Earmark Podcast | Earn Free Accounting CPE

Attention: This is a machine-generated transcript. As such, there may be spelling, grammar, and accuracy errors throughout. Thank you for your understanding!

David Jani: [00:00:00] Half of the people that we spoke to, 52% specifically have experienced a data breach at some point in recent times. So the last two years or beyond that, basically. And while I don't have any specific information on whether AI is the cause behind that, what we do see is a bit of a worrying habit, which is that most companies don't have clear guidelines on how their A, how their AI tools are used with sensitive data.

Blake Oliver: [00:00:37] Are you an accountant with a continuing education requirement? You can earn free Nasba approved CPE. For listening to this episode, just visit earmarked.app in your web browser, take a short quiz and get your certificate. Welcome back to the Earmark Podcast. I'm Blake Oliver Capterra just dropped their 2026 Accounting Software Trends report on how AI is changing accounting, work skills and strategy. More than half of accountants are now using AI in their software, and 89% say it's delivering ROI. But nearly half are checking every output, and a third are catching errors at least half the time. My guest today is David Janey, senior content analyst at Capterra and the author of the report. David, welcome to the show.

David Jani: [00:01:24] Hi, Blake. Thanks for having me on.

Blake Oliver: [00:01:27] David, before we dig in, tell us a little bit about your role at Capterra and the kind of research you do.

David Jani: [00:01:34] Well, okay, so I'm a senior analyst at Capterra, and the research mainly focuses on software applications and software uses. So I mostly spend my time looking at accounting software. Um, and for this, we look at how software is used in businesses, a selection of software and the sorts of decisions that businesses need to make and the questions that they might ask when selecting software, basically.

Blake Oliver: [00:02:00] So tell us about this software trend survey. Who did you talk to and what were you trying to learn from this report?

David Jani: [00:02:08] So what we did was we asked 500 accounting managers in the US about software that their companies are buying or that they're using. We looked at elements like the benefits and challenges of adopting AI tools, and also how companies are managing operations and modern challenges in the current moment.

Blake Oliver: [00:02:29] So was there one single finding that surprised you the most? What did you learn that if I if you had to pick one thing you would would share with me?

David Jani: [00:02:39] I'll have to say there's a couple I would want to pick, but if I have to just pick one, I think it's going to be about the staffing really, because I think despite a lot of, you know, reports say predicting the end of accountants, it's not really what we found. We found that automation is present in accounting work at the moment, and companies are adopting AI into their tools, and a lot of accountants are using them. It's not necessarily affecting hiring decisions in the same way. There's a lot of companies that are actually more focused on upskilling than they are on elements such as automation through AI to replace employees.

Blake Oliver: [00:03:18] That's interesting that you say that, because I have felt this way for a long time following AI and automation, starting with cloud computing in the in the shift that happened with cloud in accounting, we didn't see jobs drop. We saw them grow. We saw more and more bookkeeping and accounting work being done by accounting firms and client accounting services. Cast has been growing year over year for ten years now. And I think last year it was like 17% in accounting firms in the US. So we also saw this, like if we go way back to the invention of the electronic spreadsheet, Visicalc, Microsoft Excel, there are predictions back then that accounting jobs would disappear now that spreadsheets can do the work. But here we are, and we have more work than ever.

David Jani: [00:04:12] It does seem to be the case. Yes. I think in some ways, you know, everything old is new again, in that sort of in the trend in terms of the trend at least.

Blake Oliver: [00:04:21] So I want to come back to that staffing question because that's a big one for everyone. But first, let's talk about AI adoption. Right? Where are we today? How widespread is it in accounting right now? Are we still in the early adopter territory or are we now more in the mainstream? It's been about a few years since GPT 3.5 came out. So I'm curious to get your take. What is the survey data telling you about that in terms of adoption?

David Jani: [00:04:53] Well, this was really interesting because I think we've gone beyond the point of it being, you know, companies testing the water with this stuff. It does seem that accounting software is commonly used with AI in it. So we saw more than half of the accountants that we surveyed saying that they use it in their accounting software now.

Blake Oliver: [00:05:11] More than half using it in their accounting software. Now, I guess that that makes sense, right? Because we are seeing, uh, QuickBooks, we're seeing Xero, NetSuite, Sage, all these different accounting gles add ons, right? Different types of apps that we plug into the Gles. They're all adding in AI, AI agents. Um, there's this big software stack now in accounting, we're using dozens of apps, often with each client perhaps, or internally. I'm curious to know where in the stack is AI showing up the most. What categories of tools are leading on AI?

David Jani: [00:05:56] Just to clarify, what do you do? You mean like as in terms of features or in terms of like the actual types of software?

Blake Oliver: [00:06:02] Um, I guess I'm thinking like in terms of, yeah, like the areas of the software types of software, like when we say AI in accounting and we're using it in our software, what does that mean?

David Jani: [00:06:13] Well, it seems to be showing up a lot in areas such as, um, how the users interact with it. So for example, the most common use of AI in accounting was through chatbots or assistants for the most part. There's also data entry, which I don't think came as much of a surprise, because I thought that would probably be where it would end up being before the survey, and I think that was proven true. So that's about like half of the it's just under half in this case, but a lot of it's coming through fraud and risk detection, which I think was quite interesting for this in this case.

Blake Oliver: [00:06:48] So how does AI adoption break down by company size? Are our small firms keeping up? Is this mostly mid-market? Is this mostly enterprise? Where is it the strongest?

David Jani: [00:07:02] It's kind of strong across all different types of businesses. That was the interesting finding from this. It shows what we found was that most smaller companies aren't adopting it at the same rate as bigger companies, which is pretty much where you would expect things to be. But smaller companies are still finding are still adding it to their stacks. It's still there. They're a bit more tentative, and the ways they're applying it tend to be a little bit more cautious, I would say, compared to bigger companies, which are much, you know, using it for much more sophisticated situations. But for smaller businesses, it's still there. It's still common.

Blake Oliver: [00:07:39] So you mentioned chatbots. It seems like when I talk to accountants, pretty much everybody is using chat bots. 8,090% or more. We're using it for research analysis when we talk about AI enabled software. Is that what we're talking about? Are our firms choosing software that is specifically AI enabled, or is this adoption happening because the AI is getting bundled into the tools they already are using?

David Jani: [00:08:10] Well, we've got a bit of a split in this situation, and it seems to be a bit of both. Is the simplest answer to that now. It's most common that businesses are paying for it, so they're making a conscious decision to adopt it. So that's happening with over half, but really just a few percentage points fewer. And you get a similar and you find that people adding it as a as a free add on. So let's just use the comparative numbers. There's 52% will do it as a paid add on and Uh, 8% will have it just added by the vendor.

Blake Oliver: [00:08:49] So it's a mix of both.

David Jani: [00:08:50] It's about it's about equal, but with a slight preference towards the paid option. However, there is another section of users who are also adapting it by asking for it specifically. So asking for a. There they're changing their software specifically or choosing the software for AI features specifically. So in some cases it's juices, juices, switches. But I'd say that the preference tends to be more for paid, but only just. It's very close.

Blake Oliver: [00:09:26] So let's talk now about where AI is delivering the value, because I think that's what we want to know as accountants is, is where can we in our firms get the most value from this? You mentioned accountants are using chat bots a lot. Data entry. We're using it for data entry, getting the information into the system. Fraud and risk detection. Um, anywhere else that you notice that accountants are using AI AI actively day to day?

David Jani: [00:09:58] Well, aside from.

David Jani: [00:10:00] Those select examples mentioned before, we're also seeing it in elements such as predictive analytics. We're seeing cash flow forecasting as well, um, invoicing, smart invoicing and elements such as bank reconciliation. So it's been used in quite a wide number of cases, but there does seem to be a bit of a coalescence around analytics and process driven tasks.

Blake Oliver: [00:10:29] That makes sense. I mean, we're already using it for analytics. Uh, at least we have been with chatbots in a lot of ways. All right. Let's talk about the ROI. The number that jumped out at me from the report was 89% of AI users are reporting positive return on investment. What is behind that number? What is actually delivering the value there?

David Jani: [00:10:52] For the most part.

David Jani: [00:10:53] It does seem to be productivity. So that was the most common answer. So half of the accountants we surveyed said that they saw productivity gains as their biggest reason for ROI or benefit. Reduced errors was another major one, which was also around half just a little bit below. And then there was a little there was some focus on specific tasks such as faster, close and reconciliation.

Blake Oliver: [00:11:21] So one of the problems with AI is that we are using it more, we're trusting it more, but it's also seriously flawed, or it can produce seriously flawed outputs. Hallucinations. What did you find out about how often accountants are catching errors in the AI output.

David Jani: [00:11:40] Well, this was something we wanted to look into specifically because, as you say, it's an expected outcome of using some of the tools right now because they're they can be very effective as borne out by some of the benefits mentioned, but they're also not 100%. And you can't necessarily entrust very sensitive information that you would use in accounting or sensitive processes that could be followed by compliance regulations without some care taken. So that seems to lead to a situation where, well, 48% of the people that we spoke to are checking all their outputs manually. And that is just to make sure that the errors and the veracity of the information and the validity is all correct, and it's for good reason. Basically, there are quite a lot of mistakes made. So there's around a third. So Specifically say that they see errors in their AI outputs more than half the time. So it's still an area where people are being cautious and perhaps for good reason. For the moment, just because it's better safe than sorry, really.

Blake Oliver: [00:12:54] So how do you reconcile that with the 89% positive ROI number we just talked about? If we're having to check everything and we're catching a lot of mistakes, where do you think the productivity is coming from?

David Jani: [00:13:10] Well, based on some.

David Jani: [00:13:11] Of the other findings, particularly around the sorts of roles that are being affected in AI by hiring at the moment, not necessarily because of AI, but in general, my feeling is it's freeing up some time from some tasks, but it's not always freeing up time from other tasks that you might have to do manually. So it's, it's creating some, some gains in some areas, creating some extra work in others. But maybe it's given the fact that so many people are citing productivity gains. My, my guess, but this is just my guess is that it is creating some extra extra room where you can do other things too. But yeah, it's still still caution necessary.

Blake Oliver: [00:14:02] So even though the AI is helping us work faster, we're having to review more, but it's still less time, right? Like if we net the hours that we, uh, saved on doing the work with the review that we've added, we're still coming out ahead.

David Jani: [00:14:21] I think, I think that seems to be what is borne out from the numbers, but I think it's important that businesses still keep a close eye on the ROI of these situations. And, you know, really check that it is delivering those gains carefully.

Blake Oliver: [00:14:36] Are there any like specific tasks where AI is reliable enough to trust? Or are there areas where it's like still really error prone? Like, what is your data suggest? Or just what is your thought on this having having researched it.

David Jani: [00:14:50] We don't have any data on that particular nuance, unfortunately. So I can only give you my best guess. And I think it depends on the complexity and the sensitivity of the task. So if it's going to be something more day to day. You probably don't need as many eyes on it, but I still think I would recommend keeping an eye on it because as we saw in the data, Eric, checking and error reduction is a thing that AI is quite good at. But I think it's also you need to give AI a hand in the same way that, you know, you might want a second pair of eyes on something. It's good to have a second pair of eyes on the AI. I think that's my takeaway from it.

Blake Oliver: [00:15:33] All right, let's get into security. This is a big concern for the accounting profession. What is the data show about security data breaches in accounting in general. But also when it comes to AI if you've got that information.

David Jani: [00:15:52] Well okay. This is an interesting picture that we see. And first of all, there's some bad news, which is we see a lot of people are affected by data breaches generally. So this is more of a bigger picture situation. So half of the people that we spoke to, 52% specifically have experienced a data breach at some point in recent times. So the last two years or beyond that, basically. And while I don't have any specific information on whether AI is the cause behind that, what we do see is a bit of a worrying habit, which is that. Most companies don't have clear guidelines on how their A how their AI tools are used with sensitive data. So the case where there are the most guidelines seems to be around employee and payroll information, which you would expect, but that's only 49% of the entire survey sample, which.

Blake Oliver: [00:16:53] So less less than half. Less than half of the respondents have AI guidelines on putting payroll and HR data into AI systems.

David Jani: [00:17:01] That's correct. Yes. Which was a bit of a surprise. Um, obviously accountants accounting information like you say, is super sensitive and you've got to be super careful with making sure that it doesn't get out there and breached in some capacity. And yes, information like employee and payroll, bank reconciliation and customer billing and payment, those were the most, um, covered by any kind of guidelines. But this is below. Below 50% in all cases.

Blake Oliver: [00:17:30] Yeah, it sounds like we've got some work to do in the profession on, on creating those guidelines, because, I mean, the productivity gains from AI are significant and people are going to want to use it. And so if you don't give them guidelines, they might just upload payroll information into a into a AI chatbot. And hopefully they're not doing that in a system that isn't owned by the company. Right. At least if maybe they're using a free AI tool. And in that case, the terms of service usually say that the company can train on your data. And I mean, that would be a terrible situation, right, where you've got somebody uploading like payroll reports into the free version of ChatGPT. And now OpenAI can train its model on that, that sensitive information. We really need to step up.

David Jani: [00:18:28] I completely agree with that. Yes, I think that would be one of the big pieces of advice I give from these findings, which is guidelines around this are they don't seem like much and obviously everyone is rushing to get the get AI tools, use their data with AI tools and so forth. But it's a huge risk factor that does need to be addressed.

Blake Oliver: [00:18:49] So when you ask accountants how they perceive the risk of AI in their software, what do they say? Does their perception line up?

David Jani: [00:18:58] No. Um, quite the opposite in fact. So what we found was there was most, it was most likely that people addressed it with a kind of feeling of minor risk or insignificant risk. Um, whereas a few people see it as a moderate risk, but many people don't see it as a critical risk. So in this case, 48%, sorry, 49% see this as a minor risk, Whereas only 3% see it as a critical risk. And with 3,536% seeing it as moderate and 12% as insignificant. So it really skews more towards people not being too concerned about it. They don't see, um, a cybersecurity risk from the introduction of AI features posing much of a threat. So this might be why there are so many people who don't have guidelines. Unfortunately, there just isn't maybe a perception of the risks that are at play.

Blake Oliver: [00:19:57] All right. Let's pivot to people. We talked about this staffing crisis a little earlier, and I think you said something interesting that that is is enlightening about this because there's a seeming contradiction, right, where AI is making us way more productive in accounting. And you would think that would reduce headcount that companies would eliminate accounting and finance roles, but we're not seeing that yet. And I think you kind of hinted at a reason why, which is that we have to review the outputs of the AI. Still, there's a lot of review that has to happen. So, I mean, that's my guess as to why we haven't seen, you know, job losses in accounting yet from AI. I'm curious to know more about this. You know, what is your data show about the state of staffing and accounting right now? Is it is the talent shortage still as bad as it's been for all these years? Is it is it improving? Is it the same?

David Jani: [00:21:02] It would seem to me that there's still there's still a problem for accounting in terms of hiring people. 73% of the of the accountants we asked say their companies have trouble with retention. And there are hiring challenges, basically. And this is driven by a few factors, the factors, the biggest one being a lack of qualified candidates, though.

Blake Oliver: [00:21:23] What roles are firms having the hardest time filling?

David Jani: [00:21:26] Well, this is quite interesting because what we're finding is not really, um, it's not super advanced roles, but it is kind of mid level. It's financial analysts. It's specialized accountants mostly. Um, so for example, when I say specialized accountants, I mean tax or cost basically. Mhm. And these are the hardest roles to fill. So around a third say that they struggle to find financial analysts. And just below that sort of around a quarter to a third say that the specialized accounting roles are the hardest to find in this current moment.

Blake Oliver: [00:22:01] So the talent shortage continues. Our firms using AI to solve this. What what strategies our accounting teams leaning on?

David Jani: [00:22:14] Well, basically it seems to be mainly through upskilling that people are trying to get around the issue. So making the most of the accountants that they already have. I was, um, as I mentioned at the beginning, this was quite surprising, is that it's more than double the amount of people who are doing, um, automation to try and get around this. So 21% use AI to fill staffing gaps compared to 40% who are doing the upskilling process basically. And then you've got conventional methods of hiring, such as hiring through job boards, which is a third. And there's even more who are actually using graduate intern programs to try and fill the gaps in their staffing. So I think that some good news for entry level, for a start, but it's one thing that I would take as a little bit of a as a little bit of a caution with that is that it's similar. So AI to fill the gaps is 21%, 23% for graduate intern interns filling the gaps. So I think there's more competition at that level, but it doesn't seem to be this it's not replacing human accountants. Basically. I think it's the takeaway from this. If anything, it's probably driving a little bit more competition in the middle of in mid-career roles and is encouraging companies to try and upskill to get around this, to meet the meet the specialization that comes with running some of the tools as well.

Blake Oliver: [00:23:51] Right? Because we need more experienced people who can review these AI outputs. So we need to upskill our staff so that they can actually do that effectively instead of doing the work. That seems like, uh, quite a challenge.

David Jani: [00:24:07] It's an interesting.

David Jani: [00:24:08] Problem.

Blake Oliver: [00:24:09] Okay, so let's talk about what's still manual. Let's talk about what is being done by AI. I guess we actually already did. I'm curious to know if you have data on like what is still being done manually? What is, uh, you know, so we can compare that to what is being done by AI. Like what do we have to still do manually, at least some of the time?

David Jani: [00:24:35] There's quite a lot still being done manually. Um, the most common answer to this question was financial reporting. So just over half are still doing that. And when we say manually in this case, when we ask that question, uh, what that means is, is they're using perhaps spreadsheets or a few other like manual processes rather than doing it by hand on paper, for example, accounts payable and receivable are also something that is still quite commonly managed this way, as well as billing and invoicing and payroll. So it seems to be a lot to do with, um, with the, uh, the payment and reconciliation kind of elements.

Blake Oliver: [00:25:16] Well, you mentioned spreadsheets, so that's always a question when new technology comes along. Uh, how are Excel and Google Sheets doing? How are they faring in this AI era?

David Jani: [00:25:27] Well, rather.

David Jani: [00:25:28] Unsurprisingly, they're not going anywhere. Um, they're very, you know, yesterday's, um, accounting replace accountant replacement is still among us. And there's still accountants basically, but there's still 51% of people using Google Sheets or Excel for this for their, some of their financial data. So it's still with us. That's the simple answer.

Blake Oliver: [00:25:54] And probably will be until the end of time. At least it seems that way. Uh, at least at the end of my career. Anyway, it's going to it's been around for what now? Um, 40 years, right? I think, uh, 40/40 years since the invention of or the release of Microsoft Excel, the first version. And here it is still going strong. All right. So let's talk about the future. What challenges do you see accountants facing wrestling with over the next 12 months? What's the what's the top on your list?

David Jani: [00:26:31] Well.

David Jani: [00:26:32] The main problem seems to come down to budget and budgeting and forecasting. However, it's interesting because it seems to be a balance between day to day work such as that. And the other one is figuring out how AI is used or how they can use AI most effectively. Um, and yeah, that doesn't surprise me. I think, given everything we've talked about so far. That's, it's, um, a question that many accountants are using. But then I think with so many different ways to apply AI, as you already mentioned in tech stacks, there can be integrations, there can be full stack, you know, whole platforms that have got it built in. It's trying to understand that, I think, but also just trying to understand the kinds of features that it can apply to and where it fits into their businesses. Basically, it's trying to get into that, but doing it in a way that makes sense.

Blake Oliver: [00:27:31] So budgeting and forecasting and determining how to use AI, that's the those are the top two. Is that, um, concern about budgeting and forecasting tied to the economic volatility that we've been living with for the last few months or years.

David Jani: [00:27:51] That's my.

David Jani: [00:27:52] Theory. Yes. Because what we saw in the last survey that we did before was that, um, elements such as inflation interest were the biggest worries for accountants for the well at that point the coming year. So we're now in it. I think that factors into it somewhat.

Blake Oliver: [00:28:12] So I'm going to ask you to, uh, predict the next survey results. Right. You've done this before. You just did it. Uh, if you do this survey again in 2027, what do you think is going to move the most? Where do you where do you predict it's going to go? I know that's a bit unfair, but I'm going to ask it anyway because, you know, you've you've done this before. You've seen what moves. What do you think is going to go up or down.

David Jani: [00:28:39] Yeah. That's an interesting that's an interesting question to try and answer right now. Obviously with everything changes so quickly. Um, I would say that it would be interesting to come back and look at where those applications of AI and the features that people are finding most useful, usable, useful and that they're trying to target. Whether that changes much, I think would be an interesting thing to try and look at whether they. The return on investment remains. But I'm particularly. I would be particularly interested to see what happens with the security question and whether anyone has had any kind of data breach from AI. I think that's yeah, those numbers would be interesting if maybe a bit scary and hoping if I'm making if I want to make a positive prediction, though, I'm hoping that we would see higher numbers. I think we would. I think that's I think what will happen as these tools become more mature and as people become more integrated with them, is that more people would have their have their guidelines updated, so they'll have special guidelines set up for how data such as payroll and cash flow, as we mentioned, would fit into that, and what rules need to be followed to ensure that it's secure and safe, basically.

Blake Oliver: [00:30:03] Yeah. There's so many concerns about AI security. Um, you know, we mentioned this risk of employees putting data into AI systems where the terms of service don't protect the data, but I'm actually less worried about that than I am about prompt injection. This idea that you can submit documents to an AI system that has hidden text in it with instructions, and then there's not a defense mechanism for this AI agent that's receiving the document. You know, let's say it's a bill or an invoice going into an accounting system, and it reads that hidden text and follows the prompt and then sends data out to a, to a threat actor, to somebody who's trying to steal from that organization, or maybe change like payment information, things like that. I mean, I feel like we're just at the beginning of that kind of fraud or cyber security risk?

David Jani: [00:31:06] I would agree with that. I would say that it's something that gets me it gives me some worries, um, as well, just because it is, it is a very new and rather sophisticated way of extracting information from a company or from a, from an organization. We still haven't, we still don't really know enough about it. I think that's the problem. It's it'll be interesting to see, I think, where that develops and how that affects companies, because it will be big news when it does, we'll hear about it, that's for sure. It's going to be something that's going to weather it. I think the question is whether it will be a big it will be bigger companies or smaller companies that bear the brunt of that, given the AI adoption and the size of bigger, you know, the the prize of being able to get into bigger companies for their data, it's probably going to be one of them. But yes, this is, um, this is something that will be interesting to watch.

Blake Oliver: [00:32:05] All right. So let's say I'm an accounting firm leader or I'm a CFO and I'm trying to decide where to put my AI money. I've got budget to put AI somewhere in my firm or with my team. What would be your advice to me? Where should I bet my bet my money? Where would I, where should I put AI in place to get the maximum benefit?

David Jani: [00:32:31] That's a very big question. My based on what we've seen in the data so far. I think what we probably would probably recommend is that companies focus on their individual cases. Really. It's very nuanced, understanding how how certain features fit in with the kind of accounting you're doing. And let's, let's remember accounting is such a wide web of different kind of specializations and necessities depending on what sector you're in. I think it's very important to try and map your to map things around that. But if we're talking more generalist, just to answer your question properly, it seems to be that elements such as data entry are quite useful, as well as anything that gives you a chance to make your planning easier. So in terms of features that have helped, for example, accountants map around the cost changes and volatility we've seen in recent times. There's things like predicted cash flow and analysis dashboards. And now analysis we know is already quite embedded with AI. But I think it's going to be more to do with tools that can give you some predictive modeling.

Blake Oliver: [00:33:46] David, where can listeners find the full report if they want to dig into this more on their own? Where can they find that and where can they follow your work in general?

David Jani: [00:33:55] Well, you can find the reports on capterra.com at the moment. Um, it's on, uh, one of our blog pages and that's where a lot of my work can be found in general. Uh, we have an accounting section, but we also have sections that are dedicated to other topics such as HR, uh, project management, and various other important, um, business software use cases.

Blake Oliver: [00:34:19] David, thanks so much for joining me. And, uh, thanks for putting together this really interesting report. You can read the full Capterra 2026 Accounting Software Trends report@capterra.com. We'll put the link in the show notes. And if you want to earn free CPE for this episode, you'll find the course in the earmark app. Go to earmark.app in your web browser, or get the free app from the App Store. I'm Blake Oliver, thanks for listening and I'll catch you next time.