The Real AI Race in Tax & Accounting Isn’t About Prompts

Welcome to the Growthminded Accountant podcast where our experts will share best practices on running your firm in the digital age.

This podcast is brought to you by  Pro.

Let's get started.

Hey everyone and welcome back to another episode of the Growthminded Accountant podcast.

My name is Lee Reams.

I am the founder and CEO of CountingWorks and our sister company taxbuzz.com.

And today's episode, I want to kick off something a little bit different.

Um, and then over the next several months, I'm going to stay on this theme.

So, we're going to go uh we're doing basically a series of episodes focused entirely on AI, operational intelligence.

It's something I want you to understand why I think that is the future of tax and accounting firms and what I believe is actually happening underneath all the noise right now.

So you hear it, you read it on Reddit, you see the articles on CNBC, you see the influencers, and honestly, I think a lot of the industry is focusing perhaps on the wrong things.

So right now, everywhere you look, the conversation is about prompts.

Um, we were talking about how to write prompts a year and a half, two years ago, right? about AI tools, AI agents, automations, uh chat GPT, Claude, Gemini, you know, you're starting to hear terms like MCP servers, uh you're hearing custom GPTs, workflow hacks, right?

And basically, how do I save time?

And look, I think some of that stuff really matters and a lot of it is genuinely exciting, but I also think we're very early in understanding what AI is actually going to become inside tax and accounting firms.

And more importantly, I think a lot of firms are underestimating how much this is going to reshape the actual operating model of the profession itself.

So, in this episode, uh, this is really what I call a cornerstone episode.

We're going to start by building a foundation. uh and because before we can talk about other things AI agents workflow automation advisory systems and all the futuristic stuff that everyone's excited about we first need to understand what problem AI is actually solving and I don't think the answer is AI writes emails faster okay I think the answer is much bigger than that so let me start with this right now most tax and accounting firms are drowning drowning drowning drowning in operational friction not lack of intelligence not lack of talent and I'll explain why but more operational friction.

Okay.

So, honestly, I think that's more important to say out loud because for years, technology was supposed to simplify firms, but for a lot of firms right now, technology has actually created more operational complexity, more software.

I, you know, we we have clients, we talk to them all the time.

You know, we had a firm come in, I think they had 13 software products in their tech stack, and I'm like, half of these are really complex um software products, right?

So, how do you train everybody?

Who knows how to use the software? who's using it correctly, how does the software talk back and forth, you got more logins, you got more disconnected systems, you know, more tabs open, you know, more places where information lives, right?

And they're all in their own little silos, right?

So, you know, more things staff members are expected to remember.

I think that's really important that creates friction just in it creates stress and just doing your job, right?

So, now on top of all this, firms are also being told they need to become AI experts overnight.

And yes, that is what you're hearing on social media.

So at the same time the profession is already dealing with staffing shortages.

You got the boomer retirements declining accounting graduates.

You have rising client expectations.

I think that is a big deal.

You got software sprawl and massive operational pressure you know during each of your busy seasons right that is really really tough.

It's like how do we get all this work done in a compressed time zone.

So that's the environment firms are trying to navigate right now.

And I think one of the most dangerous things happening in the market right now is that there's a lot of AI theater.

Uh and I see a lot of people affected by it.

People that I respect are like they're buying in.

You know, I'm like, well, it's not exactly as easy as you think.

Okay.

So, a lot of that, you know, certainty being sold uh where I don't think certainty doesn't I don't think it exists yet.

Okay.

So, people promising fully autonomous firms.

Um, and there's a huge danger here because firms can become very cold very quickly.

And if you're in a relationship business and you don't have a relationship anymore, what do you think happens?

What is what is at risk?

It's your it's your clients, right?

So AI employees, one-click advisory, firms running themselves, agents replacing entire teams, uh, in fully automation and tax prep overnight, right?

So that's what's being promised out there.

And I want to be very careful here because some of these technologies are real.

We're using them.

We're developing them.

But some of them absolutely will reshape the profession over time.

Okay.

But I also think there's a huge gap right now between demos, you know, the ability to create a prototype, right, and operational reality.

And I think that's something a lot of firms are missing right now.

So, I'm going to go into it, especially in a profession built around, think about it, accuracy, compliance, liability, trust, and of course, judgment, and more importantly, context.

And I think context is something people don't understand when they're using a lot of these AI tools.

So, tax and accounting is not social media, okay?

It's not ecommerce.

If Spotify recommends the wrong s song, right, nobody cares.

If AI AI gets a tax position 5% wrong, that becomes penalties, audits, liability or even damaged client trust or maybe even missed uh opportunities to save.

Right?

So, the tolerance for error in this profession is fundamentally different and I think firms need more voices helping them think clearly about what AI actually is today.

Okay?

Versus what people are projecting it might become someday. uh you know because right now most firms aren't struggling with intelligence, they're struggling with operational chaos.

So think about you know what firms deal with every single day.

They have disconnected software, emails everywhere.

Someone sending you a Dropbox link versus your client portal versus a Google Drive document versus your uh you know Gmail, you know, versus WhatsApp.

Who knows, right?

Um you got workflow bottlenecks.

I think everyone can uh see how that works. whether a junior person then waiting for a senior person to review something.

Uh staff trying to remember where clients stand I think is almost impossible today.

Uh you have onboarding gaps.

Um you know a lot of people then don't even do engagement letters right or they're not using proposals.

They have scope creep everywhere right or they miss advisory opportunities.

Uh inconsistent follow-up uh communication breakdowns tribal knowledge I think is a huge issue.

Uh and this is creates operational chaos during the busy season, right?

And more importantly, this is constant context switching.

And honestly, most firms have operated this way for years and years.

The owner knows everything.

The senior admin knows where things live.

The tax, you know, manager might remember the client history.

The staff kind of understands the workflow.

And a lot of firms are operating on institutional memory and human effort holding the entire machine together.

And for a long time, you know, that worked or at least I would say it worked well enough.

But AI is exposing something really important right now.

It's exposing how fragile many operational systems inside firms actually are.

I think that whole concept of I'm going to do it the same way I did it last year is starting to now it's becoming a little bit of a red flag and people are paying attention.

So AI is accelerating the speed of work dramatically.

Tasks that took hours now take minutes.

Well, wait a second.

How does that affect your income if you're billing everything by, you know, your time, right?

So, that's where it's starting to disrupt even your business models.

You know, how what is the value that you're you're uh providing your clients, right?

These are all things that I think are really important to understand.

Um, you know, processes required multiple multiple touch points can now happen almost automatically.

Client communication can scale. that whole what we see all the time in reviews specifically the onestar reviews my CPA my EA does not respond to me you know I send something in it's a black hole AI can solve that problem very quickly document analysis at scale right I mean we have technology we developed last year scans all of the client uploads organizes it in context summarizes it and now gives you opportunities to a to use AI to create tax plans tax strategies even to write respons responses to inbound messages, right?

Workflow coordination can scale, but here is the problem.

Um, faster chaos is still chaos.

So, I think and what you're seeing is a lot of software products are bolting AI tools on top of software that's already clunky, right?

And I think a lot of people need to start rethinking an AI first workflow and how things work.

So, um, I think that's the part many firms haven't fully processed yet.

And because right now a lot of professionals think the AI race is about who has the best prompts, who uses chat GBT best, who builds the coolest agents, right?

Or who automates the most tax uh tasks.

And I don't think that's the real race.

The real AI race in tax and accounting is operational intelligence.

That's something I brought up earlier.

This that's kind of basically the whole premise of what I'm doing with  is we're creating operational intelligence that now can be acted upon with the human in the loop.

So when I say operational intelligence, what I really mean is this.

A firm where the systems, your tech stack, whatever you want to call it, your AI actually understand that your clients, the workflow, the timing, the communication history, right?

The opportunities, the advisory opportunities, and then you know what needs to happen next.

So without everything relying entirely on human memory and that's a very different future than just AI writing content.

Okay.

So, because tax accounting firms are not content businesses, they're operational businesses built around trust.

Now, I would say timing, your client relationships, obviously, compliance, advisory, and institutional memory.

So, everything connects to everything else.

So, a tax strategy impacts bookkeeping, for example.

Bookkeeping impacts advisory.

Advisory impacts cash flow planning.

Right?

I think you get it.

I'll keep going though.

Communication impacts your retention, right?

How responsive am I?

Are my clients leaving me because they're upset with me?

Uh entity structure impacts tax planning.

Uh you know, QPS, I'm going to sell my company.

Well, if we had done this and waited 5 years, this would be your tax minimization, you know, uh opportunity.

Uh life cycle timing impacts recommendations.

You know, where are people?

Are they boomers?

Are they Gen Xers?

Are they millennials?

Are they Gen Z's, etc.

So all of that will kind of affect what is the planning that you're doing for clients and how do I now change my messaging even my tone and the way I communicate based on where people are and I think this is where a lot of the market is getting distracted right now people are falling in love with AI outputs and you see it often too people can be build a prototype and lovable and you know if you know how to use it and it's beautiful and it's fully functional but it's not it's not software it just kind of shows you what it could So the real uh value is not that output.

The real value is the intelligence layer sitting underneath the output.

Okay.

So that's the context.

Where do we have all this data?

Do we have demographic data about our client?

Do we have information of their latest tax returns, their financial statements sitting inside an AI tool securely?

Right?

But now the AI can be looking all that and putting that into your inputs when you're working with clients. uh with your workflows, the client history, the timing, the operational memory, right?

And the guardrails, it's really important to understand.

So, you know, people are kind of going rogue and I'm creating agents, but without the proper guardrails, you can ask the same task 10 times and get 10 different outputs.

And people are seeing that and I'll just go back to a tool like lovable for designer programmers where you could be working in a a design, for example, and then the AI tool just goes rogue and starts designing something different. and you're like, whoa.

And the problem here is they're not the guardrails.

That's really important to understand.

You want to be consistent.

You want all your technology to if it's doing a tax plan, you want them all to look alike.

You want them to use the same methods to get there.

Each client is different, but you want the output to be framed kind of consistently.

Um, so that's where the future gets interesting because everything firms um I was going to say every eventually firms are going to start to realize hey AI without this operational intelligence creates fragmentation.

So and I've seen it already.

I'm using Zapier and I'm going to use this tool and I'm going to use that tool and what does it do? it creates disconnected workflows, uh, inconsistent outputs, and then trying to connect all these MCPs that are external that aren't designed specifically for that tool, you know, it creates a lot of problems, and it creates scattered client context.

Uh, and I think operational dependency on whoever built the automation.

So now all of a sudden, you have an AI CTO inside your firm.

I don't consider that scalable and I don't think that's even sellable, you know.

Oh, here let me see your tech stack.

Oh, we use cloud for this. use chat GBT for this it's my personal account right then where's the CRM where's the security how do I know my client data is secure you know I can go on and on right and that becomes what I consider fragile infrastructure so you know we're already starting to see it happen uh you know I just I gave that example I you know I talked to a firm using six different AI tools you know one for their meeting notes one for email drafting one for workflows one for proposal generation one for document summaries one for chat chatbot support and eventually the owner finally admitted I'm spending more time managing the system than actually running my firm.

And I think that's the danger nobody talks about or at least doesn't recognize or at least kind of admit, right?

They're just like, well, I'm trying to save time or I'm trying to be smarter than others, but they realize very quickly and we're seeing it in developers.

We talked to developers who clients have cut back on programming staff.

Well, I can use, you know, AI tools to write software and all of a sudden they're all coming back with their tail between their legs going, well, it's not as easy as I thought, and I need help.

Can you clean this up?

So, you know, a lot of firms are accidentally becoming the CTO of software stacks they never intended to build, and now they're managing prompts, integrations, workflow logic, automations, APIs, MCPs, training, troubleshooting, and and what really is this becomes operational fragmentation.

So that's not what most tax professionals signed up for.

And honestly, most firms today are operating with what I call human middleware, right?

So meaning humans are manually acting as the integration layer between all these disconnected systems every single day.

So the admin remembers what happened, the manager remembers the client history, the owner remembers perhaps the opportunity, the advisor opportunity, right? staff members remember the next step, but humans are carrying operational memory for the entire firm.

And that model becomes incredibly fragile as firms scale.

So here's another I'll call it a pretty simple example.

Think about how many times a client calls your office and somebody says, "Let me look into that." So, the team starts digging through your emails, maybe your CRM notes, the tax software, the last tax return, perhaps financial statements, perhaps old conversations in the in your client portal or via email, right?

Uh perhaps they're looking at spreadsheets, you know, and staff members are trying to reconstruct all this context manually, right?

So, that's not a tax problem.

That's an operational intelligence problem.

I think um you know, think about the advisory opportunities you're missing.

If you had all that data all together that could instantly be analyzed, you know, think about it.

You know, right now you're only getting advisory opportunities for clients when someone remembers, oh, I you know what, they said something like that in the Zoom meeting or someone notices something.

Maybe there's a trend.

Oh, they this guy just they just hired five more people.

Oh, maybe we should have been consulting them or are the employees independent contractors, right?

Or the cost of each person and what is that going to do to cash flow?

We need to rebudget, right?

Um, you know, or or someone doesn't have enough time anymore, right?

So, what happens when operational systems can continually service uh surface these things?

So, uh entity change opportunities, right? estimated tax issues perhaps retirement planning triggers uh cash flow risks or even a tax strategy opportunity automatically that is the beauty of what AI can do okay so that changes the economics of the entire firm and I've I've called it you know solopreneurs are basically super people at this point you know small firms boutique firms can compete with the big firms if you're using the right tech stack and operational intelligence so I do realize all of This feels a little bit overwhelming and I don't think you are behind.

This industry specifically the tax and accounting industry is always a few years behind but this time this disruption is so severe that you have no choice but to jump in here right uh but I think the industry as a whole we're still incredibly early in this transformation.

So most firms are still trying to figure out, you know, what a AI even means operational and you know, and then you have the people that are so anti, you know, I will never use this.

I, you know, what are you talking about?

The problem here is their competitors are going to be and their clients are going to be getting better client experiences from those people that do.

So if you feel like everybody else understands this, everybody else has it figured out, or you're somehow late to the game, I'm going to tell you, you are not.

Um I quite honestly feel there's so much noise, there's so many false prophets out there that u you know all of us are learning this in real time. you know, there's promises and even I'm going to go back to uh both Antropic and OpenAI CEOs have kind of started to backtrack some of the things they were saying three months ago, 6 months ago, and they're starting to realize, well, wait a second, you know, and maybe I was doing that to raise money.

Maybe I was doing that for Wall Street.

But the reality is, you know, that's bad for the economy, that's bad for the world, and it's not true.

You know, that's kind of like pie in the sky kind of thinking.

So the future though is orchestration.

Okay.

So connected operational intelligence.

So system that's intelligently coordinate a bunch of tasks that you do.

So your workflows and what I mean by workflows this could be internal workflows and an engagement or it could be the onboarding process.

So, the idea of a proposal once signed kicking off an engagement letter that's once signed uh kicks off an intake form that once received creates to-do lists for staff and assigns them, right?

Um, you know, coordinate how AI models get contacts and can communicate to each other, right?

Uh, the communication with your clients.

I talked about onboarding advisory opportunities surfacing opportunities for you 24/7 while you sleep, right? and all of a sudden come in and today on your dashboard it says hey Lee you have six opportunities you know client A just had some children you should talk to about 529 planning right um this will affect retention as well and more importantly client relationships do not be a cold firm do not use technology and lose the human element uh and that's going to be your downfall if you're not so think of this as inside one centralized operational layer that's where the industry is heading and honestly I think this is one of the reasons we've been thinking about max our AI you know knowledge base basically and ac very different internally we do not believe the future of tax accounting is another chatbot another AI feature or another disconnected app we believe the future belongs to firms with connected operational intelligent systems systems capable of helping firms scale relationships um create their own narratives You're starting to see even in the SEO marketing space.

I know Neil Patel had a um a blog or a vlog I saw yesterday and it's like, "Oh, you need to think about narrative.

I've been talking about narratives for two years now.

Um and creating your own narrative, your own personality, telling your story uh because this is now integrated into all of your AI tools.

So when you write a blog article or you write a proposal, it has your firm personality in it." Okay?

And more importantly, when AI tools are looking at brands or clients or prospects are having conversations about topics, if you're an expert in that and you have a narrative in your firm on your website, on your social media posts, when you you post on LinkedIn, that information now is context that you are now part of that conversation.

Right?

I think that systems also will reduce operational drag. this whole, you know, knowledge here, knowledge there.

Uh, obviously the uncover of advisory opportunities, I think, is one of the biggest ones and doing it on autopilot.

Being able to service your clients in a way you never could do before.

And then what this will do is create consistency across the entire client life cycle.

Um, you know, because ultimately the firms that win, I think, over the next decade will not simply be the firms using AI. there'll be those firms who operations have become more intelligent.

So, I think that is the real opportunity.

That was kind of the the whole point of this podcast.

We're not here, we're not talking about replacing humans.

We're talking about amplifying them, okay?

Not adding more chaos.

We're talking about reducing friction, not more disconnected tools where you have 12, you know, software products in your stack and you got Zapier trying to connect them all.

Um, we're talking about smarter systems.

So, uh, if you're listening to this and wondering, okay, Lee, you know, what should I actually be doing right now?

My advice would be first, stop chasing every new AI tool, stop trying to be your own virtual CTO.

The firms winning long-term are probably not going to be the firms with the most software.

Okay?

I think second, start identifying where operational friction exists inside your firm today.

So, where are your workflows breaking?

Where is the communication getting lost?

Where are the opportunities getting missed?

You know, where is the context disappearing?

Right?

You know, I I even talked about, you know, note-taking from a Zoom meeting interview with your client again if the right disclosures are done.

But now you have all that context sitting there.

It's not in a note that a, you know, a junior person was doing during a meeting, right?

And more importantly, staff wasting all their time rebuilding information manually.

That is where operational intelligence actually matters.

That's where we're really moving the needle.

And third, I think you need to focus on systems that can start centralizing this communication, your workflows, the client context, right?

Operational visibility so you know exactly where everyone is and institutional knowledge.

So wait a second, everything doesn't just live in one, you know, staff member, you know, because the value uh of future firms will increasingly come from how intelligently the firm operates. um not just how good a tax return is, right?

Uh does that make sense?

So, you know, the intelligently the firm operates operationally.

So, what does that mean?

Well, if I've now put together a system where we were we had all these human gaps before and now the AI is able to do this at scale, I am now, you know, it's not just again the quality, you know, putting out a tax return.

I'm now a full-on adviser, right?

I'm having AI assist me in identifying all these opportunities for my client to help them get a better financial outcome.

So, you know, I know there's a lot of noise.

I know this is a lot of like top level stuff.

Uh but don't panic, okay?

We are incredibly early in this transformation.

You know, I genuinely believe this industry, this profession is entering one of the biggest operational shifts that I have ever experienced.

Uh and I am excited and scared about it to be honest.

Uh but I'm more excited because I think the firms the biggest fear I have and the reason I guess I have fear is not enough firms are understanding what's happening and I think there's so many misconceptions out there.

You know they're forgetting about their relationships.

Uh the relationship layer should be your primary thing having a unique narrative.

I I talk to clients and prospects about this.

You know we do podcasts, we do webinars about it and people still don't I don't care.

Well, I'm just going to be generic.

I'm just going to put out this cookie cutter website.

Why?

Why do I care?

My website doesn't get any more traffic.

Well, no kidding.

It doesn't get any more traffic because the entire search ecosystem has changed.

You are not optimizing for geo type of optimization where the chat bots will be able to talk about you, right?

And then more importantly, if you can figure out how all this operational intelligence can, you know, start moving the needle both from efficiency, saving time, and more importantly finding these advisory opportunities, I think you're going to have an enormous advantage over the next several years.

So, uh, what I'm going to do is build upon this.

This is kind of my beginning pitch about this.

I over the next several episodes, we're going to go a lot deeper.

We're going to talk about why AI without context hits the ceiling.

I think this is really important.

Um, I listened to a podcast.

It was the all-in podcast.

The Salesforce CEO was on it and he talked about that's like the big thing people don't get.

You know, AI by itself without serious context and guard rails goes rogue and all the all over the place and it's inconsistent.

It can make mistakes.

Um, but AI with context and guard rails can do incredible things.

We'll talk about why uh 95% accurate is could still be 100% wrong in taxes.

Um obviously in accounting, right?

That's a that's a comment Theres Tucker from Black Line, one of our board members always says to me, you know, it's like, yeah, do you realize that, you know, 90% right, 97% right is still 100% wrong in our industry?

And I think that's really important and maybe some people are playing things a little too loose when they and a lot of people who are trying to build tools in our vertical don't understand that.

I think we'll also talk about the hidden risk of DIY AI stacks.

Uh I'll talk all the issues there.

Um multi- aent systems.

We'll talk about what is operational intelligence in more details.

We'll talk about how you can use AI to find these advisory opportunities.

We're going to talk about how this affects your firm valuation and what you're selling, right?

And what the modern operating model of tax and accounting firms is probably going to look like over the next 5 to 10 years.

So, uh, I believe the industry is entering one of the biggest transformations it has ever seen.

So, I I'm not discounting this.

I just want to kind of bring people back from the the the ledge, so to speak.

So, you know, we're incredibly early.

Um, so hopefully this is kind of getting you to think about AI a little bit differently.

Hopefully you enjoyed this episode.

So, make sure you subscribe, share with other tax and accounting professionals, and stay tuned because we're just getting started, right?

So again, thanks for listening to the Growth Money accountant podcast.

And remember, the firms that win in the AI era won't necessarily be the firms with the most tools.

Uh they'll be the firm with the clearest operational intelligence.

So again, thank you for listening and we'll see you next

‍