056 Is It Too Early For Accounting Firms To Jump On ChatGPT?

Welcome back to Old Timey Tuesdays, where we help people overcome old timey accounting isms. Today we're talking about. We're just going to have to wait and see on that one. It still feels too early. The old wait and see, the thing that kept people off the cloud for decades, the easy out when you don't want to have to go through that change. Let's talk about it. There's valid situations for it and there are a lot of invalid situations for it. Come on in. It's Old Timey Tuesday. So last couple Old Timey Tuesdays. Are we gonna call it Tuesday now instead of Tuesday? A couple weeks ago we talked timesheets. Last week we talked change management. This is for me, all the stuff that I needed to hear like 10 years ago to help me navigate an old timey firm and actually make a positive difference. My goal is for this not to be complaining about old timey accounting. There's enough of that happening out there in the world. Let's give folks who are in these situations some practical advice for how to move that firm forward just a little bit, make a positive difference and create a bunch of little change makers out there in the accounting verse. So cloud tech is a really good example of this whole, well, let's wait and see kind of mindset. Because, I mean, there's still people saying this about cloud tech. Do we really know that it is secure or should I just keep it on this thumb drive instead? Should I keep my work papers on paper instead? By the way, the whole thing going on with Trump and the we saw the banker boxes of documents in the last week, man, I just love. There's, there's so many jokes to be made for like a traditional accounting firm and the way that information is managed. There's. Maybe we steer clear of that, but cloud's a great example of, like, where people, I think, waited a really long time because there was this scary newness to it. And the easy thing to do was to say, I'm just gonna have to wait and see or I'm gonna wait for, you know, bigger governing body X to give me guidance on this thing. Sometimes that comes, sometimes that never comes. But at least you can put your finger on that as a way of keeping you from having to go through that change. And this conversation's getting especially timely and more important because of the rate of change right now because of just how much is changing and how fast things are changing. And it's a hard one, admittedly, the more people you have in a practice, the Harder it is to manage. And one of the, one of the things that really gets in the way here is anytime there is change like this, you have to navigate a lot of misinformation. Some of that misinformation is like intentional misinformation. Some of it is just like people not understanding how the underlying stuff works. But all of it creates this cloud of uncertainty, you know, like just the dialogue of, do we really know how this works? Like, all of the stuff like that just creates this era of uncertainty that creates inaction. And in the cloud migration days, like it was all the desktop people doing this, right? Like you had all these articles of like, of like just kind of sowing these seeds of like. No, yeah, no, you're right. We don't actually know. So we should do nothing. And there's so much of that happening in the AI space right now. I've said this before. Ultimately what matters for us is understanding how data retention works with these tools. The capability overhang or the abilities of these AIs to do things. You know, we're still learning, we're still through prompt engineering, almost like developing new skills with the same tools. That like, does not change my calculation for what sort of information can go into AI tools. That is just what the tool can do. The retention policies of this stuff, besides like the fly by night, you know, stuff that we shouldn't be using at all, the retention policies is straightforward, pretty clear cut in the same way that the cloud apps that we use are today. But there's just a really big kind of dialogue and narrative of uncertainty around this and that. That sort of uncertainty can absolutely take the excitement and the possibility and like the positive momentum out of things. I was at some talks at Engage last week where it's really easy to spend a half hour talking about cool, exciting possibilities. And then one person raises their hand with a question or something that's just incorrect. And just like that, it takes the air out of the room for, you know, something that is maybe just completely untrue. Or it's that, that story of, you know, I have, I have a friend that had this or that happen to them, or a client happened, this thing happened to us. It's just like this totally off the wall story. And these things, like, until we educate ourselves, these things will keep us from leaning into our really helpful change that will make ourselves and our teams more productive. So I all that is to say, like, be mindful of those off the wall stories, those anecdotes that usually just align with how somebody feels about a certain thing. Like those are the things that we connect with. If you're already feeling this way, you hear this story, like you're just going to cling to that. I think that's human nature. So stuff like, you know, Samsung employees posting source code into ChatGPT, yeah, that's bad. That's a thing that they shouldn't have done. But then Samsung goes, and as a rule, no employees are able to use ChatGPT at all. And to me, that's a great example of a missed opportunity because there wasn't a more explicit set of guidance there around the ways that you are and are not allowed to use that thing. People within organizations will misuse information, like without the proper controls in place. But until you make those rules, and like you say, here's what's okay and here's what's not okay, you're living in kind of this limbo land. But people will see this news story and they will cling to it as if the fact that Samsung doesn't use ChatGPT is a reason that your firm should not use ChatGPT. And I don't, I don't want to fixate on ChatGPT right now. Like, it is just a something that is very front of mind for all of us and like, okay, what's the responsible way of doing this? So when people are kind of navigating a change like this and you end up stuck in that let's wait and see sort of mindset, I think it's usually a symptom of there being a lack of like an explicit process for managing change and assessing the value of change. I think within an accounting firm, first and foremost, you have to look at change on the basis of merit. Like, is this actually a helpful change? Is this actually a meaningful change? And in most cases you don't know the answer to this question until there's like a degree of testing that happens. And it doesn't need to be anything big or ambitious. Like this ought to be small scale tests. And I think we actually in accounting firms are so focused on like, how do we get this stuff done? That's just on the other side of the windshield right now that we don't stop to take the opportunity to like run different types of tests to see what works and what doesn't work. You know, accounting practices, for example, you've got a monthly cycle that you can run tests on. And these tests are like investments in your future, like learning more about what works for your team and what doesn't work for your team, U.S. tax firms are on this annual cycle, and it's what makes tax firms feel so slow to change and all change feels so high stakes because it's as if you only have one time a year to implement change. And boy, you sure hope it works. Otherwise you're like a year behind the eight ball and you got to go back to the beginning, man. That's not the way to make change in a tax firm. Like, you have to think smaller bite size tests that you can like run in parallel so that if you do go and make like a full sail change, it's something that's already been validated, that you've already done before. But I can tell you in my experience, like, just from running a tax practice and most of the tax practices I talk to, whenever you make change, it feels like this big, super consequential risky thing. And you sure hope it works when we shouldn't really be making any change like that. Like, we have to find ways to do smaller scale testing that could be test sets of clients, like maybe subsets where you're trying different things. It could be literally burning hours of the day just running test projects that are not like things that you will actually bill for. I think accountants, we can be so ROI focused that the notion of like doing testing just for research's sake can feel like pure waste. When, like, oh my gosh, like, this person could be getting paid. We could be getting paid for this person's time, right? But what we fail to factor in is, is the outrageous cost of getting change. Wrong. Of all the training that goes into it to go down a certain path and then realize this isn't the right path, you change course and attacks practice. Oftentimes, before you know it, you've lost, like years of progress changing course. Hey. This episode is sponsored in part by firm360. You know,360 because it handles all aspects of your firm in a single, single platform. You don't want to hop around all these. All these different places to do different things. That's why it's 360. You think they've got. Hang on, sidebar. You think they've got like VR headset support. You see the new Apple Apple VR thing. Wonder what that would look like, right? Okay, let me tell you about Amanda Spivey. Amanda approached Firm 360 in the second half of 2020 because they needed a platform to streamline their processes. Enable and enable remote work. Remember that old chestnut, remote work? Oh, a lot of people having to rip that band Aid off coming into Covid before firm360, their team had to haul boxes of work documents between home and the office if they wanted to work remotely. That is disgusting. That is absolutely disgusting. But you know what kind of level with you that was happening. That was happening in my practice more recently ago than I would care to admit. But let me tell you, thanks to firm360, they had the opportunity to add more clients. First three months they were seeing improvements in team productivity, better transparency into what everybody was working on. And it was because they could pull all that stuff together into one place all in one place. 360 put those goggles on. That's not a thing yet. Maybe soon. If you're still hustling information in a whole bunch of different places, check out firm 360. Link in the video. Description this episode is sponsored in part by dark horse CPAs. If you're a regular listener of the show, you've heard me say that Dark Horse CPAS is a platform CPA firm about five times now. And I know that each time I've said that you're like what the H E double hockey sticks? Does that even mean? It's a community of accountants operating under the same brand, building their practices collaboratively together. Turnkey resources. You're not reinventing the wheel. All the stuff that solves for you and will get you through. You know the hard things about launching a firm, about not making the same mistakes everybody else is making. Dark Horse brand will help you attract the talent you need and the clients you want. Buddy their tech team. There's a tech team. Hang on. Tech Team Alert will ensure you have what you need to automate the mundane and augment your. Your advisory muscles. Advisory muscles? Do you have those? Dark Horse is the only CPA firm built exclusively for CPAs. Unlike a franchise, they will work with you one on one in group settings. Get you everything you need. You're no longer a lone ranger. There's a better way. There's a better way. You got a whole squad behind you now. To learn more about how Dark horse is helping CPAs build practices from scratch and scale existing practices within a community that helps each other continuously up level their game, head over to Abetterway cpa. So do you have a way to like test this stuff? Like do you have a testing methodology within your firm to be like, I'm pretty sure that this is better but what's a small like bite sized way that we could test this out? And there's a term called kaizen and I first heard about it when learning about Six Sigma. I don't know that it actually has anything to do with Six Sigma, but. But this concept stuck out to me big time when I was in graduate school. Cause I was like, yes, like accounting firms, like we get this wrong, we need to lean into this more. Like yesterday I plugged this into ChatGPT and just had a conversation about it and got a better understanding of it. Kaizen's a Japanese term that means change for the better or continuous improvement and involves making continuous incremental improvements to an existing process rather than making large radical changes. And I think in firms we generally just do the latter. We're like, oh, we're gonna make this big scary change to this new system and we think it'll work and we hope that it'll work. And honestly it's probably feeding that feedback loop of change being hard and scary inside of accounting firms. And your team gets exhausted by change because the only change that we're making are, are these big dramatic changes. And there occasionally will be times where you have those rip the band aid off moments where probably you got to change a bunch of stuff. But I think what we don't do enough is incremental improvements like find those small testing environments. So a few other things about Kaizen here kind of eight notes. ChatGPT provided 1. Empower employees. Kaizen is a bottom up approach that encourages all employees, regardless of their rank, to suggest improvements. I've worked in environments where it was safe to do that. In environments where it wasn't safe to do that, there's like this old Toyota Six Sigma story of how they like the first week or something they would have new hires. Where did I read this? Was this an emf? They would put new hires in like this red box or something at the end of the day and they couldn't leave until they had given like one or two bits of feedback for how something could be improved. And obviously there's a really scary thing for new people, but it also like broke that sort of, I don't know, respect or fear of not knowing what you don't know and that keeping you, you know, preventing you from giving feedback. And I've been at firms where it was like, buddy, that's not your job. Like your job is to like get back in front of the conveyor belt and like make more widgets. But I've also been in other firms where there were dedicated spaces for this and systems by which this information was captured and like even just the act of capturing and documenting those things I think shows your team like you do, assign a greater value to those inputs that you're providing you with. Second thing here for Kaizen, create a safe space for suggestions. I guess I kind of just got there, didn't I? So like what is, what is the actual place that you provide to let them do that? 3. Use small incremental changes Unlike some approaches that advocate for radical changes, Kaizen is all about small consistent improvements. These small changes can add up over time and create substantial improvements. And then accounting practice, I don't think that's too hard because you're on a monthly cycle in a tax practice where maybe you're on an annual cycle, maybe that feels hard because that cycle is so infrequent. I would say again, find a way to like manufacture a smaller scale test environment. Maybe that is running small tests on that annual cadence that would lead to bigger change if they were successful. Even better version of that is probably like what is, what is like your lab? Like what is your research lab in your, in your tax firm? Is it a set of like sample projects that people can do from A to B? Like how do you manufacture this live environment that you can use for testing? Again, there's a, there's a ROI part of accountants brains that makes that hard to. Hard to embrace, right? Because like you're just. Is this just all time that you're throwing away but it wasn't. What it doesn't take into account is the meandering path that your firm will take with these big ambitious like educated guests kind of changes. Like I've shared how I ran a firm through a practice management system change that in the end probably cost us three years of going in the wrong direction. And I did all the pre work I possibly could. Like I researched this stuff for like a year and this was maybe five or six years ago. And it was the wrong decision. And it was this big ambitious like huge change and we got it wrong and it cost years. And like I did a lot of work but ultimately we probably cut the wrong corners when it came to making that decision too quickly without involving enough people in the firm to help make that decision. And that that accountant's inclination to cut corners to like reduce the cost and think about ROI totally doesn't factor into account the cost of the wrong decision. Couple other things, standardized successful practices. Once an improvement has been tested and proven successful, it should be standardized across the or. So what does success look like for that test? Continuous training and then review and measure have regular reviews that should assess the effectiveness of these tests, quantitative metrics, yada yada. So I do think like a common thread I can see in my firm experience is oftentimes that let's wait and see, it's not quite there yet sort of approach is a product of not having a solid testing methodology. That testing methodology like that for you ultimately becomes like, where does the rubber meet the road? Is this worthwhile or not? This is how we validate that. The other side of the is it ready yet? Discussion is what do we understand about this tech right now? And what does my insurance carrier think of this? And when it comes to like the nerdy, like the more technical aspects surrounding tech of is this responsible or not? You have to have a partner on your team that can help answer this question for you. Whether that is an IT group, you know, that works with other businesses like yours, so they have a deep understanding specifically of what you do. Whether this is an internal expert, like that has to be somebody's job. And I'd worked in environments before where like the partner types, the decision makers, they held themselves out as, as ultimately knowing better than the subject matter experts. But like the right approach to that is you've got to have a partner that will like help you make that technical decision. Not partner from the traditional like, you know, equity partner standpoint. But you've got to have that subject matter expert either in your organization or, you know, maybe your contract IT group who can make that call for you and be able to say like, given the state of everything today, here's the right way to do this and here's the wrong way to do this. They will not all give you the same answer in the same way that all insurance carriers will also not give you the right answer. It's been, it's really interesting how the like the E and O market is developing around ChatGPT stuff right now. What I have seen some initial feedback from carriers has obviously been very risk averse and sort of like very basic level, like kind of leaning into AI tropes, a lot of which is untrue. So when you are working developing those tech partnerships of who are the people that are going to validate those decisions. When you are picking an insurance carrier, like you have to factor in like how much these groups like our subject matter experts, how much they will lean into innovation versus how much they will do just what we're talking about, let's wait and see. And so like how you position your firm, like you gotta be in alignment with those subject matter experts with your carrier and all of that to like ensure that everybody's operating on the same page. But all that is to say not everybody will be at the same place in a given time. You're going to have more legacy minded IT groups who are just going to be risk averse and be like, nope, can't use it. Because honestly the easy thing to do and like the risky thing to do in my mind is just sit back and wait until it becomes mainstream. And there are some changes where they're going to play out over a long enough period of time to where like you're probably going to be fine. Like did the people who did the people who waited 10 years to go cloud get hosed? Probably not. They probably threw away a whole bunch of time because you weren't able to lean into like cloud integrations and then you're on your back foot when you come into the age of cloud accounting and all that stuff. But I do think the price of not leaning into change has never been higher and traditional, like tech change. Like you probably run it through this testing approach and you're like, yeah, so you know, if your tax practice, should we use taxcaddy or should we use sure Prep? You know, you figure out a way to test that and that may get you an incremental advantage where you think, okay, this is actually going to save us like 10% of our time, you know, year over year, 25% or something like that. ChatGPT and language models right now I think are like a big enough change kind of on a different level to where you really have to be having this conversation and beginning to define what are the okay uses and the not okay uses. And ultimately the goal is to have a document within your practice to say here are the ways everybody's allowed to use it and here are the ways everybody's not allowed to use it. And even if that has to start at a relatively conservative place, that's better than no document. Because right now the world we're living in is we're in this situation where like if we provide no guidance to our team members, they are like the equivalent of the high schoolers going out and generating papers with chat GPT, right? Like laziness is kind of the ultimate catalyst for innovation. If there's this tool that I think will help me do my job better or help me do it in half the time, I am really motivated to become an expert in that tool. And until you put some sort of boundaries around how it's allowed to be used, you better believe people are using it in irresponsible. Ways. Right? Like, like the document that enables a certain type of use is the same document that will like ensure that when people are leaning into it, they're doing so responsibly. This episode is sponsored in part by Client Hub. This week on Tales from the Hub. That's right, it's back. Remember last week we were talking about super smart accounting firm and they were celebrating how everyone could stay on top of things from clients from Client Hub's jobs dashboard. That's right, I said celebrating. They were celebrating it. Now they're getting ready to head to conferences. It's conference season. I've been talking about conferences a lot lately. Right. Man, they can't wait to take their partnership with Client Hub to the next level by doing some irl, glad handing, back slapping and a lot of other analogies like that. I tell you what, it's one thing to chuck support requests into a system to email with that support team, you know, share your new ideas for features. It is a whole other thing to stand in a booth with the delightful people that you've been talking with and actually build, build some human relationships. Client Hub, co founded by a former firm owner. It's why they just seem to get it. It's why when you flip the switch on Client Hub you're like, wow, these people understand me. They understand my pains and they understand why all I want is for my client to respond to my request. Well now, super smart accounting firm, they can't wait to see Client Hub at scaling new heights in a couple of weeks. You're going to be there, I'm going to be there, Client Hub is going to be there. Let's all hang out. Even if you're a Canadian, they're going to be at cpp, CPB Ignite in September this year. I got no idea what that is, but if you're Canadian, you probably do. That's it for this week's episode of Tales from the Hub. Check out Client Hub. If you're going to be at these conferences, let's hang, swing by their booth. You can meet the actual people behind it. You can have the best, you know, meaning people, the people with the best intentions working for you. But I still always approached my job and as having to assume the worst, having to put them in a place where it was impossible for my team to accidentally do something that they're not supposed to do. And so I think with to bring this back to kind of old timey accounting and if you're in a situation that feels sort of legacy and how do you make a difference? I think, as is the case with so many of these conversations, it's easy to fixate on that thing you're struggling with right now, like stop using timesheets when it's like, no, the issue is the partner doesn't know how to value bill, so if you pull that away, they genuinely don't know how to bill. Or we gotta use CHAT GPT when like the underlying issue is like you actually don't maybe have the right people around your firm to determine and help you develop that document or what are the, of what are the right and the wrong ways to use it. And when we're in these situations, I think it's really important to, to figure out like what is the small win that you can enable for a stakeholder that will start like walking them down that path. Like you can have your, your maxims of what is the ultimate way to do X, Y, Z. The reality is human beings, before they can ever be there, like they've got to take 100 steps between where they are today to get to ultimately get there. Like that's just how we work. So what are some, some small wins that you could enable that will help start to open people's mind to a certain possibility that maybe you're really hot on? And just to use ChatGPT as an example, because this is something that is like front and center for us right now. What are some like easy wins that you could swipe and begin sharing with your team? Stuff that we've talked about. I've got three videos on ChatGPT now on the main YouTube channel. Stuff like using ChatGPT to look up unknown merchants when you're doing bookkeeping. That is something that like, I don't know that anybody's going to have a security concern with. That is actually a really big time saver for me. In my experience, if there was a type of change that would require some sort of investment like developing a document or pulling in a third party expert to validate, is this okay? I first had to validate like, no, this is useful today and this is the ways that this will save our team time. And so for me, in retrospect, if I could have given myself advice, it would have been start simple. Like you're thinking, you're thinking really big and ambitious on the long term impact of it. But that's not what's going to change the trajectory of a firm. What's going to like that first step is what is that really small bite sized win that you can enable that will begin to open people's mind to where you're already at. Right. Like you're already like 10 steps ahead. You're not going to like, shortcut them to get straight there. You got to enable something really simple just to get started. Hopefully that's helpful. If this is something you've had to navigate within a practice and not even specifically AI or Chat GPT or anything like that. Would love to hear like, what were maybe some positive examples of how you enabled that change. I'm, you know, there's a ton of people that listen to this show who are in more traditional firms, and I don't think the answer there is always go out and do your thing. Sometimes it is, and sometimes those people will in the future go out and do their own thing. But there's a lot of really positive change that you can drive. And like, ultimately making the profession better is going to be a combination of going out and creating this amazing, compelling firm, but also like intrapreneurship and like being that change maker within a firm. Like, I've been that guy before. It's hard sometimes. It feels really thankless. But that's also a really good way to impact the profession. Specifically on ChatGPT, I came away from engage really with this kind of on my mind. We need like, just more work needs to be put into people just getting on the same page about what's okay and what's not okay right now. Like, I have my own ideas of what's okay and what's not okay because I have an understanding of GDPR and HIPAA and like, what's the norm in other industries. But getting to it to more like practical. Here's an example document that you can use within your firm of the right and the wrong ways to use this right now. Like, we just got to get down to more practical, like, brass tacks, stuff like that that's going to take thought leadership from. I actually think it's a great opportunity for IT groups within our space to like, come out with some thought leadership and say, like, hey, here's the right and the wrong ways to use this right now. There's definitely wrong ways to use it, but those wrong ways to use it should not, like, throw out all the great, all the helpful ways of using it right? And ultimately, like, the best, most powerful use of Chat GPT will not be like the thought leaders and the nerds like me and the people like y' all who turn tune into, you know, daily shows like this because we kind of geek out on this stuff, it will be when the entire team can flip the switch on that and you've got 10 times as many people using and innovating and coming out, coming up with really cool ways to use this stuff. Like that is where we'll find the really cool ideas and really make an impact on how we do our work. That kind of the democratization of that technology, getting it in more people's hands. So if there's a big thing for the profession to work on right now, I think it is getting us from where we are today to a greater level of democratization and how we can use this stuff. I was disappointed at Engage actually how much fear mongering there is still happening. And I would say that's not actually coming from the aicpa. Like they've actually been pretty progressive with all this stuff. It's a lot of the people, it's a lot of the thought leaders in our space, honestly still like, don't have a good enough understanding of the technology to be comfortable with using it in specific ways. But we'll get there, but we'll get there and we're gonna do our part to kind of help push some people along that journey. Have you seen any examples of like firm policy documents for ChatGPT usage yet? I would love to see like if somebody's got a starter version of that or if there's a group that has put something out like a state society or an IT company. Like surely somebody's got to have something like this as a starting point. It doesn't mean I will agree with it necessarily. And I do think these documents often like lean really heavily, kind of on the conservative end of the spectrum. But we've had some really great privacy developments specifically around Chat GPT just in the last few months from, you know, countries like Italy, pushing back on like, hey, how are we ensuring that like kids aren't using this irresponsibly? And like, how do we get better transparency into what, into data retention and all that stuff. We've had some really good like steps kind of in how this has progressed in the last couple few months and I hope it'll keep going down that path. But when that happens, it actually opens up more possibility. It's like, here are the ways that I'm comfortable using it today, but as the like it all gets more transparent and we get more kind of tailored functionality in how we can use this tool. Like for example, Turning off in ChatGPT, you know, the conversation retention and turning off your prompts, training the model, which is now a thing that we can do. Like, this opens up new possibilities and how we can use it. And like, I'm of the mindset that I want to give my team all of the tools they can have to be successful, especially as we're coming into an age where productive output, like a big variable in productive output, is what are we capable of getting out of the language models that we can now use to support us? Because an expert with ChatGPT is capable of getting an entirely different thing out of ChatGPT than like a beginner. And so, like, as these language models become more a part of work, we got to start exercising that muscle and developing that skill of what are the. What are the types of tasks that we can just throw to a language model that will help us get that work done in more. A more efficient way in order to get the most out of those things? Like, we've got to start learning. Like, we've got to get from that place of not knowing anything about it to taking that first step. Because right now, like, that stuff, it's definitely a learned skill. So any ideas you have for how we can enable this for our teams, Any helpful documents you've seen in this regard when it comes to doing this within a firm, like, think small, like, what's the little itty bitty win that could be one little prompt that you create for your team and you share with them and you say, hey, we can use it this way now isn't this cool? This will save you some time. Like, what's that itty bitty win that you can get that you can get banked to just, like, start trending things in a more positive direction. That's it for today. We got Q and A Wednesday tomorrow. Got any queues? Check them in the comments and I'll see you tomorrow. Sa.