customers complaining about the slow queries, I guess I still would lean towards it being technical debt.'cause it's not exactly a bug yet. long as the app's working, it just might be slow. so I guess it just, there's a lot of nuance to all this, right? Like that, you could look at it a lot of different ways.
Ryan:I have a bucket, you know, like along with bugs, design defects, refactors. I have a SIP column in my backlog for security infrastructure and performance and everything just sort of, right. And I try to pull out of each of those buckets something each week so that I'm always moving like the highest priority in the queue ahead. there's at least five performance things that cause me. Infrastructure headaches at any given time that I need to go and refactor the code to fix the performance problem, and I just haven't gotten to it because I got these other priorities. Right. So to that extent,
Brendan:Yep.
Ryan:it feels like technical debt because it's taking me away from building, you know, the vision of the products to handle something that is purely. Code related, right? Like it has no outward facing value because performance does affect the end user, but it is not perceived by most customers as a value add. It's like they, they view it as a baseline.
Brendan:Mm-hmm. Well, that, you don't mind, let's talk about your buckets there for a minute.'cause I think that gets at some of the stuff you talked about to begin with all the different options for technical debt. But it sounds to me, if I'm hearing you right, your backlog is several different columns and they each have a theme. And I think there's definitely something there that I like about it. I've never really that before. But I don't know, maybe there's not a whole lot to it, but it seems like that's an interesting way to break it up. At least
Ryan:Yeah, sure. I mean, we don't do anything fancy. We've tried every, project management tool, JIRA, pivotal Tracker. I've used Excel. I've tried a lot of different, Solutions in the past.
Brendan:I.
Ryan:and what I've settled on, at least for now, is Trello.'cause it's just sort of infinitely flexible and I can quickly move things around. It's still not perfect. but we're a small team, so it's good enough. and I basically just, you know, I'm looking at it now on the screen and I'll just sort of read through it right on the first few columns I have like new bugs, new design defects, new feature requests. We have a mobile app, so I have a special column just for request for that. And then I have a client commitment column. So if our sales or support team promises something to somebody, I give it, it's like, you know, high priority. and then we have a column for paid client dev, which I. Slightly different than a client client commitment.'cause there's the paid component, but it's essentially the same thing. anyone on my team can come in and add anything to any one of those columns at any time. Then I literally draw a line in the sand and I say, you shall not pass. Right? Like, and then from there on I've got my feature roadmaps. So like the priority features that I'm thinking about that I want to get to. and then I have areas of responsibility. this all actually maps back sort of like combination of, a way of organizing your digital life into projects, areas, resources, and archives. see, is it on my bookshelf? The para method. simplify, organize and master your digital life by Tiago Forte. I organize. Everything files. Dropbox Drive this board. My email inbox follows this method, and it was like a huge unlock, which could be a whole nother episode.
Brendan:Hmm.
Ryan:Areas of responsibility. So like from a product side, we have a product-led growth strategy, so that's one column. I've got Refactors, which is sort of a catchall for any of these things that you were just talking about that's like, Hey, I just need to reorganize this code because I know that I wanna build this thing, but until I. Fix the domain or database modeling. Then I actually split out security and infrastructure in one. So like anytime I have to upgrade my database, upgrade Heroku, upgrade rails, Ruby, any gems that goes sort of in security and infrastructure,'cause those are tightly coupled. Then I gotta call column for performance, which is all the, you know, slow endpoints that people complain about that I need to fix. you know, there's not too many, but there's enough where it warrants its own column. Then I have from then on out, I've got like three columns for bugs, bugs, high priority, medium, and low. And as things come into that first part of the board from the team, or from support or from whomever, I will move it into one of those, you know, columns. Sometimes I say no and I just archive it like straight away. I'm like, no, this is not something we're gonna build. and sometimes I say, yeah, this, this is an interesting idea. I don't know yet, but I'm gonna put it in low priority. We'll come back. in the context of bugs, low priority is like, Hey, this only really is a bug, like once every so often. It's not happening to a, you know, a important high paying client, right? Like we can manage the expectation, we'll get to it when we can, but I'm not gonna sweat it. three columns for design defects, same priority, three columns for any feature requests that come in. Same priority. and then I have a chores column for like, anything for that pretty much doesn't fit in one of those other things that has to get done. Then I have a column for product ideas. So anything that's not on the roadmap, but we might come back to at some point. And then I have a done column and that's pretty much my entire tracker. I use a lot of color coding and whoever posts the card gets tagged on it. and that way we can always just sort of, people can come in here, filter for their name, see all the cards they posted where I put them. Is, and so. Then I just make a habit of every morning, the first thing I do is I open up the boards. I use some tags too to be like, you know, critical bugs or things that I know I have to get to soon. So like, I'll come in every morning, filter the board down to that tag and be like, all right, what am I gonna work on today? Right? Like, that way it's outta my head. I don't have to think about it. Everything is on there, and I'm just very diligent. Every week I go through the error logs and Honeybadger. I go through the error logs and Sidekiq, or the other background job processors or any support tickets that have come in that week and I make sure there's a card for it. and sometimes it's a duplicate card and that's okay. Right. more people have reported it, more people are asking about it. They use that as a signal to help prioritize.
Brendan:Yeah, that's, it's interesting, it sounds really detailed and. It's a lot different than most of the systems I've dealt with, which I think the most intricate I've gotten with like a backlog is basically you've got your input or intake or whatever you wanna call it, column, and then you might have some variation of, we're never gonna do it, and you toss something there or you've got, you know, we're gonna do it, but we haven't prioritized it yet. it sounds like there's a lot to your backlog, but it also sounds like you have a pretty good process to keep up with it.
Ryan:The key is to, I think there's two keys, I try to keep up with new stuff coming in and always sort of comment back to whoever posted it to make sure that they know what I'm thinking, it's important for customer success to understand like, hey, yeah, this is a bug, but it's not a common bug. So like they can then if, you know, if I say it's a low priority bug, they can communicate that back to the customer. To make them feel heard, but manage expectations. And if it's a critical bug and I say it's a critical bug and I'm gonna fix it, they know where it is in the queue. and so that part's sort of important. The other unlock that I had was simply that, you know, outside of the things that I deem high priority or tag as critical, I don't worry about duplicate cards. I don't worry about trying to keep the entire backlog organized and succinct and limited to like, you know, the big ideas. I just put it in there and it's there for me to reference. And, you know, when I sit down, you know, whether it's quarterly, every six months randomly or at the end of the year, and I plan out like the big themes, then I take a deeper dive and, you know, look at the themes that are emerging. the, the one thing I did is when I. Do the intake. I use labels to sort of. Categorize it as like, you know what a lot of people would call like epics or sprints or, you know, whatever word you want to use. So like if I know there's a lot of stuff coming in around a particular feature, I'll make sure it's tagged with that particular feature so that when I'm ready to work on it, I can just filter the board down to that tag. Now I see everything bugs, design defects, feature ideas, chores, all related to that one thing. And I'm like, all right. It's that I say, I got a six week period to work on this chunk of work. What can I get done in six weeks? What are the priorities? Like, you know, I have to get this bug done. So let's start there. And then this has been like, you know. I have six cards about this particular feature. Right? Well that six cards is a lot, right?'cause that means six different people at minimum have asked about it. It was probably been raised on a call or two with the team. There's probably some support tickets to support it, and I can be like, all right, well this is like a no brainer. Like the themes emerge. Themselves, as long as you know where to look. But I don't sweat trying to keep everything organized in any particular order. I just say, Hey, if it's important enough for it to be worked on, it's gonna naturally rise to the top. And if it doesn't, then I still have it and I can, as I knock stuff off, I can come back and, get the next thing that rises to the top. And that has worked fairly well. For a relatively small team. I don't know how this would hold up, with a bigger team, but it has certainly been the most dialed in. I've been in a backlog. JIRA was a sort of a nightmare to manage, when I tried that just doesn't have the sort of same flexibility that I needed.
Brendan:I'll put it this way, I'm impressed you keep up with all that because I mean, I do have a bigger team, but at the same time, like I've had smaller teams and I've been a lone developer and I've never gotten it quite that organized in any time. So it's kudos to you for being able to do that.
Ryan:The para method is awesome. projects as Paris stands for, projects, areas, resources and Archives. And essentially the idea is for projects that you're actively working on, you should have a folder, you know, for that in your email, in your file system, and whatever things you use to organize your digital life. And anything that's related to it just goes in that folder so you know exactly where to look for it. And then when the project is over. You can just move that whole project folder to archives, right? You don't wanna delete it, you don't want it to cloud your, your mental space of around other current projects, but you may. Like the example, I think one of the examples from the book is like, oh, I'm planning a birthday party for my wife, right? So I have a project folder for that, and when the party's over, I put it in archives. And five years later when I want to throw her another party, I just pull that out and it has all my venue stuff. It has all the guest list, everything's there, and all of a sudden you jumpstart that process. And it works for business too. But that was just an easy example. And then areas are sort of ongoing things that you're responsible for. So in my life, I have a folder for my wife, a folder for my kids, a folder for any charitable stuff that I do, folder for coaching or any hobbies or things that I wanna, work on continuously or personal brand or whatever it may be. And then you have resources, and this is where everything you put for, you know, I found this blog article, right? Another place you can do this is in bookmarks, right? I found a blog article. I found it interesting. It's not particularly actionable on a project. I just wanna keep it around. I might need it to search for something later if it comes up. So, you know, the whole idea is you organize your digital life and you create, I don't know if you've heard the term second brain.
Brendan:Mm-hmm.
Ryan:accessible when you need it. this is just a way of organizing that information.
Brendan:Yeah, that makes sense. I certainly follow that, that last part there, the second brain, I dunno if I've ever called it that, but I know exactly what you mean. I think the same way all the time about. offloading things outta my brain that I don't either don't want to remember or don't want my brain to have to remember or just know. I never will remember it if I don't do it, is the way I think about it a lot. I try and do it for work stuff too. I have found, I use pen and paper a lot more. I've tried digital a lot and I've run into, I'm not even sure what the right answer is. Probably. That's part of the reason why I go back to pen and paper, but I find that just like having the, the. A piece of paper in front of me is just easier for my brain to remember, Hey, look at your last page essentially. and I try and do kinda like a to-do list sort of thing. I would love to get better at it digitally though, for sure.
Ryan:Well, it's interesting because. The second part of para, once you sort of have the structure that I just described, right? The idea is you put stuff in a queue as you come across it, and then once a week you're supposed to sit down and sort of organize it and synthesize the information down to like a summary I never got to that level. I still use pen and paper every day, especially for like product designs or as I'm working on, coding problems. I prefer the pen and paper. What I try to avoid is this long running list of to-dos, right? if it's a to-do that, I wanna remember it goes in the tracker and I have the equivalent of a tracker for my personal life that I call a planner, right? It follows the exact same system, but it's all for the personal stuff that I need to remember and I can just go in I mean, it's been evolving over years, right? Like I feel like I have it dialed in now and it's working for me, but like there, it was a constant struggle. But now I have a place that I can come in and, you know, within five to 10 minutes every morning I can see what the priorities are, write down my hit list of things to do. Usually I do it the night before, I actually write it down on a notepad, like a sticky note and that's my hit list for the day. I don't worry about anything else because I know as long as I'm diligent about putting it in there in the first place, it's there. I can come back, I can scan. and that's freeing just from the point of view that I don't have to keep it all in my head or rewrite my to-do list 500 million times, which is something I used to do, especially when they got long. I want all the to-dos in the order that I wanna do them. That day, day and every day that order changes 'cause of different differing priorities. And it was just like I spent more time writing to-do lists than I did actually doing work,
Brendan:Yeah, and I've been there too. what tends to happen to me is I will start with my goal is always to have just like two to-dos for the day, inevitably my sheet for that day turns into 20 to-dos, and so keeping them paired down is very difficult. I would agree with you. I feel the same way with like, there's two, to-dos. If I get those two done for the day and none of the other 20 things, then I still feel like I've accomplished something for that day. I think this kinda
Ryan:Right.
Brendan:of the stuff we talked, with John Nunemaker about just like organization and how do you feel like you're actually accomplishing something for the day. And yeah, I think I've got more work to do. As long as you're aiming in that direction and you're not just letting everything just overwhelm you and you're always just taking the next thing up and not paying attention to all the stuff that's maybe more important than that next thing. that's the goal you should aim for, at least to start.
Ryan:Yeah, I think that's right. And I also try to be intentional with my calendar, right? Like I have the buckets that, you know, we talked about that are in my, in my tracker, and I have time every week in my calendar to at least work on something from that bucket, I have a couple hours a week that are dedicated to doing the, rails upgrades, which is sort of in the security and infrastructure bucket. I have time every day to work on high priority bugs and design defects. I have time every day scheduled to work on the key sprint project that I'm working on, and then everything else I let sort of. Sprinkle down. You always see this on social media with the college professor that, puts the big rocks, in the jar and then put smaller rocks in the jar and then sand in the jar and they're like, it's full. And then they put water in the jar. It's like, that's how I think about my,
Brendan:Hmm.
Ryan:My priority is right. I put these blocks of time in my calendar. I know those, those are my big rocks. They need to be moved forward, right? Like blocking high priority bugs and my sprint feature one and two. Then I put some smaller rocks in there. my security infrastructure, performance stuff, my new greenfield features that I wanna work on. And then I let the admin sort of be the sand in the water, right? There's always, you know, support tickets. I'll pop in and do those, payroll, finances, KPIs, reporting, right? Like all that. filters in to fill in the gaps and still a work in progress from a, you know, execution standpoint. Some days I do really well with that process and other days go off the rails and that's fine. The next day I come back and I reset.
Brendan:Yeah. I think that resets the key part too.'cause I get the same thing, like I said, my list grows every day. But then I don't go back to that list the next day. I start fresh, you know, at least come up with the top priorities for that day. and for me, a lot of times, I don't write as much code as you do. And so for me it's mostly keeping the ball moving for my team. Do I need to, unlock people? Do I need to review things to get something deployed or, stuff like that. just thinking about the tech debt of it all and where we kinda started.
Ryan:Yeah,
Brendan:and I don't know, maybe you and I should brainstorm and come up with a different term for tech debt at some point,
Ryan:I've given up, right? I think in theory, that's a good idea. Let's try to be more specific to find the term. The truth is everyone uses it. It means something different to everyone else. I think it's okay to just let it be sort of a catchall. Anytime someone asks me about a technical debt question, we just have a conversation like this and I think that's fine, Like I don't even want to give it anymore. Brain power than that because I think it's just, I don't know if it'd be meaningful, does it matter? We can all have our own definition of it, and we all pretty much mean the same thing. I very rarely run into someone who's like, oh, I have this tech debt, and they say something. I was like, that's not tech. That, like, I very rarely have that reaction. What were you gonna say?
Brendan:That's a good point. No, you're good. You just made me think of something else though, which is I would agree with you. Like it's not really worth coming up with something else. I would a hundred percent am on board with like, why argue about it. Like kind of the generic thing is tech debt. And the same with like sprint, like, I don't like the word sprint really. But, and even like the, the tool we use right now for Kanban calls them iterations, but everyone knows the term sprint, so I don't fight it anymore. It's just, okay, we're just gonna do a sprint and that's fine.
Ryan:And that's totally okay because it basically means the same thing, You can take the parts that, I mean there, there's how many words for the same thing? Sprint iteration, you know, project Epic, right? Like, I kind of use those all interchangeably and I don't sweat the, I know there are, you know, in some frameworks there are technical differences in how people define'em, but I just, I don't really care.'cause everyone knows what you mean.
Brendan:Yeah, I don't care so much about that part, like the, the framework, I don't really care what it calls it. The, the thing with a sprint specifically that bothers me is just the idea that like. Your boss is saying, we're gonna be sprinting all the time. Like, when's the next sprint start? It never ends. Right? Like, that's not a sprint, but that's just me probably, you know, just doing, doing my thing. And maybe that's not how people are even thinking about it, but that's what bothers me about it anyways.
Ryan:Well, that's because you care about your employees. So that's a good thing, right? You don't wanna burn 'em out and have 'em feel like they're always go, go, go, right? Like developer happiness is a real thing, and if you can cultivate that, they'll be way more productive than if you try to maximize their output.
Brendan:I mean, is there anything else you want to add about tech debt? or just that, that concept of keeping up with the Legacy app and where we started?
Ryan:you know, I've been talking a lot, so I don't think I got anything else. You wanna have the last word here.
Brendan:Yeah, I don't have a whole lot else to add. I think it was a good conversation around, you know, it means. And I actually think that the other stuff we talked about was probably more interesting. But, you know, it, it does feel like this is a topic we could talk about again, and maybe we, you know, if we have a guest on, they could give another flavor and, just add some different color to it.
Ryan:let's definitely revisit it There's plenty more to chat about.
People on this episode
Podcasts we love
Check out these other fine podcasts recommended by us, not an algorithm.
Code with Jason
Jason Swett
Standing in the Fire
Very Good Software
Remote Ruby
Chris Oliver, Andrew Mason, David Hill
IndieRails
Jess Brown & Jeremy Smith
Startups For the Rest of Us
Rob Walling
Vibing with AI Code
Brendan Buckingham