chang: killing saly, one agent at a time | the disruptors - 卡塔尔世界杯常规比赛时间

fieldguide’s ai rewrites the rules that have held audit back for decades.

this is a preview. the complete 1-hour video episode, with commentary and transcript, is first available exclusively to pro members | go pro here

sponsored by tax season readiness: practical steps for a smoother busy season | see today’s special offer

while older partners worry that today’s technology will prevent staff from learning, chang responds: “i don’t know about you, but my first two years as an auditor, i wasn’t really learning audit strategy. i was learning how to test something as fast as possible.” instead, fieldguide intends to fast-track junior auditors into more strategic roles where they manage ai agents and think like senior associates from day one.

an aicpa model from several years ago predicted that firms would shift from a pyramid to a diamond shape, but chang disagrees. he sees a narrow pyramid built on a true apprenticeship model, rather than today’s churn-and-burn approach. a pyramid shape allows for the inevitable attrition as audit staff leave for industry jobs or other opportunities.

chang encourages firms to think about their next decade, not just the next two or three years. “audit transformation is not easy. it typically involves pain, and pain is guaranteed,” he says. the winning firms, according to chang, “won’t be doing the same as last year.” instead, these firms will be “rethinking the servicing equation,” which means that manual work will be done by ai agent teammates directed by a “superpowered staff” and overseen by a senior, who is more like today’s managers. instead of a bottom-heavy pyramid, firms will have leaner audit teams and will be able to “offer everyone, people at every level, a more advanced career.”

“ai enables humans to do uniquely human things,” chang explains.

10 key takeaways

while many audit software vendors promote digital audits, this often amounts to merely moving paper binders to the cloud, which doesn’t lead to significant changes in productivity.
generic ai tools often lack essential security protocols, accuracy, and the audit trails required by regulators.
purpose-built ai like fieldguide’s ai agents are trained on gold-standard data, with aicpa-approved language, and reason like experienced auditors.
saly isn’t all bad – it informs future audits with client context and nuances. ai agents can combine the knowledge base of the prior year with fresh approaches tailored for the current year.
nothing drives a client crazier than an auditor who keeps repeating the same questions.
legacy vendors struggled for over 20 years to move to the cloud. building ai is more complex, so don’t expect it to deliver in five years.
using chatgpt to manage your personal life is a good way to learn its benefits and limitations. however, for audit, the bar needs to be much higher.
retiring partners have no incentive to invest in technology, while younger partners want to build for the future, making consensus difficult to achieve.
firms that think or make decisions like corporations can move more decisively and with more speed and precision than partnerships.
the biggest challenge in audit transformation isn’t technology, but rather a lack of clarity or a practical roadmap.

more about jin chang
jin chang is the ceo and co-founder of fieldguide, the leading agentic ai platform for audit and advisory firms. a former big four practitioner, chang spent years in the field wrestling with outdated tools and manual workflows–experiences that inspired him to return to the profession as a founder and build the platform he wished he had. under chang’s leadership, fieldguide has become the trusted ai partner to nearly half of the top 100 cpa firms, helping firms unlock new capacity, improve quality, and elevate the work of practitioners. chang’s work has been recognized by accounting today’s top 100 most influential people and cpa practice advisor’s 20 under 40 list. chang speaks regularly on ai, audit innovation, and professional services transformation, bringing a grounded, team-first perspective shaped by his background as a cpa, technologist, and first-generation founder.

transcript
(transcripts are made available as soon as possible. they are not fully edited for grammar or spelling.)

liz farr

welcome to accounting disruptor conversations. i’m your host. liz farr from 卡塔尔世界杯常规比赛时间 and a special treat today, my guest today is jin chang, ceo and co-founder of fieldguide. welcome to the show, jin.

jin chang

thanks liz, for having us and excited to be here.

liz farr

well, great. now i i’m always tickled to find people in audit to be guests, because that is an area of accounting that really i think it’s just it’s dragging its feet moving into the future. so i’m really glad to have people in audit on the show. now, when you co-founded fieldguide, what were the pain points in audit that you wanted to solve?

jin chang

well, it’s interesting that you point out the fact that audit has been dragging its feet, and i would certainly agree. now the good news is i think there’s a lot of movement happening now in the audit, but going back to your question of how and why, i co-founded fieldguide, i always like to describe fieldguide as the company i never imagined starting, and that’s because i started my career as an auditor, and i, while i appreciated the career and the people i worked with, and i still maintain strong friendships with today, i didn’t like the manual work involved. and we all know in the audit life cycle, there’s a ton of room for human judgment and critical thinking, but there’s so much work being done and so much time being spent on manual testing, ticking and tying things that even back then, when i was practicing, i would ask myself, why aren’t machines doing this better faster? why did i get a four year degree to do this matching exercise between evidence and samples being tested? so even as a junior auditor, i would question why i was doing things manually. so frankly, fieldguide is the platform i wished i had, and the ai agents that we’re building are the ai team members i wish i had to collaborate with. so yeah, you know coming boiling it down to really trying to automate the manual work so that human practitioners can really focus on expert judgment, client advisory, client service, building relationships, building a faster career trajectory, so that they have a promising career that i didn’t see as attractive back then.

liz farr

yeah, yeah. you know, in my time in public accounting, i was often the warm body from the tax department that got pulled into audits. and so they, they would always give me, oh, here’s the easy stuff. do this, and i just like, i don’t understand this. this is really manual, and i have to do this checklist. i don’t understand that. it just made me hate doing that work, because it was so manual and so tedious, and i didn’t really see how the piece i was doing fit into the whole

jin chang

i love that. first of all, i love that you use the word hate because i didn’t use that word. but my deep rooted feeling towards the manual work is actually of hate. i hated doing that. and the thing that i always like to emphasize is you mentioned something around how you didn’t really have time to understand why you were doing things. you were just fixated on doing the manual work as fast as possible. a lot of audit partners and leaders asked me, if ai is automating the junior work. what are my people learning? and my rebuttal to that has always been, i don’t know about you, but my first two years as an auditor, i wasn’t really learning audit strategy. i was learning how to test something as fast as possible, and typically, same as last year, so

liz farr

exactly, exactly, yeah. now. now, what would you say are the biggest challenges facing audit in advisory today, touched on some

jin chang

sure now we touched on kind of the manual work. but if we were to zoom out for the entire industry, i think overall the audit, and i would argue, all of cpa services, but when you look at audit and assurance, there’s this rising set of client demands and expectations. really they want to have quality assurance at reasonable fees. and reasonable fees implies not the highest fees, right? but what that also means is, you know, quality assurance and reasonable fees don’t often go hand to hand. they’re often opposite levers, right? because lower fees implies fewer hours, implies lower quality. let’s call it. what that usually ends up looking like is lower prices, but still holding a high bar for quality, resulting in lower margins and burnout of talent, because as junior auditors, you still have to deliver that quality at reasonable margins, at your target margin, that also drives a bit of a what i like to say is a downward spiral of burnout. junior auditors are like, i need to deliver this quality. my leaders are saying, i have to deliver it in this budget. you still have to do the job. so that’s what’s creating the burnout and retention issues in the industry? well, in part, of course, other factors are the salaries aren’t quite as competitive, the promotion and career path doesn’t always look that promising or clear. so those are kind of the macro issues of the industry. so when i speak with cpa firm leaders, audit and assurance leaders, a lot of them are starting to look for generative and agentic ai to help them solve both quality and fees, to improve both at the same time. and we have the great fortune of having this transformational technology in front of us that can help us accomplish both, at the same time, better efficiency and quality. now, of course, with any hype cycle that you and i have seen through the years, there is a ton of hype and noise. in fact, in most hype cycles, i would say all there’s more hype and noise than substance. but the key in every hype cycle, my belief, and i believe everyone at our company, is that we need to find the real use cases, the real practical use cases that are valuable for an industry, not just market ai, for the sake of marketing ai, and if we focus on truly valuable, impactful solutions, we will be rewarded for that with really great usage and amazing partners and customers, right? so, yeah, we’re trying to sell but help solve the quality, the fees issues. and we believe that, and we hear this increasingly, from cpa, from leaders, we believe that the winning firms of tomorrow will employ a blend of both human expertise and ai augmentation

liz farr

exactly, exactly, and that that kind of segues into my next question, which is, you know, how do we design a future of audit that serves the clients while also works for people? and you just answered parts of that by using ai agents to help keep the fees reasonable yet also ensure quality so what else? what else can we do?

jin chang

well, reasonable fees and great quality assurance typically happens in the practitioner workflow, but when it comes to client facing interfaces, enormous opportunity for ai leverage too. in fact, we are hearing from firms that ai is helping to provide a more modern audit experience for their clients. you can imagine in the client facing interface, a lot of the pains you typically hear from clients is there isn’t enough transparency. i have to wait for an email from our audit team, and the audit team is trying to do their best, but they don’t have great project management, right? so they’re doing their best with what they can, right? but that’s what the clients see, lack of transparency. now, they also get a ton of repeat requests because auditors don’t have great project management tools, they’re trying to juggle many audits clients at the same time. so often you’ll get repeat requests. sometimes well intended, the clients will forget to respond to things right? so there’s a ton of areas for improvement. let’s call it in the client interface. yes, what we see in the modern audit experience is when, when an auditor requests something of their clients, and the clients upload evidence or integrate their accounting system to facilitate data into the audit process, ai should do a first pass of whether the pbc meets the quality bar for the audit team, right? you know, the typical life cycle today is the client responds to a ppc request and they might hear back from the auditor a week later, okay, right? the worst possible thing is seeing from your auditor you missed this one sample selection, right? so this is one use case of many that we are starting to roll out where the agent is basically telling the client, hey, we think this is going to be a high quality response to your auditor request, or, very quickly, let them know, hey, for one of the 25 samples you responded to, you’re just missing the evidence. or it’s, you know, it’s perhaps not meeting the quality bar. that saves just practitioner time, but also clients too. so that’s one way we are enhancing the client experience, right? the other way we’re really enhancing the client experience is what we like to call, internally as our domino’s pizza tracker. so it’s been a while since i’ve ordered domino’s, i’m trying to save a few pounds, but but last i saw when you order from domino’s, you can see when it’s being prepared, when it’s in the oven, when it’s picked up by your delivery person, or when it’s just a few minutes away. so fieldguide, we provide that transparency and it’s under full control of the audit team, because we understand every audit life cycle could be, could have nuanced differences, but we enable the audit team to communicate in real time with their clients where the audit process is, and we’re starting to incorporate more and more intelligence as well so that we can coach clients and practitioners on where project management is.

liz farr

well, that sounds, that sounds ideal, you know? and i remember in my days of sending off the pbcs and going through and going, well, they provided eight of the 10, but these two they didn’t provide, or they provided something, but it’s not what we asked for, that’s right. so something like that would be just a huge payoff and and this sounds like this would also reduce burnout and overwork among auditors, which is really important, since we have a huge talent shortage, it’s been going on for a long time. but am i? am i hearing things correctly that that would help?

jin chang

i mean, when i reflect back on being a junior auditor, every day, usually every night, i would go through my open items list, try to organize my pbc requests and then send another request list at the end of the day that itself took an hour plus every day. we’re effectively automating that phase almost entirely, right, so that the request lifecycle is more so on autopilot. what this also does is, by the time evidence and pbcs come back to the audit team, it’s basically ready to test, because it’s, in some ways, already pre-tested. we do make that distinction between request phase and audit phase, by the time it gets to the audit team, it’s already been pre scanned by ai and then we also have an ai agent to do the testing and documentation of results.

liz farr

oh, wow.

jin chang

so when evidence or pbcs come back. our ai agent will do the initial test against the test attributes that the audit team has determined or the audit programs, and our agent will basically follow those instructions, test all the evidence within seconds, document the results, whether they met the bar or there was an exception, and also automatically ties the evidence to those samples selected. so you can imagine even a reviewer in the audit team coming in wanting to look at the piece of evidence, not having to shuffle through things, just hovering over that sample selection, and out comes the evidence that was, you know, relevant to that,

liz farr

yeah, that makes it a piece of cake.

jin chang

i mean, you know, turns out there’s more grunt work to automate than we initially expected. but, you know, just a quick data data point for you too and your audience, we are hearing firms start to say we have this eight hour test or task that we are automating down to under an hour now.

liz farr

wow.

jin chang

huge improvements,

liz farr

huge improvements. yeah. and so that means that for that seven hours you’ve saved, the audit team can be doing other things that are a little bit more valuable,

jin chang

absolutely.

liz farr

and a little bit more useful.

jin chang

valuable for the client, valuable for client relationships, valuable for their own experience too. yes, i’m trying to help avoid junior auditors thinking, why did i get this four year stem degree to do this manual work?

liz farr

yeah, exactly. now. now, speaking of technology, for a long time, digital technology and audit was just things like excel or your your binder tool i used engagement, but these days we’re seeing intelligent audit. so what is intelligent audit? you’ve already talked about some of it. you know, the the agents that look at the samples and can do the testing for you, which i think is great. so, what is intelligent audit?

jin chang

great question. because right now, i think what we see in the industry is folks talking about a digital audit or a digital audit binder, right? really, what it means is, it’s the same type of work we used to do on paper, and it is now digitized in the cloud. well, not always in the cloud, but, you know, some in the cloud. let’s call it. and when i see like our philosophy is that we should be pursuing step change improvements in human productivity for a space that really needs more productivity. we don’t have an unlimited supply of people, right? but what we observe in the market is most vendors just wanting to digitize the audit process, right, and calling that the next gen solution, i would argue, marginal improvements at best, but not true step changes in human productivity, not eight hours going down to one hour more like eight to six or eight to five, right? so our approach is, let’s there are certainly things that we need to preserve in how we currently do things, but let’s take a leap forward in human productivity. it starts with upfront intelligence during the planning phase. so taking a true risk based approach, but not always a risk based approach. i would argue there’s an argument that for some private companies, smaller audits, it’s actually not that efficient to do a risk based methodology. but we do believe that fundamentally, risk based is the right start. i. uh, now, if the current generation of even risk based methodologies, if we’re to simplify what they do, they effectively take in client context and then tell the humans what to do, right, right. go through these checklists, these test steps, audit programs, and yes, there’s some value to that, to doing to doing it automatically, i would argue, not every methodology does that right automatically, but we believe it should be an automated suggestion of audit programs or procedures, but that alone is still what we consider a digital audit. where we need to get to is the machine shouldn’t just tell the humans what to do. it should take the first pass, right? why? why are we still doing the same manual testing that is driving junior auditors away from the profession? i don’t think a risk based methodology would have kept me in the profession. what would have kept me around is automating the painful testing, right? so once a methodology suggests go, you know, go through these steps, we have agents that will perform the first pass of those tests. it will be involved in the test procedure design. we have agents that write pbc requests that go straight to the client. when the clients upload evidence and respond, our agent does the initial analysis. by the time it gets to the audit binder or audit team, it’s test ready, and our agents do the first pass test and documentation of results. what it surfaces in that process is, if something doesn’t meet the bar, it proactively suggests to the auditor, your client is missing this piece of evidence, or we analyze this evidence and it’s not quite meeting the bar. would you want me to send a follow up request to your client? now, we’ve designed it that way, with intention, for a human in the loop, approach to review before it gets back to the client, but you can see how there’s intelligence built in more and more right where the machine is starting to say, hey, we analyze this. we’re going to have to follow up with your client about these things, these additional pieces of evidence or additional inquiry, or so on. and then it’s just kind of, in some ways, starting to coach humans what the next step should be. so our vision for the future of intelligent audit solutions or an ai powered platform is, it’s actually in our company name, fieldguide, we always had this vision where ai would get so good that it would start to coach or guide practitioners in the field, start to preempt what the next step should be. so you know, i would characterize a lot of our agents right now to be augmenting staff one to two level work, maybe senior one work. what we’re starting to see is some firms really pulling us into more senior levels of work. so building agents for senior associates, managers, for partners, specific agents for the planning phase as well as reviewer phase. so reviewer agents, planning agents. now we’re starting to work with some firms to try to anticipate what a peer reviewer might look for as well. so you can see that we’re starting to employ more and more kind of advanced techniques that where agents start to mimic more of kind of senior level practitioners.

liz farr

i love it. you know, all these things that the ai is doing are all the things that i always resented doing, you know? and as i’ve talked to other people who’ve used technology in audit, i wonder, you know. well, why are we still doing this manual ticking and tying and matching? and you know, it just doesn’t make any sense, and just moving your audit into the digital realm is not really much different than taking your paper workpapers, because i worked long enough ago that i did do paper it’s not much different than taking that paper binder and threading it through a scanner and then you have this, like nice pdf. so i’m really glad to hear that a lot of the awful stuff is going away.

jin chang

yeah, we’re well, thanks for rooting for us in that department. yeah, obviously we, we share those views quite a bit. so yeah, want to improve the lives of auditors, and, you know, make it an attractive career. so yeah,

liz farr

now, now you just talked about the ai agents that do a lot of the work for you, but we’ve also heard risks about using ai, that the data may not be safeguarded, that ai hallucinates, or that audit quality may be at risk, and especially when everybody you know, all that anybody knows about ai is those chat bots or chat gpt. so how does field guide address those kinds of risks?

jin chang

great question. and actually, these are the same risks, risks that we talk about regularly, too. now we always caveat this by saying it’s not a blanket set of risks across all ai solutions. in fact, this is why we make the case that we should be, you know, pursuing purpose built ai solutions that really take into consideration the industry requirements, their tendencies, their behaviors, the regulators and standard setting bodies who who oversee them and govern them, the level of documentation, the kind of reasoning that the ai needs to go through, the auditability, the traceability of those steps. these are all real risks, security, privacy, of course, these are all real risks when you use a generalized chatgpt-like tool for professional uses right now, i’d be the first one to say chatgpt is a great product. i use it almost every day in my personal life. it helps me plan things on the weekends when i don’t have time to do research, it helped my wife and i kind of plan out our honeymoon. wonderful use cases like that, right? but in a professional setting, the bar needs to be much higher, right? using chatgpt in your personal life, i always encourage because it helps people understand both the benefits and its limitations, and it helps you get comfortable with both the benefits and limitations, right? so it starts to really kind of upskill people to know how to use ai. now in the professional setting, there are many other considerations, especially in a professional services vertical like the cpa firm vertical, we can’t just use something that, you know, we use in our personal lives. we can’t just put our sensitive client information and data into a generalized chatgpt-like product. you want to have and you want to partner with vendors who keep in mind the basics in security and privacy practices, not just having a proper, robust soc 2 attestation over their systems, but also pursuing ai compliance as well things like iso 42001 and so you want to see that the vendor doesn’t just treat security and privacy as a check the box exercise, but they start to really consider the risks around ai and control, building, building in controls to mitigate and manage those risks so that within a safe environment, professionals like cpas can make the most of generative ai capabilities while mitigating or eliminating the risks involved. now that’s on the security and privacy front. what about in practice, hallucinations, accuracy? this is where further domain or industry specific tuning comes into play. if you were to drop in a set of test procedures and evidence into chatgpt, it will probably do a 70% decent job of performing initial tests. but when you really double click and look under the hood and you read what it’s writing, it doesn’t use always aicpa approved language or recommended language, it doesn’t reason and explain the steps like an auditor would reason and explain their steps. the accuracy rate if it’s not trained on or tuned against, we call it gold standard data or files, then the performance is maybe 70% at best. but what we found is by really tailoring our ai offerings around auditor specific use cases and feeding it, feeding our ai agents with real gold data to serve as gold examples, then our accuracy rates greatly exceed 90-95%, even close to 100% at times. and when we see that in our data too, when we see that practitioners are accepting without revising our ai outputs. that really indicates to us that, well, this is really starting to get to auditor grade performance.

liz farr

i see, and now, now you mentioned earlier that you are creating bespoke ai tools for some of the firms you’re working for. so what are these? you know are, what are these kind of things do? i think that’s well, i’m curious.

jin chang

that’s a great question, because our observation has been that most software solutions in our industry are saying they have agentic ai, but when you really look at their products, they basically plugged in a chatgpt wrapper experience over some knowledge corpus. now, is that valuable? yes, but i would argue only marginally. the other kind of pain point we typically hear is firms trying to use chatgpt outside of their audit platform, in which case auditors just don’t use it. because when you have ai that resides outside of your audit platform and they don’t talk to each other, it’s almost, it almost takes more time to feed context into chatgpt than it does the same involved, right? so we hear that the usage isn’t great. so with fieldguide, we have the end to end audit lifecycle on one platform. it’s super integrated end to end. it’s not like we have a separate client portal, a separate audit binder, a separate financial statement drafter. everything is integrated end to end. on top of that, we have ai agents that are embedded into the workflow. so we talked about earlier having a request pbc analysis agent, right? it doesn’t feel like an auditor has to use a separate tool. it’s just embedded into our pbc workflow. same thing with testing evidence. auditors don’t have to separately, transfer evidence into another system, separately, write out what the procedures and risks involved are. and instead, in fieldguide, it’s embedded into the binder where our agent will proactively say, hey, do you want me to test this for you? and what it starts to feel like is, i like to say, even for a junior auditor, their work starts to feel more like a senior or a manager, where you are instead of writing instructions to an offshore staff and waiting for the results overnight, you write instructions, or in some cases, the ai agent suggests to you what the instructions should be. and then you basically tell the agent go or coach it a little bit, say, refine your approach in this way. and then you. um, because it’s embedded into the workflow, it just feels like another teammate that does the work in front of your eyes within seconds. so instead of waiting overnight for your offshore staff, you can go to bed knowing whether the tests worked or not. yeah, this is,

liz farr

this is great. i mean, now you know what you’re saying is that the entry level people, it sounds like the entry level people can start to work at a much higher level. now, you said earlier that those first two years you spent an audit were sort of a waste of time. you know,

jin chang

my former partners don’t hear that, but

liz farr

you know and and you know, and this kind of ties into something that the aicpa put out a couple years ago that the org chart for accounting firms in the future would no longer be this, this pyramid shape with lots of staff at the bottom, but sort of this diamond shape with just a few people at the bottom and more people in the middle. but the thing that that really trips people up is, well, you know, i learned by doing that miserable two years of testing. so how are we going to teach the new generation how to do this if they don’t know this and and, you know, i’m, i’m getting the idea that we just need to change how we think of audit period is, is that accurate?

jin chang

well, first, i’ll address the latter part of your question. and then let’s, let’s talk about the diamond versus pyramid. so firstly, what i do notice is, while there are certainly kind of, i grew up with that, pain type partners out there who, for some reason, want their younger folks to go through that pain too. i don’t know if any industry will be absent of that, right? there are always going to be folks like that. but the reason why i feel optimistic about the future of the audit profession is because i hear increasingly more and more audit leaders and partners saying, i don’t want my people doing that. there’s a reason. actually, a lot of partners are telling me i hated doing that work too. so so i feel a lot of optimism for this towards a more automated and better career experience for younger folks entering the profession and and i think i’m definitely seeing a tide change, because the industry talent trends aren’t quite supporting an environment where partners can, you know, force a painful experience. the reality is, a lot of partners recognize, a lot of leaders recognize, if i don’t change how we do things here, i may never build a great team. because, great practitioners, the smartest practitioners, they want to go to a firm that recognizes you know their value. and when, again, when someone with a four year technical degree is doing matching exercises, you start to question, is this firm maximizing my capabilities? so i feel like behaviorally, the industry is changing. even the leaders, the old school leaders, even they, are shifting their tone. now, when it comes to the diamond, the pyramid, i think the aicpa was ahead of their time in in thinking about and anticipating the talent trends that are changing now, now, and i think what their main takeaway or message was around kind of having more kind of higher level work for the bulk of the work workforce, and offshoring some of the lower level. work somewhere else. if you ask me, the diamond for me, the math doesn’t work, because typically when you look when you study populations, you still need a pyramid for the future of that population to be successful. this is just basic populations. when you look at even countries like we talk about replacement rates. so in order for a healthy, long term sustainable population, i do believe it still needs to be a pyramid, but i don’t think it will be a wide pyramid. some might argue that’s a pyramid scheme. we believe, i believe, that we should see a true apprenticeship model. and i say that with intention too, because a lot of folks say the pyramid model is an apprenticeship model. i don’t agree with that. it’s a churn and burn model. it’s designed for people to burn out. if we have a true apprenticeship model that everyone in the industry says they want to offer, then we should, we should be looking at the population as more of a narrow pyramid where, where we, if anything, have fewer people entering the profession. but it’s a true apprenticeship, where you give them a real, valuable experience, where they aren’t spending all their time just doing manual testing, and you start to teach them the true value in audits, what true audit strategy and risk assessments entail, so that the work is more interesting, so that their careers are fast tracked to be a manager even faster, so that they can manage agents and junior staff, but not an army of just human junior staff. so my thinking is like, if we can narrow the pyramid, and of course, some people will leave the profession and go into industry and so on. so you do have to have a pyramid, because people drop off, right? but we can narrow it provide a more valuable experience to our young and upcoming talent, so that they appreciate the career more and they want to pursue a partner path that’s that’s our proposition.

liz farr

well, that certainly sounds a lot more promising than the old fat bottomed pyramid. and you’re giving me something to think about that we still do need a certain number of people at the bottom, because, like you said, there will be some people who enter the field and then drop off. they move to industry, or they leave accounting completely. so there will be that. so you do need enough people to feed up into management levels.

jin chang

absolutely. yeah, yeah, yeah. again, it’s math.

liz farr

that’s a good point. a good nuance there. now, auditors are really notorious for following saly, which is same as last year, when they’re creating their processes and tasks. but you know, we’ve got some tech tools. you know, fieldguide is one and just plain old, linear regression is another tool. it’s an oldie, but a goodie. you know, what are some tasks and processes that no longer serve auditors today, what are the things they should just pitch?

jin chang

love this question because yeah, the you know, one of the biggest jokes in the industry is saly right, same as last year, and it comes from, let’s face it, what typically happens before busy season, you get your prior audit files, you make a copy, and you update the dates, and that is why the foundation tends to be same as last year or saly, right. at the same time. you know what’s funny is, like, saly gets a lot of jokes. is the butt of jokes, but there’s also some good in saly too, yes, and, and you can tell, i try to look for the good in things, too. and the good things in saly is there are client contact their client data points, nuances, contexts that are important when you carry, you know, audits, year over year, that context informs your next year’s procedures. that context even informs how you interact with that client. some clients want one list a week of requests. some clients want piecemeal every day, things like nuances like that, should carry on year over year. now, what shouldn’t be carried on year over year, at least from a first principles perspective is just taking the same exact procedures as last year without really reassessing the client context and the risks that are relevant this year, right? that’s where we need to bring a fresh lens too. so how do you build a system where you can take the benefits of saly and the first principles thinking of a fresh audit. well, our belief is that you have ai and agents and just a general platform that can roll forward the key things from last year and key lessons from last year. effectively, what that means is our agents are informed of last year’s audit. but it doesn’t just take last year’s procedures blindly. it takes it as a reference point. has an understanding of industry guidance and methodologies and the current year client circumstance, and then further tailors the procedures for the current year context. right? obviously, kind of theoretical conceptual, but i’m trying to kind of make it digestible for the audience too. but, and if you think about it, and i always joke like ai, is actually more human than we think. if you think about it, what does a manager do? a real a high performing audit manager doesn’t just take saly, but also applies the latest guidance and methodology and the most recent client circumstance, and then contextualizes the right set of procedures for the current year. so really, what we’re trying to train ai to do is think more like humans. yeah, that’s kind of our take on saly, you know.

liz farr

yeah, yeah, you know. and i agree, you know, there were some good things about saly. you know, when you figure out a really economical way, efficient way, to calculate something, then, yeah, you want to keep that going.

jin chang

and by the way, clients also expect, you know, there’s nothing that drives a client crazier than an auditor that, you know, keeps repeating the same questions over and over, yeah,

liz farr

exactly, exactly. and you know, and then there, there are the nuances. you know, something that was relevant to the client last year may no longer be relevant.

jin chang

that’s right. yeah, yeah. so yeah, trying to take the best of both worlds and both prior and current should inform the next audit

liz farr

absolutely. now, what do you think keeps audit from changing? what keeps us so anchored in saly?

jin chang

some will say it’s the partnership decision making structure, because, because you have retiring partners who have no incentive to invest in tech, they just want to ride off into the sunset without disrupting anything. you have younger partners who are betting the next phase of their careers and wanting to invest in something and sometimes it’s hard to get to a consensus decision. i think there’s merit to that argument, and i do think that the firms who think or make decisions like corporations, they tend to move more decisively and with more speed and precision. i definitely noticed that. but really, i think the biggest challenge tends to be in not having clarity or a roadmap. you know, ai, there’s so much noise, there’s so much hype. every vendor these days says they have ai, they don’t. every vendor these days say they have agents. they don’t. but for the cpa buyer, cpa firm buyer, it’s hard to have clarity through that noise. so our belief is that we need to provide a great, actionable framework for the industry to benchmark themselves against peers in their ai and transformation journeys, and it’s okay to be a little bit behind. it’s also great to be a little bit ahead, but it’s most important to know where you stand, so that you can start to chart your course towards the future. so that’s why at fieldguide we you know, it’s so fascinating. at first, we thought, by building the industry’s best ai tech people will use it. a very tech startup way of thinking, by the way, right? and i acknowledge that as you know a tech startup founder, what we learned very quickly is tech is less of the issue. it’s actually more behavioral, having a practical roadmap. so what we are building at fieldguide beyond the product itself is a really robust servicing playbook. in fact, we consider ourselves internally as management consultants or a technology consultant that’s helping a firm through their transformation journey, and oftentimes that’s a multi-year transformation roadmap. so, you know, i would say, fieldguide. we take a very consultative approach. we work closely with our customers on identifying where they are in their journey, and then we make recommendations for how they can achieve the next level of autonomy and intelligence in a reasonable timeframe. and, yeah, that’s the roadmap that we’re trying to provide for the industry.

liz farr

i like and, and, you know, i like that you’re you’re looking you’ve realized that the biggest change is mindset. that’s right, it’s it’s not enough to just put the tool out there. you’ve got to also get the people comfortable with using something that’s different from the way that they have done their work for, you know, maybe five years, maybe 10, maybe 30 years.

jin chang

absolutely, the mindset shift, audit leaders being open to doing things differently, not just same as last year. yeah, and helping them see the value in that we we don’t necessarily expect everyone to see the value in ai automation right away. we expect, we expect a lot of hand holding to be clear. so yes, whether it’s just answering security and privacy questions, whether it’s understanding where the best ai use cases are in the audit lifecycle. we’re here for it all.

liz farr

fantastic. now, speaking of the future, where do you see the profession in 10 years?

jin chang

i love this question because, and actually, what you’ll notice about our messaging as a company is we encourage firms to think about their next decade. and the reason we say that is because we find that some firms are making decisions just for the next couple years, next two, three years. and let’s face it, audit transformation is not easy. it typically it involves pain, and pain is guaranteed, i would say so. we encourage friends to think, okay, when you select your technology partners, don’t just look at the product today. try to forecast what you’ll need in the next five to 10 years, too. and is this partner going to keep up with those trends? right? and if you know, if your legacy vendor struggled to get to the cloud the last 20 years, don’t expect ai in the next five years. i would say ai is harder to build than cloud. now to answer your question directly, we view we believe that the winning firms, 10 years from now, won’t be doing the same as last year. will be rethinking the servicing equation, the math involved in delivering an audit engagement. and what do we mean by rethinking the servicing equation? it’s understanding that junior level manual, manually intensive grunt work should be done by your agent teammates, maybe your staff 1 human is already thinking about thinking like a senior one and starting to direct agents to do the manual work. so instead of having four or five staff on an audit, you might have one super powered staff, one senior who also directs, manages. but theoretically the senior could have more of a manager capacity. maybe the senior managers, instead of spending too much time on client work, they’re actually selling work to clients, building relationships with clients. so you can see how the servicing equation can really shift right. instead of being a bottom heavy pyramid, you have a leaner audit team with agents that help with the junior work, and then you offer everyone, people at every level, a more advanced career. so that servicing equation, we believe, is already starting to happen. we see a handful of firms that are really leaning into a different servicing equation. so we’re very optimistic about that happening.

liz farr

yeah, it’s it’s bringing the human in to do more of the human things, so that the managers, instead of frantically trying to review 25 audits, and you know, oversee all of that, they can be focused on building the relationships that lead to additional services, which leads to additional revenue for the firm, and which makes everybody happier, because not only do you have more money, but you have this other really cool stuff to do. that’s not just an audit, right?

jin chang

that’s right, it took the words out of my mouth and actually liz, there’s so many things that we say very similarly, including, what i talk about in the industry is ai enables humans to do things that are uniquely human. so yeah, we share a lot of the same kind of philosophies and lingo. i would say,

liz farr

yes. well, i’ve been steeped in this world for quite a while and thinking about transformation in the whole accounting profession for some time. so anyways, it has been just wonderful to have you on the podcast. jin, and you know, and i hope that listeners want to connect with you. so where is the best place to find you?

jin chang

thank you. liz, it’s been such a joy having our conversation, and i’m sure the first of many fun conversations to come, but yeah, for the listeners here, look we at fieldguide. want to be the leading ai partner for the cpa industry, and we have solutions, but also a consultative approach to meet you where you are in your ai transformation journey. there’s a ton of noise out there, more noise than we can all handle. so our goal is to help you cut through that noise with a responsible framework approach. and at the end of the day, we just want to be helpful advisors consultants in that journey. you can find more information about us at fieldguide.io and one of our consultants or account reps can make sure to be a great thought partner in your journey.

chang: killing saly, one agent at a time | the disruptors

this is a preview. the complete 1-hour video episode, with commentary and transcript, is first available exclusively to pro members | go pro here

sponsored by tax season readiness: practical steps for a smoother busy season | see today’s special offer

subscribe to 卡塔尔世界杯常规比赛时间 podcasts anywhere: apple, google/youtube, spotify, iheart, deezer, amazon music, audible, player fm, audacy, rss

sponsored by tax season readiness: practical steps for a smoother busy season | see today’s special offer

related posts

kless: profit is a result. flourishing is the purpose | the disruptors

hartman: upwork to “40 under 40” in 3 years | the disruptors

vilms: the power of people in a tech-driven world | the disruptors

brolin: blueprint for empathetic leadership | the disruptors

telka: transform fear into fuel | the disruptors

meyer: tax plans in 90 seconds? believe it | the disruptors

leave a reply