Randy Johnston and Brian Tankersley speak with Sasha Orloff, CEO of Puzzle, about why artificial intelligence should be viewed as an accounting assistant rather than an autonomous replacement for accountants. Orloff argues that accounting is fundamentally an accountability profession: AI can prepare work, identify anomalies, execute firm-defined rules, and accelerate close processes, but a qualified human must review, approve, and remain responsible for the books.
The Accounting Tech Lab is an ongoing series that explores the intersection of public accounting and technology. Visit on YouTube: Accounting Technology Lab – YouTube
The conversation contrasts probabilistic generative AI with deterministic accounting workflows. Puzzle’s approach lets firms describe policies in natural language, confirm the system’s interpretation, and preserve those instructions as repeatable agents. Examples include automatically preparing depreciation for computer purchases and recognizing annual prepaid rent according to a fixed schedule. These agents preserve firm knowledge, create audit trails, and reduce repetitive clicking without silently changing records.
The speakers also explore accuracy, trust, staffing shortages, and the limitations of legacy accounting platforms. Orloff says firms adopting governed automation can close faster, improve accuracy, scale without proportional hiring, and potentially give staff significant time back each month. The central message is optimistic but disciplined: firms that combine AI speed with human judgment, transparent controls, and explicit sign-off can deliver better client service, higher margins, and more rewarding accounting work.
Transcript
(Note: Some errors may appear due to automated transcription.)
Brian F. Tankersley, CPA.CITP, CGMA 00:00
Welcome to the Accounting Technology Lab, brought to you by CPA Practice Advisor, with your hosts Randy Johnston and Brian Tankersley.
Randy Johnston 00:09
Welcome to the Accounting Technology Lab. I’m Randy Johnston with my co-host Brian Tankersley, and we are super pleased today to have as a guest Sasha Orloff, the CEO of Puzzle. Sasha, we’ve and Brian and I have met in the past, and you know our opportunity to talk with him today is really a pleasure. So, Sasha, what would you like the listeners to know about you in terms of background?
Sasha Orloff 00:33
All right, let’s start. I guess for anybody that didn’t make it to Scaling New Heights or a AICPA emerge, more puzzle we built. build accounting software for accounting firms, and I think one of the things that really kind of drive us are two parts. One, we think what I mentioned on stage there is we think AI is going to be one of the biggest booms for accounting, and accountants are going to make more money because of AI than ever before, and have a more important role. And two, I think accounting is getting more fun now than ever before because of AI. It is scary, but like, let’s talk a little bit about why I think it can also be fun and interesting and exciting again.
Randy Johnston 01:15
Yeah, makes beautiful sense. And what did you do before starting Puzzle, Sasha?
Sasha Orloff 01:20
My background has been mostly in like math and software, so I built fraud models and risk models. I built a lending business. I built a payments business. I worked for a nonprofit in Central America, helping sort of poor women in rural areas start and grow businesses to feed their families. I worked at the World Bank on their technology team, and sort of always had a sort of a passion for where software can help sort of make economies, industries, and communities stronger.
Randy Johnston 01:49
Yeah, super. Well, you know, our listeners who’ve been with us for a while know a lot of probably too many TMI, if you will, personal information about Brian and I. But because I have a computer science degree and a light physics degree, and first wrote AI on Lisp in 1975, so over 50 years ago. Yeah, been around the block once or twice. So I appreciate some of the neighborhoods that you’ve lived in, and see, I think your position and thinking that the future of accounting is neither autonomous nor manual, where AI can really handle the repetitive prep, and the accountant gets to make the judgment calls and then signs off. So nothing really reach the reaches the books without the humans deciding. I think the firms that will that will really win in the end will use AI to make their accountants more valuable, not really replace them. There’s so much talk like that. I think’s just false. So well, I
Brian F. Tankersley, CPA.CITP, CGMA 02:45
mean, it’s you know again, you can look back at the beginning of my career in the early ’90s at Coopers and Librarian with a five, seven, and 14 column paper. You know, so many times I don’t have it here in front of me now, but I still have the metal ruler that they gave me on august 24 1992 when I walked through the front door the first time, and you know they gave me that and a mechanical pencil and a bunch of paper, and off I went to work. And I had to buy my own 10 key even. And so I mention that to you here because I am so glad we are in the computerized world we’re in today, as opposed to the paper world that I was in, I walked into 30 something years ago, and I think the people that come after me in 30 years are going to go, “What do you mean we used to key stuff into software? You know, they’re going to look at that and they’re going to get there. It’s going to be like us watching Star Trek, where they talk to the computer, you know, and going, “Wow, that’ll be cool when that happens someday, you know. So it’s. I think you know. Just for perspective here, folks, we’ve been through these things for a while, and I think I’ve been told that five or 10 different technologies are going to eat my job, and accounting is not where you need to go. So many times over the last 30-five years. So I think Sasha’s right to ratchet back the fear here.
Randy Johnston 04:02
Yeah, so you know this idea that AI will make accountants more valuable and not obsolete because the person that’s in charge of the books will be in command of them. So it’s really the accountant, not the AI models, that’s the ultimate decider. So I think Sasha, there were three or four things that I heard you say at these various presentations over time, and one of those is just about fear. So you know, one of the questions that people seem to ask is, “Will AI replace accountants? And maybe the better question might be, “Who’s accountable when AI does the work? And that shift is really the whole conversation, and we’ve set that up a little bit. But talk to our listeners about what you’re thinking there.
Sasha Orloff 04:46
Yeah, the two narratives I think that are easy. I was actually thinking about this on my bike ride to the office before chatting with you guys, and yeah, I think it’s really, it’s accounting is just a complex industry. Everybody that’s an accountant. That and anybody that works with accounts knows that. Now, if you’re a press or a venture capitalist, it’s kind of easy to oversimplify and think about accounting as a categorization exercise because you’re not in the weeds. You’ve never done it before. You don’t understand. We’re trying to codify every type of potential money movement of our entire economy, with full data, with partial data, with no context of data, just as cash out of the bank account, and so it’s so easy to oversimplify this. So I think then it gets to two really key points: is this technology will be a step forward. It is a step forward in capabilities in software. When we went from we could put stuff on the internet to we could put stuff in the cloud to we have APIs and we could share data between stuff to now kind of some semblance of intelligence knowledge work. But two things come to play. One is who is accountable. Accounting is invariably in the name an accountability exercise and job. If you generate some marketing message and it doesn’t resonate, you maybe lose a client. If you generate inaccurate books because quote AI did so and you have some autonomous or self driving system, you know who is getting is IRS going to be like you know oh no it’s cool AI did it that’s okay don’t worry that it’s and
Brian F. Tankersley, CPA.CITP, CGMA 06:19
I will say that you’re exactly spot on there because the IRS actually released the new version of Circular 230, and they’ve said they they’ve said some troubling things that they think if you bill by the hour you should you should rebate the AI benefits back to the client, and they also said they also said exactly what you just said that whoever signs it is responsible, and I think that’s the important thing here because the world is not black and white. The world is gray, and we don’t have certainty on everything.
Sasha Orloff 06:49
No, and this gets AI can only just like machine learning and regression and before that, it’s only as good as the data inputs that you can provide. And so, you know, if it says cash out in the bank account, or your bank statement is a little blurry, or there is no invoice or receipt. If you just see a transaction, like what is that? Who the AI isn’t going to be any better than a human and not knowing what to do. The difference is an AI is going to sound really confident when it guesses the wrong way, and so it gets down to who is signing off on the books. And maybe if you’re a very simple sole prop content creator. You just have like a very side hustle business. You know it might be good enough, but if you are a partnership, you have small business, you have investors, you have debt, you have the IRS. Like somebody is going to care about the accuracy of the books, and to care, you need somebody to sign off. So I don’t think we’re ever in a place where, you know, any meaningful, sizeable small, medium, or large business is going to want to save 1020, 3050, 100 $200,000 to like hope that AI did it and sign off on liability. Like I don’t think that’s just that’s not a thing.
Randy Johnston 07:59
Yeah. So as I’m listening to you on that, a phrase I’m using this year, and I don’t like my phrase, is “good enough, right? Because so many people are trying to buy very specific solutions. In some cases, overkill as opposed to getting it good enough. And of course, Brian was our primary author on all the AI-powered ledgers, which you know we’ve known you through the years as you were evolving this platform as one of those AI-powered ledgers. But you know, in your case, I think the options that people are reflecting on is the old manual way, which was accurate but slow and didn’t scale, versus autonomous, which was fast, where the books could close themselves and so forth, but I think you’ve determined that there might be a good third path with the workflow that’s following the firm’s rules, and we did talk about that in another accounting technology lab agents, where we have the agents applying rules consistently and preparing the work, and then we have people reviewing and so forth. So help us understand what’s going on in your thinking on this.
Sasha Orloff 09:09
So it’s a good question. I think you let’s say let’s take the acceptance that this technology exists and it is getting better over time. It’s getting better because there’s now reasoning to explain some of the black box logic. But the technology is evolving, and that happens regardless. I think how you implement that technology has a lot of implications, and I think it depends on who your customer is. So, for example, our customer is the accounting firm, and what does that mean? It means we defer to them as the experts, and we are a tool to help them amplify and scale their judgment. That’s why I say I think accountants will make more money in the future. When we look back at just to get a very specific example, when we look back at like the two. Major automations that we’ve had in the accounting practice-it’s the bank feed. It’s getting data from the bank into the system where the general ledger is, so you don’t have to download and upload. And it’s the if-then statement: if this vendor, then this category. You know, there’s a lot more to accounting than that. And I think when you think about what are the capabilities of agents or a more complex automation, it can do if-then statements. Now it shouldn’t because we already have code for that, and it’s cheaper and faster and more deterministic. But you can actually build deterministic agents for accountants as the experts. And so, if we take the example just to bring this to life, agents or automations using artificial intelligence are can be probabilistic or deterministic, and the way that you make them deterministic is you need to rebuild and design the architecture for determinism. Now that that’s a bunch of buzzwords, so let me just kind of get a little more specific in here. When you type something in to our AI agent platform in Puzzle, it is the way that it’s designed is it will say if you opt into it. Of course, this is not required or optional because again, we view the accountant as the expert. It will say here is what you wrote. Here is what I think you said. I’ve looked at whatever transaction, fixed assets, depreciations, amortization schedules, reconciliation. This is what I think you said. Here is what I think the application of what you said is. Does this look right? And if you say you have to type in yes or confirm or some sort of positive affirmation, which will log into the audit log and save everything, it can then execute exactly what you want. You can schedule it so it can happen when you sleep or when you log in. But is your judgment preserved? Your intellectual property preserved? It just doesn’t require you to log in and do the clicks every time. And then the difference is because we sort of wrote little tiny software programs or functions for each individual step. What it does is it saves those exact steps in that exact order, and it can run the same thing every month deterministically, meaning the exact same way, the exact same way every single time. And that is really important for accounting. And so the two examples to bring this to life are, not surprisingly, we’re a software company, so we buy a lot of computers for engineers. So you would have a policy that says you buy this very expensive computer, you should depreciate it. Let’s just say it’s 30-six months straight line, no salvage value. Use the start date at the in-service date as the transaction date. You never have to do that ever again. It will just say if there’s a computer in this bank account or whatever level that you describe it, just depreciate it. Now you review it. It’ll draft everything. You can say just draft it. I want to hit send or automatically depreciate it, and I’ll just review. Or we did this a lot at Scaling New Heights. It was really fun. We just had crowds of people and we were just building agents in real time. There was another one that said, “Hey, listen. This client, this specific client, they prepay their rent annually. So every once in a while, I don’t know when it’s going to be, but this landlord XYZ assume this is a prepaid rent expense.
Sasha Orloff 13:16
If you see it, draft the prepaid rent expense, put on the balance sheet, and then expense the rent every single month according to this schedule, and you just save it. And the agent looks every single month or every day, however long you schedule it, and says, “Find this thing. If you do it, you have to do it. There’s all of these nuances between every individual client that you have to remember. You keep in your note sheet. You have to look individually. Well, like you can preserve that knowledge and intelligence, run it deterministically every single month, and not have to remember. There’s like you know 1000s of these things that we built a couple 100 at Scaling Heights, but this can now be codified and saved and preserved for that firm’s IP. Now, if that person who runs that client’s books isn’t you and they leave your firm preserves that intellectual property, so that client’s books aren’t at risk. There’s so many fun applications of this. That’s what we basically did for like three days straight at Sailor Heights. Did a crowd of people, and we’re like, “Tell me the most crazy, weird, nuanced thing. Let’s build an agent and just like show you how powerful this could be, but also how safe it can be. It’s never going to make changes. You have to type in yes, so you don’t log in and then wonder what did AI change? Did it change my chart of accounts? Did it reclassify historical transactions? None of that happens, at least in Puzzle, because that’s how we implement the technology for our clients.
Randy Johnston 14:35
Yeah, makes great sense. Well, you don’t know this about Brian and I, but I serve on the board of directors of a bank core processor, and Brian’s wife is a VP at Regions. So we have some ties to that banking industry, but also I installed some of the first ATM networks, and you know the banking system cold. And here in the U.S. it’s kind of a jigsaw puzzle compared to global. So you know you appreciate that. More because you’re doing it, but you know this conversation about agents is important because we believe that agents will be like smart staff that we’ve had historically, and you know on this you’ve I think demystified it to a degree. So where this agent is handling a defined task, where it categorizes and reconciling and does matching and flags anomalies. All those are pretty important, but also you’ve had the outcome surfacing that to a person so they can make a decision and say go. So you know that idea that it’s a fast junior, you know, staff manager senior that never really signs anything but to kind of preps things that seems to make pretty good sense. So when you think about agents, what do you think your best definition of an agent in plain English is? Because we teach so many people, I’m still working for what I consider an ideal token definition and an ideal agent definition, and so forth. So what do you think an agent is in plain English?
Sasha Orloff 16:01
An agent is a piece of software that executes a very specific task.
Randy Johnston 16:08
Yeah, and your deterministic piece here is actually fascinating because so much of AI is so flexible and probabilistic, as you would say it, and the world of accounting is a little more precise, unless it’s you know two plus two. What do you want it to be? But you know, in the big picture of things, that’s a nifty way to think about it. And the other thing that’s so popular, Sasha, to talk about right now is this human in the loop decision and discussion, because we want the review. Brian properly called out the IRS Circuit 230 information on that, but I’m listening to presenters talk about human in the loop, and I actually know they don’t know what they’re talking about when they’re talking about it. But that’s neither here nor there because everybody’s saying it, and to me, the real test is when and how the human acts. So we can review things, you know after the fact, but I think puzzles approach and your approach is trying to have it reviewed before it posts. If I’m getting your strategy right, and in fact, then the accountant becomes the responsible party. So you know, I think that also makes sense. And I just happen to have the Accrual Intentions book. You know, Brian got set in Ron Baker’s session, and they talked about that. And we are hopeful to have you know Alexis Kingsbury as a future guest, probably. Well, frankly, but in the big picture, that’s a very interesting approach for the human in review. So, what do you want? What do you think the listeners need to know about human in the loop?
Sasha Orloff 17:50
So, you know, let’s let’s. I’m as you’re talking. The thing that’s coming into my mind is if we accept and appreciate both the importance and the weight of the accuracy of accounting, and we abstract technology aside from that. What do we do to get to best practices in accuracy? You have somebody prepare, you have somebody review that is independent, and you kind of have somebody oversee it. And obviously, as you get bigger and bigger into public companies, you have even more reviews. You have preparers, and you have reviewers. You have controllers. You have CFOs. You have auditors. You have auditors of auditors. There’s a lot of best practices because it is so important. And the bigger and more complex your business gets, the more weight that exerts. So if we just think about how can technology accelerate the best practices, assuming the importance of accuracy and the weight of it at different stages? Agents can play, or task automation can play, a really important role here. And so you can create tasks and agents that are a preparer. You can also create tasks and agents that are reviewers, but you still, at all of this, want to have that level of control and sign off and accountability and auditability. And so, how we design it is the same thing. There’s a set of tasks that you create as your firm, your judgment. You just define it in natural language. We have built also a very complicated set of agents together that do some reviews to gap to variants to anomalies. Look at clearing accounts. A lot of things where it’s easy to kind of oversee, especially as businesses get bigger. That’s a more complex sort of agent orchestration. We just make it so you click a button that says review my books. It doesn’t change anything again, but it will list out between five and 20 different things. It’ll mark it high, medium, and low. It will describe here’s what I saw. Here’s why I think it’s worth review. Here’s why I think it’s high. Here’s some potential options to fix if you want, and then it spins up an agent to do this. So in the privacy of your own browser ad. An accountant, you can not only take your judgment, look at all of the different ways it’s easy to make some mistakes, and we’d say it tended to captures about nine out of 10 things that you would do in a review anyway. And then actually, sometimes it catches stuff that you know you’d forget because we’re humans and it’s hard to oversee lots of different clients at the same time. But still, at every single step, it is who is the human, who is the accountant that designed this agent or this task, and it will say, you know, AI on behalf of Sasha for this specific thing, and then you’ll have the reviews, and it’ll say who clicked the review and who did it and who made any changes. But at the end of all of this, it still needs somebody to write in yes confirm I am signing off that these books are drafted and finalized and so human the loop can get blurry and get important I think at the end though it is not it is who is accountable to the level of accuracy of the books and our software is just a mechanism to help you get to your desired level of accuracy as fast as possible. Some businesses might not care that much. Some care a lot, and some care to the weight of third-party audits.
Randy Johnston 21:09
Yeah, and that makes great sense. As you were talking about it, I just think about when people haven’t slept well enough because the kids were up the night before. You’re not feeling well. How easy it is to make mistakes, and I know at certain points in the day I’m better than other points in the day. So you know, you just think about all this review that’s going on. So you know, your
Sasha Orloff 21:31
what happens if your glasses are blurry, and then all of a sudden I can’t say like, “Hey, I’m sorry, I missed that transaction because I forgot my glass cleaner this morning, or like I took them off or they left in the other room. Like these things happen, and this can be designed to be a mechanism, yeah, not to replace you, but to make you the best version, to make your judgment executed to the highest degree of confidence, and that is what we see. That’s what we build. That’s what our firms tell us they like about what we build for them, and it is very possible.
Randy Johnston 21:57
Yeah, and you know, it reminds
Brian F. Tankersley, CPA.CITP, CGMA 21:59
so. So the analogy would be then the analogy would be then that instead of a virtual reality that is driven entirely by AI, it’s like an augmented reality where it works alongside us and helps and makes us better.
Sasha Orloff 22:13
As long as you can articulate what you want to happen in your moments of clarity between coffee, between sleepless nights, like it doesn’t. You don’t have to be in that moment. You can just when you have a good night’s sleep, you define your month close process as a set of agents. The things that you want to review, you can save them as a firm as templates. You can save them individually, and just let it prepare and help you become your best version. So in the moments of sleeplessness, in the moments of insomnia, in the moments you stayed up late, in the moments your kids were crying, to the moments of perfect clarity, like it can help you be the best version of yourself all the time. Yeah,
Randy Johnston 22:52
and that whole concept, which many people refer to as flow, you know, when you’ve really got that right, oh my, you can do some wonderful things. But for some of our younger listeners they won’t recognize this sighting, but the five-it’s like the $5 million man of accounting. You know, we just throw all these specialty things that are augment the skill so much. And I’ve actually been enjoying Neil deGrasse Tyson’s “You Know Take Me to Your Lead” book on aliens and the sightings that he’s done in that recent release. Well, you know the citing you had earlier of accuracy and trust, and domain fit. You know, you we talked about AI, which is probabilistic versus deterministic, and this accuracy thing still means that some of those entries are going to be wrong. So, you know what is protecting firms in most cases are the process around the AI. So again, the review, the traceability with the human approval, and you know it’s not some sort of a slider to try to get the right accuracy and so forth. But you know it’s also why we’ll call it third-party systems that are closed transactions. You know they’re going to do a different thing than I think that Puzzle might be able to do, and that trust really is the firm’s product that is here. So, just talk to us about where you see accuracy, trust, and domain thing.
Sasha Orloff 24:14
I think it gets down to the point that we started this conversation, which is who is accountable to the books, and if you’re using a system, or if you’ve implemented AI that is making changes without your knowledge, or without your design, or without your judgment, you’re actually creating more work and more stress for yourself. And so those are helpful in some contexts, but our sort of view is the accountant is the expert. You just need too many clicks to get that expertise implemented. And AI is a supplemental tool. It’s an augmenting tool, just like software, just like design, just like bank feeds, just like invoicing feeds and spend feeds, and etc. That helps you just. Your judgment in a better way. Now there is risks because if you don’t design an agent the right way, it can make changes. And I don’t say that to fearmonger. Like it’s better to play around with agents and accounting and AI, and just get comfortable with it so that you can figure out what you feel comfortable with. But we’ve taken the approach that everything is opt in, meaning you start in a clickable way, just like the legacy. You can opt in by transaction. You can opt in by balance sheet. You can opt in by reconciliation. It’s a transparent system in which you choose the level that you feel comfortable with. And what we see is some people just go all in, and the results that we’ve seen from firms are: we’ve closed all of our books on the fourth day of the month. We’ve seen results from firms saying we’re giving everybody the last week of every single month off until we get more clients. We’ve seen firms say I can scale three times the size without needing to hire more people, and it is hard to hire now than ever before. This is not us playing marketing. All we do is we sort of try and amplify the results of our firms, but it is possible now, and it is design. We design in a very, you know, what we think is the best way of safety. So, because we know not everybody is as AI pilled as I am, sitting in San Francisco, like using AI across every aspect of my personal life and my family life, that is okay. Like people, you need to. Everybody has their own comfort level, and and our system is designed to where you have to opt in and you choose your comfort level. But we generally see that after people start playing around with a little bit more, they’re like, “Can it do this? Can it do this? Can it do this? Can it do this? And we don’t promise it’s going to solve all of the things that you can possibly think of and do and all that uses, but it’s getting pretty close.
Randy Johnston 26:44
Yeah, understood. And you know, I’m as I’m listening to you think about that. We do try to help our listeners understand the rails and governance that need to be around AI from both a privacy and security perspective. So it sounds like you’ve got a good bead on that. Further, as I I’m just listening to it again old school, but in the old days of computer networking, Novell Network locked everything down, and you had to open it up. And unfortunately, Microsoft took the opposite of you, open everything up, and then try to close the Pandora’s box, and they’ve never really quite got the security Pandora’s box reclosed. So, in your point here on the labor, people routinely disagree with me, and I’m fine with them doing that. I’m thinking as these AI tools progress, that outsourcing will become less of a deal. Yes, we have talent shortage, but the fact of the matter is, we’re going to get so much leverage. I believe with AI-powered tools that as people move away from legacy accounting software platforms into the new AI-powered ledger, I think it’ll be far better to do accounting because you won’t have to do all that tedious work. A rule of thumb that we’ve used over and over again. So, listeners, I apologize for repeating. But if an accountant is keying data, I’ve got a broken process, and this really allows you to not be manipulating the data. You can actually look at the reporting and apply more judgment. And I hope that as things progress, we’ll see more advisory opportunities and do that advisory work, but that’s also a bit of a slow slog right now. So, what do you see in terms of this talent shortage, and maybe you know other things in the future?
Sasha Orloff 28:36
Well, I want to also like acknowledge that if the accounting industry and the accounting software industry is one of the first cloud softwares and one of the first softwares of all time, like so, when we look around at the a level of capabilities across almost every aspect of every other business, we’re seeing really rapid advancement of capabilities, and the accounting industry has been slower to adopt. Now, I think the root cause of this is not the fault of accounting or accountants. It’s just the fact that 30 years ago you would build software a very different way than you would build today, and so it is totally okay to have said, listen, I’ve been using the same software for 20 or 30 years. I know where to click to do everything. Of course, you know a lot of legacy players change their UX all the time. You know, but like I have a workaround for all the things, and I have comfort. And like I’m near retirement, or I just don’t want to change. That’s okay. What you’ll do to solve is you’ll hire people or you’ll outsource, and that’s a totally acceptable and reasonable solution. Now, I do think that the results that we are seeing from the firms that are leaning really into sort of both puzzle and the AI capabilities, as long as well as other people in the market, is they’re able to do. The same amount of work in an unfathomable amount of short time, but without sacrificing accuracy. In fact, they say that all of these tools actually make the books more accurate. They have less client questions about what about this and what about this, and they’re able to deliver a better client experience, faster results, and more insights along with that delivery, so you get happier customers that get the results faster, and you get, you know, maybe a week or two off of every single month of their current client base. You know, let’s not talk about tax time. That that’s there’s always anomaly there. So, like to be fair, like there’s different ways to solve the talent shortage or for an individual firm, not as an industry, but for the firm, and that is okay. But do I think in a year or two years that the majority of accounting preparation review is going to be clicking on a screen? No. AOL still existed for 10 or 20 years after they came out. It was just a different technology ago, and when you have the dial-up and you want to control the whole interface, but a lot of people felt comfortable for 20 years after Google and broadband and modems and T1s took over. It is okay. There’s a market for everybody, but I do think that those that lean in are going to make more money. There’s just no way to say you can work less at a higher margin with happier customers for a long time. Like if those are the two ultimate choices, it’s hard to you know that old model. It can’t be around forever. People will they’ll lower prices, which we hope they don’t. They can raise prices because they’re delivering a better value product. They’re taking off a week or two almost every month outside of tax season. Like every month, could you imagine having a week of vacation every month and making the same amount of money? Like this is a paradigm shift that is just starting to emerge this year. So it’s okay if you last year you were like, “I tried AI; it doesn’t work. Agents were not commercially available until January of this year. This is a brand new thing, and so it’s okay that like you’re not too late. You can tip your way and work your way around.
Randy Johnston 32:15
Yeah. So Sasha, saying that I’m going to pick on you here a little bit because obviously gray hair and the beard, and as it turns, yeah, it is a little, and you know, Brian and I have earned that too. But I was reflecting on the legacy software because I was the original designer, or assisted a lot of these SaaS-based products, and I hadn’t actually registered. It’s right on 30 years ago because we did that work in 9798 right? And you know, so QuickBooks Online and Sage Intacct and NetSuite, you know, I sat in the room when we were doing the design on that stuff using the tools of the day. You wouldn’t build products like that today. That’s just not the way it would be. So notice this and the concept, by the way, of the week off per month, I’ve lived that life where I took a week off every month, and people said, “How are you doing that? And so, you know, nowadays we’re trying to automate enough that Sunday can actually take about a week off a month. And you know, yes, we’re working probably harder at other times than we should, but that choice should be yours. You shouldn’t be a slave to your accounting software. So, Brian, I’m sorry I stepped on you there.
Brian F. Tankersley, CPA.CITP, CGMA 33:24
Oh no! Thank you for sharing your time with us. We’ll be back next Saturday with a new episode of the Technology Lab from CPA Practice Advisor. Have a great week.
= END =
.
Sign in to get access to this free resource, and all of our whitepapers and reports.
Download this content today!
Register Now Already registered? Click here to Log In