007 How To Manage Capacity

Welcome in my friends to the most narcissistically named show on the Internet. Today we're going to talk about capacity planning, how to do it, the Ultimate Framework trademark. Spoiler alert. There isn't one. You just, you just don't want to do the work. And so today I'm going to convince you why you need to do the work. If you're a tax person right now, I, I don't know that this is if this will be helpful or a kick in the gut because you're in the heat of things. Maybe more aspirational, some ideas for how to, how to think about capacity planning going forward, but that's it. Today, capacity planning stuff, little bit of AI nerdery. You know what it is. Jason Daly. Okay, today's topic comes to you courtesy of Alison Reef Martin. Have you considered a daily on capacity? Not enough. Too much. How to know when to hire other than when you hit the wall of not having capacity. Oh, that's a sad feeling. Boundaries. Is the game Monopoly better than the game Clue? I've never played Clue. Played a lot of Monopoly, never played Clue. Okay. Yeah, I mean I think most firms are operating, operating on the age old approach of they work really hard until they don't have enough people to do the work and then they hire the next person. And that's, that is capacity planning is just the level of angst and how far everything else is falling behind. There is a better way, man. Capacity planning is one of those things that everybody's like, oh, could you, could you show me your playbook for that? What do you, what do you use for capa? I haven't found the right app for capacity planning. And you know what it all is? You know what all that stuff is, is procrastination. Because if you sit down, you know exactly how to capacity plan. It's not hard. It's just a little bit of math that you don't want to do. Dump all that stuff in a spreadsheet. Like, that's okay. I'm getting ahead of myself. But when it comes to capacity planning, I've seen like talks and webinars and all these things and a lot of it is very specific to your firm. The way that you measure these things and ballpark it and how you plan ahead and that sort of thing. And I am 1000% guilty of this, as I am most of the things I tell people not to do online. It usually comes down to just doing the work of literally outlining every project and like putting together a plan for the Year. So let's start with that, like just acknowledging that capacity planning isn't this arcane sorcery that maybe you just haven't found the right course for yet. It is not making the time. It's usually just you haven't made the time yet to actually think through all that stuff and put something on the proverbial paper as a starting point. So like different approaches for how to measure this stuff depends on the data that you've got. This is one of the reasons why I always tracked time across a team. We didn't generally bill based on time, but we still tracked time. Right answer. There's very dependent on the size of your team. Like I was running a good sized team that was distributed everywhere, onshore, offshore and so it like I didn't have that sixth sense of how's everybody doing. But depending on the size of your team, if you've got time data, obviously that works. But you can also benchmark these projects on kind of different metrics like revenue or things that will generally be a reasonable yardstick for what you're trying to measure. Because at the end of the day, what's more important here is to think of like, what does the collective portfolio look like? What volume of work is, Is everything collectively more so than nitty gritty? Every single project, if you got the data for every single project, great. But the main goal is to have something, not nothing, and then go from there. And that's. That was like, that was like my second maxim. And what always tripped me up and kept me from doing capacity planning is it's so easy to pick apart any plan, like any way of measuring it. Like they're all fundamentally imperfect. I just it and it takes time and there isn't a great way to like track those. I know more and more project management systems are giving you ways to do capacity planning within the tool on like a full like year basis. Like perspective how what am I going to be looking like in six months and seven months and being able to have like variable drivers of what that could look like. I've never really got a system to work for me beyond just literally hustling that stuff in Excel. As much as I don't like it, as much as I would love to say, oh, here's a, here's a way to do an airtable or a cool, you know, fancy way to do it this way or that way, for me it was just a lot of work. And even if you're thinking about how to deploy tech to make that easier for you I still think the first step is like this self discovery process of here's what works and doesn't work for our firm. You're absolutely not going to get it right on the first try, but on the second try you're going to know a lot more than you did the first try. And so it's something that gets better over time. And it's really not until I think you've done that discovery that you even know what you need out of a tool to help you with capacity planning. If you, like, wildly disagree with me and you think there is some ultimate framework for capacity planning, I would love to be proven wrong. I would love to be able to take that pill and be like, oh no, this is the way to do it. I've just sat in a lot of those things and been like, yeah, no, there's no way around me just chucking all this stuff into Excel as my starting point. So really try not to get hung up on the notion that this system's going to be imperfect is a reason not to do it. That is where I always got stuck on capacity planning. Frankly, that's just where I've got stuck on a whole lot of things in my life like, oh, I can't fully automate the last mile of this thing. I guess we'll just have no system for it then and it'll still be reliant upon me doing it. The fact that you're never going to get to something that's perfect is not a reason to not do it. Because some plan is better than no plan. Just like some process is better than no process. Right? So you got to, I guess, start investing in that and it will get better over time. And maybe the first version of it isn't mega helpful, but to get to the fifth version that is really helpful, you gotta start with the first version in terms of who's the right person to do that job. Depending on the size of your firm, the makeup of your team, if you're pretty much anything beyond a solo shop, that is something that a member of your team, even if they're like a non pro, non accounting or tax pro type of person, can absolutely help with. All it takes is being organized and being able to hustle the data that's in all of your different systems to put something like that together. Because the other, the other thing that this can kind of kick off is then who is overseeing accountability and, and as you do that capacity planning, who's ensuring that you're still on track and like raising their hand if Things are going off the track and maybe you start missing targets. That's an area in my experience, where a really good admin can be invaluable. And so if you pull them in early on, the process of how you do this capacity planning so that they're in that from day one, they're going to be much more helpful longer term when they're able to see how that stuff is progressing, making sure everything's on target, being the ones to raise their hand and bring it to your attention when things aren't on target, as much as it's. It's just like anything else where it's really hard if you're running a team, not to stick your own nose into it and be like, I know better than everybody else, so I have to do it myself. Which is where everything goes to die. Because inevitably you end up having to do everything yourself, which means you do a bad job on all of that stuff because you can't do it all if you are running a team. Like this is absolutely somewhere that someone can help you with what is honestly like among your biggest stressors, right. Is when is, when do we hire next? Not only. And we say that as if, okay, I'm gonna hire on June 1st, and then on June 1st, we'll find the right person. When in reality, who knows how long that hiring search could go on? Who knows if that person could fizzle out a couple months into hiring them. And you're clear back to square one. Capacity is, in my mind, one of the biggest stressors of running a firm. And so if it's something that you can delegate to a team member, that is gold. And then my third thought on all this capacity stuff is so much of this conversation happens through the lens of we're already over capacity. What do we do now? I. It's so common. And I've just. This kind of dawned on me maybe six months ago, and now I see it everywhere. We talk about, you know, I've, like, I've talked a lot about how, like, the reason that you're too busy is because you let too much stuff in the door. You didn't say no enough times. You let too much stuff in the door. Maybe you didn't cut enough loose to kind of get out of those bad projects and into better projects. And the thing people always say when I, when I bring that up, because it is a little bit uncomfortable because it means it's fundamentally your fault. Like, if you're over capacity, you let too much stuff in. They say, you know, we don't want to let that work, cut that work loose because we just, we aren't as efficient as we once were or we've got some great workflow stuff that we can do. And oftentimes capacity issues are blamed on workflow. And workflow is the solution for your capacity issues. When I would say the fundamental issue with your capacity is that you took on too much capacity before the workflow could enable that thing. Like, you cannot. I think it's very risky to set up your future capacity and take things in and a plan according to some yet unreached level of productivity. When in reality, when you hire people, what usually happens is you get less productive because those people don't know squat. And you got to get them up to speed and they take the time of their peers to get them up to speed. But, like, what I keep coming back to there is I think we chronically underestimate how much capacity we need. So we're constantly trying to claw ourselves out of this, like, debt, this capacity debt with like, I know we're just, we're gonna automate some more stuff and that'll solve the problem oftentimes because it's a lot easier to talk about workflow and automation than like, ooh, I gotta talk to this client and fire them, or acknowledge that this is my fault because of all those clients that I picked up during busy season, that sort of thing. All that is to say, for goodness sake, stop taking in clients until you have a clear plan for the capacity for how you will serve those clients. And there's some sneaky sources of, of, I guess, like capacity not coming in, maybe where you'd expect a big one is on a whole, if you're, say you're working with a hundred clients, and this is especially relevant if you're working with business clients in general across your whole client base, that client base, those clients will generally be growing and their needs will become more complex and the scope of what you do for them will change over time. And honestly, in my experience, this was a huge source of not always revenue growth, but growth in the amount of work that we had to do. So like, my client base may grow on Average, top line, 20% year over year. And if that ought to equate to a 10% growth in, you know, our fees across a team of 40 people, like, that's 10% growth. Like, what is that? Another four people? Then I think we, we chronically underestimate the growth that will happen within our current client base and the, like, their growing needs. So So I think that can be a really sneaky source of future demands being put on your staff. The other one is just the new clients you're letting in the door. If you sit down on January 1st and you just make the the most dope capacity planning Excel spreadsheet you've ever seen that plans out the next 12 months, here's the people we have today, here's the clients we have on the list today. Like the organic growth of the clients that you already have, combined with the clients that you're going to add, obviously that's a lot more work. So the like the capacity you have on January 1st, what's required there is going to be wildly different than six months down the line really without much change, like largely holding station with the clients you have, picking up a few new clients here and there. To say you have adequate capacity on January 1st does not mean you're going to have adequate capacity on July 1st. And that's fundamentally the value of capacity planning. The other big sneaky, sneaky shift here that can be hard to predict ahead of time is the old risk of staff leaving. And nothing keeps you up at night quite as much as that. I mean it's honestly the very worst part of running a firm is team members leaving or stuff happening, like life circumstances happening with them and your job is no longer the right job for them through no fault of your own. And like that just happens. It's really hard to manage in a small firm. I talked about this in a video a few weeks ago. If you're a solo firm runner and you've got two or three staff and you bring on work according to the capacity that they enable, somebody leaves, well, shoot, there goes a third of your staff or a third of your output capacity overnight, right? So one thing I talk a lot about is diversifying the way that you get the work done across traditional hiring, across contract hiring, across contract groups from whom you buy tax returns or reviews of tax returns or month end closes, the more you can diversify. Especially in a small firm setup, if you can diversify across those different ways of getting the work done, then you don't become too exposed to any one of them and then say a staff person leaves, you can flex some of that work to the other people. So like big time key man risk when it comes to having a really small firm and being reliant on just those couple hires. When you get to being a bigger firm, it's a little bit easier because you have more people to shift that work around on. But there's no way around it. You lose a person and they're part of a team of five or something like that, it's going to put a lot of strain on the other people in that team. All of that is to say, in your capacity planning. I think what we generally do is we plan for the absolute best case scenario where we're going to be super productive, where we don't add any more work, because I think we kind of underestimate how much the existing projects will grow. And in these future plans, people aren't getting sick and staff aren't leaving and all that reality is you got to build a pretty healthy buffer in there. And if none of that stuff happens, is it that, like, is it really the end of the world if your team gets to take a breath? Like, is it really the end of the world if somebody doesn't have something to do? Like, there's so many projects that we can get staff plugged into that are not productive doing of the work types of projects. So if you do end up with excess capacity, that's totally fine and is better than the opposite. Because I actually think being over capacity is like fundamentally the largest driver of people getting run out of this profession is they're just exhausted and then their team member leaves and that stuff gets dumped on them and they're like, man, I wasn't having fun before my teammate left, now I got to do their work too. And then they leave. And that just like kicks off the nightmare, like cascading staff leaving sort of thing that I just don't want to think about anymore. So capacity planning, here's my hot take. It isn't hard. Like, of all the things we do, it really isn't difficult. We just don't take the time to do it and we get too caught up on the fact that it's going to be imperfect because there's genuinely no way to do it perfectly. But. But as with anything in life, you got to start with that really crappy version that ends up being bad and wrong and embarrassing to then have that much better version down the road. So my advice would be just start grinding through it, involve your team in it. Because there's a lot of, a lot of mental energy that goes into pulling all that data together from different places. And then the sort of methodology you use for measuring that and how you change course on it. If you only do it yourself from day one, then you're going to have to like train that next person on, well, here's all the things that we already Tried and did work and didn't work and all that. So if you have a team, try to involve your team in it and day one and just get started. If you're tax bro, don't do it right now. It'd be a good thing to look at after tax season, especially as the demands on your team generally after the tax deadline will drop off quite a bit. But I talk about how like the only way to run a firm in a calm way these days is for things to have schedules for the work to be scheduled out, for you to have clear expectations of your clients on when they need to get you things and then when you will deliver on those things. And all that stuff comes back to capacity planning and ensuring that you've got enough people to do the work at the time that you commit yourself to doing the work. And it's absolutely worth doing, especially if you can delegate it. Because for most of us it's like the biggest stressor. Right? Other thing I wanted to talk about, man, so on Monday, over the weekend, I like stumbled into something really cool with GPT. Let's talk about AI. And I called back to an old Twitter thread Chad Davis did, which blew my mind back in the early days, this was like maybe late last year. And he's like, hey, so in his classic Chad Davis, like underspoken way, hey, so I just figured out how to extract transactions from all these bank statements that, that like we weren't able to get bank feeds for with AI. And he is like this little thread and I'm like, excuse me. Well, the way he did it has gotten even more simplified now with better versions of GPT and now with ChatGPT. But basically on Monday, I ran through a way to extract transaction data from bank statements. So like, let's say you can get your hands on a statement, but you can't get your hands on a download file. Maybe you're doing tax prep and you do once a year accounting cleanup for that client because they don't maintain their books and you gotta pull all that stuff off the bank statement Somehow. GPT does phenomenally well with you literally copying and pasting the text of a bank statement into ChatGPT and saying, hey, give me a CSV of all the transaction activity with the following columns, you know, date, description, amount, and doggone it like it does it really darn well. Not even GPT4, which right now is paywalled behind $20 a month GPT plus, but GPT 3.5, which is what you get on the free version of ChatGPT, like in my testing did it without fault. So whereas before for me, in practice, this was, this was, this looked like going out to auto entry, actually we would use auto Entry for this and we would dump all the bank statement files into auto entry and like three or four hours later we would get back the transaction downloads for that. And they were generally pretty good. I don't know that we ever had any issues with accuracy. But compare that to literally logging into a free Chat GPT account, selecting and copying the text from a bank statement, pasting it into Chat GPT and you only need to grab the part with the transaction. So like if there's account numbers, you know, other stuff you don't want it to see, just don't select that stuff. You paste that stuff into GPT. And in all of my testing, it gave me a flawless CSV back in return, which kind of blows my mind because like Auto Entry was acquired by Sage, what was it, three years ago now? And I went back and did the math in terms of like what the cost is with auto entry because auto entry is, I don't know, I guess it depends on how you value that function. But effectively what I worked out with Auto Entry's pricing today is if you chucked so the bank, the credit card statement I was using to test this with was eight pages. If you chucked the entire bank statement into auto entry, like my staff always did, all pages, it would cost $5.76 to have that statement processed by auto entry. If you only chucked in the two pages from the statement that had the actual transactions on it, which is never what my staff did, it would cost you a dollar and 44 cents through auto entry. And I kind of think we like that sort of displaces what I in my mind was the core really cool feature of Auto entry because I think you can effectively get to the same place with Chat GPT, which is now free. Now it's not smart enough to do anything like validating that, you know, the total of what it pulled was correct. So for example, a higher quality version of that would go back and grab the starting balance and the ending balance and make sure that all the amounts that it gave you like foot to that correct ending balance. Another limitation right now is GPT has quote unquote context limits. That is it can only handle so much text at a time. So if you have a humongous statement, you can't chuck all of it at it at once. But in what I posted online, I was like in My testing, if you copy paste 50 transactions in at a time, it handles it totally fine. But I'm kind of amazed, like every day at these, what do they call it, capability overhangs or something like that, with these capabilities of AI that like, we haven't even yet discovered that sort of change our workflows. Taylor Hartman tweeted this morning, which I'm pretty sure is a subtweet to what I shared. It's crazy how many auxiliary apps AI has already essentially replaced. And it really is like, it's, it's, it's wild. The thing that blows my mind that I keep coming back to, that I've talked about before, is every new language model, like all of the, the testing people are doing of different types of prompt engineering. Like, nowhere is there a feature list. Like, there is, like, carbon releases an update. Here's a feature list of, like, here's all the new things you added. You don't get that with language models. You don't get that with the AI tools that we're using now. And it's like the wild west of like discovering, ooh, what can we, what can we do with this and that? Which is really, like, the exciting and, you know, especially for what I do, it's like, oh, I just uncovered this cool new thing. But it's also, I don't think, very accessible for people who are not the folks that geek on this stuff. I think we're used to, like, the product paradigm of every. Everything that tools do for us is within a product that I log into and I pay for and this and that. And AI is developing in a much more incremental way. And I think that's probably the new normal because AI develops in a more incremental way. So when GPT4 gets slightly better, or like when GPT3 got an upgrade that made it better at math. The things that it were capable of the day before that and the day after that were completely different. So I think this is actually a real blocker of people leveraging AI as much as they can today is. It's hard to know what it will do and what it won't do, what it will, like, kind of sort of do, but not always accurately. It's just a fundamentally different, like, software paradigm that I think we're going into. But that was super cool. I'm super jazzed on that, actually. Would have saved me a whole bunch of money on bank statement extractions. I'll put a link to how that works in the show notes. Any cool stuff you've stumbled into. AI wise. Any fun stuff you want to cover in this show, Drop a little comment. Tweet me. LinkedIn me as always. Love y'. All. Thanks for hanging.