I keep hearing firm leaders ask for innovation while expecting every billable hour, every deadline and every old process to stay exactly where it is. They’ve bought the licenses and sent everyone to training. Now they want people to transform the firm somewhere between finishing the client work and picking up the kids.

Then somebody spends two hours experimenting, and a partner asks why their utilization is down.

Well, there it is. Your AI strategy just ran into your scorecard. And we all know which one wins when compensation comes up.

Congratulations on protecting the old way

Imagine someone figures out how to use AI to prepare for client meetings. They test it, get some garbage back, add better context and build something useful. Now they’re asking better questions and spotting issues they might have missed.

Meanwhile, the person next to them follows the same process they’ve used for years and logs every hour.

One person made the work better. The other logged more time. Guess who gets the better review?

We love to say accountants resist change. But if changing how they work hurts their performance review, why would they risk it? Your people understand your priorities perfectly. Be innovative after you’ve done everything that counts.

You bought a license not a skill

At the 2026 ADP Accountant Connect AI Forum, I said firms need to give people time to experiment. Whether that’s two hours a week or four hours a month, pick a starting point and actually make room for it.

Tal Goldhammer, founder and CEO of Find the Tailwind, put it simply when he said, “Access is not capability.”

Exactly! Somebody has to learn what to do with the tool. That means trying something that doesn’t work, figuring out what went wrong and recognizing when an impressive answer is nonsense. Some of that will take longer than doing the task the familiar way. Of course it will. They’re learning!

And if you give someone two hours to learn but leave every production target untouched, you’ve given them homework. Call it whatever you want in the partner meeting. Your staff knows what it means for their evening.

Partners, you can change the rules

You’re the partner. You can change the target. Your staff can’t. So why are they the ones expected to figure this out?

Put learning time in the workload. Decide what gets moved, reassigned or stopped. Make sure managers understand how that time counts in performance reviews. If one partner says “go experiment” and another says, “make up the hours,” spare me the innovation speech.

And open the tool yourself. Try something you don’t already know how to do. Let your team see you get stuck. You want people willing to work differently? Show them what that looks like from the partner’s chair.

Laci Buzzelli, senior vice president and general manager of small business services at ADP, said success will be defined by “the organizations that do the best job helping their people adopt it, build confidence in it and create new value.”

New value? Can we spend a little more time on that part?

Give someone a recurring problem to explore, then have them show the team what happened. Recognize the person who helps five colleagues work better. Ask about the client need someone uncovered, even if it hasn’t become an engagement yet.

That’s the part that excites me. People who know where the work gets stuck finally getting room to fix it. Accountants discovering that they can help clients in ways they hadn’t considered. Give them a reason to build that firm.

And when someone makes an old process unnecessary, celebrate it. Don’t hand them a utilization problem.

They learned the rules from you. Now give them permission to break a few.

I’ve always been someone who likes to explore ideas, ask questions and consider different possibilities. And now, I have access to technology that will happily entertain every single one of those thoughts for as long as I want.

I’m starting to realize that might actually be a problem… Maybe all the times my husband has told me to stop re-explaining something that’s going on at work were actually helpful? Don’t tell him that!

AI Doesn’t Know When I’m Done

I use AI constantly to build companies, develop products, challenge ideas and solve problems. It’s become a natural part of how I work, but I’ve noticed lately how easy it is to get carried away. I’ll ask a question, get a useful answer and immediately think of something else I want to explore.

Then another question comes up, and another and another, until I’ve spent an hour analyzing something I probably could’ve decided 45 minutes ago.

Before AI, we had natural limits on how much time we could spend thinking through a problem. Eventually, you ran out of people to ask, got tired of searching Google, or your husband reminded you it’s time to go to bed.

There was friction in the process, and I think some of that friction is good for us. Just because AI never runs out of information doesn’t mean it’s actually productive to spend your time exploring unlimited possibilities.

Would Another Scenario Change Your Advice

Think about an accountant advising a client who’s considering expanding their business. You can use AI to analyze the financial implications, compare scenarios and identify risks. That’s valuable information to bring into the conversation.

But how many scenarios do you really need to explore before you can make a recommendation? At what point are you spending more time analyzing possibilities that probably won’t change your advice?

Put Some Boundaries Back

I think we need to start recreating some of the boundaries that AI has taken away. Here are a few things I’m trying that help me be more intentional about when I use it.

1. Set a goal before you start. Decide what you’re trying to figure out and what information you need to move forward. Ask AI to identify the three most important questions you should answer, then focus on those.

2. Ask if another answer would change your decision. Before going deeper, consider whether more information would actually affect your recommendation or next step. If the answer is no, you’re probably ready to move on.

3. Give yourself a deadline. Set a time limit for exploring a problem. When it’s up, close the chat, sleep on it, or talk to another person before making your decision.

AI will always have another answer. I’m learning that I don’t always need to hear it. At some point, you have to trust your judgment enough to stop asking and make the call.

Timesheets tend to favor what’s easy to record, easy to defend and easy to report upward. But the people creating the most value in a firm aren’t always the ones with the cleanest timesheet, the highest visible output or the neatest-looking utilization story.

In fact, some of the most valuable people in a firm are often the hardest to measure in a traditional system because what they contribute shows up in relationships, judgment, coaching, problem prevention and keeping work moving. The traditional system of time collection simply does not work in this model.

The Wrong Scorecard Rewards the Wrong Behavior

This is where the old model starts to break down. When firms rely too heavily on submitted time, they end up rewarding activity over impact. They give more credit to what’s visible in the system than to what’s actually valuable in the work. That can create all kinds of distortion.

The person who protects a client relationship, unblocks a struggling team or prevents a project from going sideways may create more value than the person who simply logs the cleanest set of hours, but the system doesn’t always know how to recognize that.

And that’s not a small issue. It shapes promotions, performance evaluations, staffing decisions and who leadership sees as indispensable. If the data source is incomplete, then the talent story built on top of it will be incomplete as well.

Relationship Builders Are Often Undervalued

Some of the most important work in a firm doesn’t look dramatic on paper. It actually looks more like:

• Checking in with a client before they panic

• Mentoring a newer team member so a deliverable doesn’t fall apart later

• Being the person everyone trusts when something gets messy or is unclear

That work is real, and it creates value. But timesheet-based systems often flatten it into background noise. Laurel’s State of Work Report reinforces just how much of senior professionals’ time gets spent in high-touch, people-heavy environments:

• Senior professionals spend up to 37.6% of their time in meetings

• 46% of meeting time involves multitasking

That tells us two things at once. First, senior people are carrying a heavy load of relational and coordination work. And secondly, much of that work is happening in conditions that divide attention and hide strain.

Burnout Doesn’t Show Up Until It’s Already Expensive

This is another place where bad measurement quietly hurts firms. Burnout rarely arrives as a neat metric. It shows up first as after-hours work, context switching, constant meeting overload, invisible cleanup and senior people stepping into work they shouldn’t have to do. By the time all of this becomes obvious in attrition, disengagement or declining client experience, the damage is already done.

Based on the Laurel report, working outside standard business hours is common, and a significant share of deep work gets pushed into nights and weekends.

This pattern matters because work that occurs outside of regular business hours can signal overload, focus debt and an environment where people are compensating for a broken system with personal time.

What makes all this more frustrating is that many leaders already have instincts about who’s truly valuable. They know who clients trust. They know who keeps the team grounded. They know who gets pulled into every fire drill because that person can actually solve the problem.

But instinct isn’t the same as evidence, and firms have spent years trying to run people decisions on data that doesn’t fully capture what those people actually do.

All of this can lead to tension inside the firm, which can be avoided with visibility-backed decisions instead of incomplete records and lagging assumptions. In other words, leaders shouldn’t have to choose between gut feel and bad data. They should be able to see the work well enough to validate what they already suspect and act on it earlier.

Better Measurement Changes Who is Rewarded

When firms start measuring the right work, the right people become easier to see. They can identify who is carrying too much invisible load, who is creating outsized value through relationships and judgment, where senior talent is being deployed poorly and where burnout risk is building before it becomes a retention problem. This doesn’t just improve fairness, but also improves staffing, coaching, succession planning and client service.

Right now, measurement is driving the wrong behaviors and rewarding activity over impact. But once firms can see the full picture of how work actually happens, they can stop confusing visibility with value and start building a system where the right people really do win. This, of course, requires firms to acknowledge longstanding system flaws in order to implement the right changes moving forward.

But how often are we asking AI to help us think?

Lots of professionals overlook one of the most interesting ways to use the technology. AI can be an incredibly useful thinking partner, especially when you’re working through a complicated decision, preparing for a difficult conversation or trying to figure out why you’re stuck on something.

You don’t have to be particularly technical to use it this way. Honestly, I’ve been doing it intuitively since the first few AI tools hit the market a few years ago. It’s become second nature for me to run my thoughts by an AI agent just to get a second opinion. I think it really just comes down to being curious, creative and willing to have a conversation.

Think about how you work through a difficult problem today. Maybe you call a colleague, talk it through with your partner or spend an hour walking around your office trying to organize your thoughts. Sometimes you already know the answer, but you need someone to ask the right question to help you get there.

AI can help with that process, and here are some ways that you can start experimenting.

1. Use AI to challenge your decisions

Let’s say you’re trying to decide whether to hire another accountant or invest in technology that could take some work off your team’s plate. You probably have a pretty good idea of the advantages and disadvantages of each option. But you also have your own biases, assumptions and experiences influencing how you’re thinking about the decision.

Try asking AI: “I’m deciding between hiring another accountant and investing in automation. What are the pros and cons of each, and what factors might I be overlooking?”

Then give it some context about your firm. Tell it about your workload, your team, your budget and what you’re hoping to accomplish. You can even ask it to challenge your preferred option. What assumptions am I making? What could go wrong? What information would change my decision?

The more context you provide, the more useful the conversation becomes.

2. Practice your client conversations

This is one of the easiest places to start. Imagine you have a client who’s frustrated about their fees. You need to explain the value you’re delivering, but you’re worried the conversation could get defensive. Tell AI what’s happening and ask how you might frame the conversation. Then have it play the client.

Let it push back on your explanation. Ask it to identify where you’re using too much jargon or answering questions the client hasn’t actually asked. Practice explaining your value until you can do it clearly and confidently.

You can also use AI to prepare better questions before a meeting. If a client is struggling with cash flow, ask what you should understand about their business before recommending a solution.

The goal is to walk into the real conversation better prepared to listen, understand and respond.

3. Figure out why you’re stuck

We don’t talk enough about how much time professionals spend overthinking. Maybe you’ve been putting off a difficult conversation or delaying a decision about your firm. Maybe you have a great idea but can’t seem to get started. Instead of asking AI to make the decision, tell it what’s happening.

Try asking AI: “I’ve been putting off this decision for weeks. Help me figure out what might be holding me back.”

Ask it to question your assumptions or walk you through the possible consequences of each choice. Maybe you’re worried about making the wrong decision. Maybe you don’t have enough information. Maybe you already know what you want to do, but you’re uncomfortable with what comes next.

Sometimes getting unstuck starts with understanding why you’re hesitating.

Give Yourself Permission to Think Differently

When I first started using AI this way, I wasn’t following a framework or trying to write the perfect prompt. I was curious. I had questions, ideas and problems I wanted to talk through. So I started having conversations with it.

The value isn’t always in the first answer. Sometimes it’s in the question AI asks me next, the assumption it challenges or the idea I hadn’t considered. I still have to decide what makes sense and what to do with it. But the conversation helps me get there.

As accountants, we spend a lot of time looking for the right answer. What if we spent more time exploring the questions?

The next time you’re wrestling with a decision, preparing for a conversation or feeling stuck, try bringing AI into the process. Tell it what you’re thinking. Ask it to push back. Keep the conversation going. You might be surprised by where your thinking takes you.

Accounting firms have been talking about the move from compliance to advisory for years. The usual conversation centers on services. What new advisory offerings should we build? How should we package them? How do we price them? How do we market these services to clients? But new buyer research from ClearlyRated suggests firms may be looking at the problem from the wrong direction.

Clients do not need to be convinced that they want advice. They need their accountants to start giving it. According to ClearlyRated’s 2026 Accounting Buyer Insights Report, 91% of buyers want their accounting firm to function as a proactive business advisor rather than simply a compliance provider. And 90% say they would pay more for proactive identification of risks and opportunities.

Sounds like great news for firms, right? It would be, except only 27% say their accounting firm proactively brings advisory topics to the table. If that wasn’t big enough of a blow to accountants, 35% say they — the client — typically initiate those conversations. Gulp!

That is a very different kind of advisory gap. One that starts at the conversation level.

Clients Already Want the Next Conversation

There has long been an assumption that accounting firms need to persuade clients to see them as advisors. ClearlyRated’s findings suggest the demand is already there. Not only do clients want proactive advice, but 94% say the advisory services they receive have had a measurable positive impact on their business. And 90% are willing to pay more for proactive risk and opportunity identification.

The opportunity is not theoretical. Buyers are telling firms directly what they want, but there is a significant difference between having an advisory service offering and behaving like an advisor. A firm can add “business advisory” to its website tomorrow. It can create packages, establish pricing and train everyone on the technical components of the service. None of that guarantees that an accountant sitting across from a client will recognize an opportunity and say, “There is something else we should talk about.” That moment is where the transition really happens.

Compliance tends to begin with a question that already exists, like how something is treated for taxes, what is required by regulations or whether a number is correct. Proactive advisory requires something different. The questions that get into advisory opportunities are around:

• What is changing in the business?

• What are you trying to accomplish?

• What might get in the way?

• What are we seeing that you may not be seeing yet?

• What should we be talking about before it becomes a problem?

The accountant has to move beyond answering the question in front of them and start discovering the question that has not yet been asked.

Proactivity Drives the Relationship

ClearlyRated found an interesting relationship between who starts the advisory conversation and client satisfaction. When the accounting firm proactively initiates advisory conversations, Net Promoter Score (NPS) is 62. When the client is responsible for initiating those conversations, NPS falls to 48. We often treat proactive communication as a nice-to-have client service behavior. The data suggests it affects how clients experience the value of the relationship.

The same pattern shows up elsewhere in the report. For example, among buyers experiencing staffing-related service problems, reduced proactive communication produced the steepest decline in satisfaction. NPS among those buyers fell to 27, compared with an overall survey NPS of 55.

Accountants tend to think about client service in terms of deadlines, responsiveness and accuracy. All of those matter, but clients also notice what you bring to them before they ask. That is where advisory value becomes visible.

You Cannot Market Your Way Into Being an Advisor

Thirty-one percent of buyers cite a trust gap when considering additional advisory services. They trust their accounting firm for compliance, but not necessarily for strategy. ClearlyRated notes that this is not something pricing or marketing alone can fix. Firms have to demonstrate strategic thinking and build credibility.

That distinction is important. You cannot declare yourself a trusted advisor. You have to behave like one. Trust grows when:

• Someone asks a question that makes a client think differently.

• They connect two pieces of information the client had not connected.

• They challenge an assumption.

• They identify a risk before it becomes urgent.

• They understand where the client wants to go instead of simply explaining what happened last quarter.

Those are conversation behaviors, and they are learnable. But they have to be developed intentionally.

For years, accounting firms have trained people exceptionally well to arrive at the right technical answer. That remains essential. The profession cannot abandon technical excellence, but the next level of value requires that accountants know what to do after they have the answer.

AI Does Not Automatically Create Advisors

Buyers are not broadly rejecting AI. Actually, the ClearlyRated survey shows that 77% of clients say AI has improved the accuracy and quality of accounting work, and 81% agree that automation has made their firm faster without sacrificing quality. At the same time, 83% prefer human judgment over AI-generated outputs for final deliverables, and 86% say they would hold the accounting firm responsible for AI-related errors.

That tells us something about the future role of the accountant. AI can increasingly gather information, process documents, conduct research, identify patterns and produce outputs. Those capabilities can free enormous amounts of human capacity. But freeing capacity does not determine what people will do with it. A firm can automate 20% of someone’s workload and still have that person spend the newly available time doing more compliance work. The shift to advisory does not happen automatically because technology becomes more capable.

While AI can create capacity for advisory, it cannot automatically create advisors. Firms still have to develop the human capabilities required to use that capacity differently. That means curiosity, strategic thinking, empathy, challenge, listening, diagnosing problems, connecting business issues, understanding a client’s vision, creating action and so much more. These are sometimes dismissed as “soft skills.” Yet, there is nothing soft about them when nearly all clients say they will pay more for the outcome.

Stop Telling Accountants to “Be More Advisory”

Firm leaders can tell their people they need to be more consultative. Ask better questions. Spend more time with clients. Bring ideas to the table. Think strategically. Those are reasonable expectations, but they are not a development system.

Imagine telling a first-year accountant, “You need to improve your tax skills,” and then providing no technical training, no review and no feedback. You would never do that, but accountants regularly tell people to become better advisors without giving them a meaningful way to practice the behaviors advisory requires.

If proactive conversation is becoming a core business competency, firms need a way to develop it like one. That means looking at real conversations and asking:

• Did we understand the client’s challenge?

• Did we ask enough questions before recommending a solution?

• Did we connect the conversation to the client’s broader goals?

• Did we challenge their thinking when appropriate?

• Did we identify a next action?

• Did we bring insight to the conversation or simply respond to what the client asked?

• Did the client leave knowing something, or thinking about something, they had not considered before?

This is the thinking behind what we are building at XcelLabs with Navi. The goal is to build the behaviors that make stronger advisory conversations visible, measurable and coachable. Because “be more proactive” is not useful feedback. It’s better to show someone where an actual client conversation became reactive and coach them on what they could do differently next time.

The Revenue Opportunity May Already Be Sitting in the Client Base

ClearlyRated reports that 95% of clients intend to continue working with their existing accounting firm, 78% are likely to purchase additional accounting services in the next 12 months and 50% expect their spending with their firm to increase. Those numbers should change the way firms think about growth.

There is enormous emphasis on lead generation, new business development and winning new logos. Those things will always matter, but there is an equally important growth opportunity sitting inside conversations firms are already having every day. The client already trusts you, wants more advice and is willing to spend more. Does someone in the firm recognize the opportunity and start the conversation?

The move from compliance to advisory will require new technology, new service models, different pricing and changes to talent development. But it will not happen because a firm launches an advisory practice. It will happen one client conversation at a time. And if firms want to know whether they are really becoming more advisory, that may be the first place they should start measuring.

I built a firm. I advise accounting firms and technology companies. I speak for a living. I literally teach accountants how to have better advisory conversations.

And now… My own software is calling me out! It’s got a lot of nerve.

Navi is an AI coaching platform we built to analyze professional conversations and help accountants become better advisors.

I missed the opening. Then I went full CPA.

Recently, someone told me that they were struggling with AI tool overload and figuring out how another product would fit into their stack. I didn’t think twice about it, until Navi flagged it for me once I got off the call.

After analyzing our conversation, it told me: “You didn’t follow up on their stated pain point about tool overload and fitting into their stack.”

They’d handed me an opening to understand their problem better, and the Radical CPA herself blew right past it.

And that’s not the only time it’s happened. Another time, I was advising a group of non-CPAs who were building a new accounting firm. We started talking about professional liability insurance and workpapers, and I went full CPA on them. I knew exactly what I was talking about, so I didn’t stop to think whether or not they did!

Navi told me to pause, summarize what I was saying in one sentence, and ask what decision they were actually trying to make with the information.

Experience comes with blind spots.

It’s fascinating to get feedback like that even after you’ve been doing this kind of work for decades. Experience makes you better, but it also creates habits and blind spots.

When you know the material so well, you can forget what it sounds like to someone hearing it for the first time. You’ve handled the same problem enough times that you can start moving toward the solution before the other person has finished explaining what they actually need.

And it’s really hard to catch that in the moment!

Navi doesn’t need to know more about accounting than I do to make me a better advisor. It can pay attention to the conversation while I’m busy having it.

• Did I actually follow the pain point the client gave me?

• Did I ask another question?

• Did I explain something in a way they understood?

• Did I start solving the problem before I understood enough about it?

Those are little moments, but they change conversations.

Thirty years in. Still getting better.

It’s kind of obvious how AI can help younger professionals grow their budding knowledge, but it’s interesting to see what happens when we use it to examine the expertise we’ve already developed.

Thirty years of experience doesn’t mean I’m done getting better. It just means the things I need to work on are probably harder for me to see.

And apparently, sometimes I need my own software to tell me to shut up and ask another question!

We stand now at a pivotal moment for the tax profession. Already, AI is causing a fundamental shift in the way accounting firms and tax practitioners work. Nowhere is this more evident than in the area of tax research, where recent advances in generative AI are redefining what the modern tax tech stack looks like. With the momentum of generative AI and the complexity of the tax environment both increasing, Blue J and CPA.com sought to understand how AI is being used in accounting firms and how this technology is impacting firm performance.

Through a joint survey conducted by Blue J and CPA.com, we gathered input from over 150 tax practitioners, focusing on their outlook for AI-powered tax research solutions. Leveraging this survey data and user data from Blue J’s own generative AI solution for tax research, we’ve found that AI is rapidly transforming day-to-day workflows—delivering measurable advantages over legacy solutions.

Key Findings from the AI Tax Research Solution Outlook Report

1. AI adoption is on the rise across accounting workflows, especially for tax research, with 54.4% of respondents already using it to complete research in some form.

2. On average, respondents have access to three tax research tools; however, adoption of these tools remains low, with 81.3% of respondents turning to search engines for tax research, despite paying for a suite of tools.

3. Poor adoption of tax research tools is largely driven by poor user experience, with 42% of respondents saying legacy tax research tools are too time-consuming to use.

4. AI-powered tax research tools can solve these experience issues, helping tax practitioners arrive at better tax answers, faster. Respondents note that it saves them 2.94 hours/week across both research and drafting work.

5. Accelerated research is one of several factors contributing to the clear competitive advantage AI generates for firms, with 86% of respondents agreeing that AI-powered tax research generates an advantage for firms.

If there’s one common theme we uncovered through our research, it’s that AI-powered tax research is only going to continue building momentum. Which leaves firms with a choice: incorporate now or risk falling behind.

The State of Tax Research Solutions

When we examine the current state of tax research solutions within firms, it’s just that: a state of multiple solutions. On average, respondents to our outlook survey have access to three tax research solutions. However, they only report using two of these three solutions on a weekly or daily basis.

Despite having access to a selection of often costly tax research solutions, the vast majority (81.3%) of respondents also reported regularly using search engines, such as Google, to perform tax research. Due to the issues that can come with using search engines for high-stakes tax research—namely, that the answers rely on uncurated sources that are often outdated and difficult to verify—this data should be concerning, particularly when small mistakes can lead to costly client outcomes.

This data also brings up an important question. Why are firms that are paying for multiple tax research solutions still turning to search engines for their tax questions? To find the answer to this question, we need to turn to a particular class of solutions: legacy tax research tools.

Where Legacy Solutions Are Failing Firms

Of those respondents who report using legacy tools for tax research, very few actually use them. Legacy tools demonstrate consistently lower adoption rates compared to other classes of solutions, with weekly or daily usage rates ranging from 50% to 61%.

By comparison adoption rates for tax-specific AI tools are significantly higher, with 88% of those who have access to AI-powered tax research tools reporting weekly or daily usage.

Aside from the budget wasted on unused legacy tools, there’s another concern: tax practitioners turning to search engines instead. Firms pay for the trusted sources and editorial content these legacy tools offer, but access to that library means little if teams can’t easily find the sources they need. While some legacy tools are incorporating AI add-ons to counteract navigation issues, these often fall short on answer accuracy compared to AI-first tax research tools.

When trying to understand the adoption gap between legacy and AI-powered tax research solutions, respondents’ complaints about legacy solutions shed light on some possible causes.

Those using legacy tools for their tax research have three main complaints:

1. They’re too time-consuming to use (42%)

2. Too expensive (42%)

3. Difficult to navigate (37%)

Beyond speed, cost, and navigation problems, respondents also complain about a lack of functionality (27%), missing primary content (20%), and missing editorial content (18%). Together, these issues point towards a poor user experience that would naturally discourage regular use.

The AI Tax Research Advantage

Whereas low adoption and slow, cumbersome research are the hallmarks of traditional tax research tools, high adoption and accelerated research are what set AI-powered tax research solutions apart.

Legacy tax research tools typically rely on complex Boolean searches to source the correct documents. From there, tax practitioners must navigate these sources and manually stitch together a cohesive response. The complicated search operators required by each tool also mean that firms often use multiple solutions, with different people in the firm each having their “favorite” tool that they’ve come to know inside and out. However, none of these provides a straightforward way of arriving at a tax answer.

By contrast, generative AI solutions rely on conversational prompting, allowing tax practitioners to phrase their questions as if speaking to a colleague. Generative AI’s ease of use is apparent even during the initial stages of adoption. After logging into Blue J for the first time, users take an average of just 2.8 minutes to generate a comprehensive tax answer.

This ease of use doesn’t come at the cost of quality, as the responses these AI systems generate are also more cohesive than those offered by legacy solutions, weaving together multiple sources into a single, comprehensive response. For tax practitioners researching unfamiliar areas, these comprehensive responses can be invaluable for quickly gaining an overall understanding of a topic. For those working in more familiar areas, these answers can serve to validate their knowledge, ensuring it’s up-to-date and accurate.

Outside of answering questions on a broad array of tax topics, generative AI-based tax research solutions can also turn those tax answers into client-ready communications. For Blue J users, drafting represents a notable portion of their usage, with 18.8% of respondents to a Blue J user survey listing drafting as their primary use case for the application. Because it’s often highly time-consuming to translate complex tax concepts for clients, automating even just a portion of drafting work can represent meaningful time savings.

Between accelerated research and automated drafting, tax practitioners leveraging AI-powered tax research solutions are realizing significant efficiency gains. Among Blue J users, 93% agree that Blue J saves them time, at an average of 2.94 hours saved per user per week.

For all the benefits of saving time with AI-powered tax research, this newfound efficiency can also produce some anxiety. Accountants and other tax professionals have traditionally charged for their work on an hourly basis, although some are moving to a subscription model. Firms still operating under the traditional approach may have concerns that reducing the time spent on a particular task will also reduce the amount that can be billed for that task.

However, firms leveraging AI-powered tax research are learning to repurpose their time savings to deliver deeper client value. Blue J customer Larson Gross Advisors has been able to redirect time that previously went towards research and drafting into conducting complex research tasks, helping clients think through key decisions, weigh their options in different scenarios and tax treatments, and understand the broader impact of their choices. At the same time, they’re also shifting how they charge clients, adopting a value-based billing model that reflects the increased efficiency and value of their work.

The AI Tipping Point

Although AI-powered tax research is gaining adoption momentum, many firms have yet to integrate these tools into their tech stack.

Interestingly, our survey results revealed a discrepancy between the number of respondents who report using AI in their tax research (54.4%) and the percentage who work for firms that have formally adopted AI-supported tools (33.1%).

This gap suggests some practitioners are using AI for tax research at firms where there is no official AI tax research solution in place. Like search engine usage, this rogue AI usage could pose a problem for firms if it means practitioners are leveraging general-purpose AI tools. While general-purpose AI tools are easy to incorporate into research workflows, the efficiency they offer often comes at the expense of answer quality. Since these tools pull from a wide range of uncurated sources, they’re prone to error, often basing answers on irrelevant and outdated sources that are difficult to verify.

Firms of all sizes stand to benefit from AI-powered tax research, but smaller teams feel some benefits more keenly. Unlike their large-firm counterparts, tax practitioners at small firms typically lack access to a team of colleagues they can consult on unfamiliar tax matters. Given that AI solutions can essentially act as an additional team member—capable of assisting with both research and drafting work—these smaller firms are missing an opportunity to supplement their teams and punch above their weight with AI.

Demographics Shift

There’s a distinctly human factor incentivizing firms to adopt AI. An unignorable demographic shift is currently underway in the tax and accounting professions. A shortage of junior talent and an aging workforce mean that firms are not only short-staffed today, but are also likely to struggle with filling seats in the future. For all firm sizes except sole practitioners, finding qualified staff ranked as the top issue in the 2024 Private Companies Practice Section (PCPS) CPA Firm Top Issues Survey.

In firms that have previously relied on junior staff to perform research, AI-powered tax research can provide a much-needed pair of extra hands, essentially acting as a junior associate. These tools can also go beyond the immediate fix of filling in for missing juniors, serving to attract new practitioners to the profession by shifting the type of work they’ll do. Rather than spending hours gathering sources and performing manual research, juniors with access to AI-powered tools can instead focus on asking analytical and strategic questions of the research, essentially fast-tracking them into the work they would be doing later in their careers. Additionally, many newer tax practitioners are already accustomed to using generative AI in school, so incorporating AI into the firm’s tech stack can serve as a powerful recruitment and retention tool.

Overcoming Barriers to Adopting AI Tax Research Solutions

Even now, firms considering AI-powered tax research are likely to face several—real or perceived—barriers to adoption. To help counteract these fears, we’ve gathered perspectives from several tax and accounting professionals.

1. Fears of Replacement. Currently, generative AI can automate research, conduct next-step analysis, and draft research memos and client emails—freeing up time for higher-value advisory work. However, generative AI can’t exercise human judgment, make ethical decisions, or foster client relationships. As such, tax practitioners should view AI as a way to augment their capabilities, rather than replace them. Focusing on uniquely human skills and delegating to AI only where appropriate will allow firms to balance the efficiency gains of AI without detracting from overall firm performance.

2. Security Concerns. Not all generative AI applications have the necessary security measures to adequately protect client data, with some using information submitted by users to train their models. When selecting an AI tax research solution, look for applications that adhere to enterprise-grade best practices, feature a secure and encrypted database, are SOC 2 compliant, and are certified by an independent auditor.

3. Research Quality. Concerns about biased responses and hallucinations are common, particularly in light of recent, high-profile instances of these errors. However, response quality and accuracy continue to improve through retrieval-augmented generation, content curation, algorithm improvements, human-in-the-loop approaches, and user verification of output.

4. Cost. While it’s true that tax-specific AI research solutions are more expensive than their general-purpose AI counterparts, these solutions are still significantly less costly than legacy solutions, which have significantly lower adoption rates. As you consider the cost of AI-powered tax research, also consider the value of fast, accurate answers, as well as the cost of not adopting, which can result in less efficient research and fewer opportunities for higher-value strategic advisory work.

5. Retraining Time. Generative AI is designed to be intuitive, relying on conversational prompts rather than Boolean searches. This intuitive approach means your team already knows how to use it, even before opening the application. After logging into Blue J for the first time, it takes users an average of just 2.8 minutes to generate a comprehensive tax answer.

6. Resistance to Change. Tax research has been done a certain way for a long time, relying on a set of tools that many members of your team may be comfortable using. Change won’t happen unless they get a chance to experience all of the benefits generative AI has to offer firsthand. Some firms consider a trial period or phased rollout to allow team members across all levels to see just how naturally new technology can fit into their day-to-day workflows.

What’s Next?

The findings of the Blue J and CPA.com AI Tax Research Solution Outlook Report make one thing clear: AI is no longer just an experiment for tax practitioners. Already, AI-powered tax research is creating measurable improvements for adoption rates and research speed, providing a tangible advantage over legacy tax research solutions.

For firms still evaluating whether to take the leap, the decision window is narrowing. Those who delay will find themselves in a widening gap, competing against peers who are faster, more informed, and more efficient. And while barriers to adoption remain, the many firms already thriving alongside AI are proof that these challenges are surmountable.

The future of tax research is here. Now, it’s up to tax practitioners to decide when they’ll join that future.

But the finished product isn’t the job. The real work is the process of creating something.

The story isn’t the transcript

I heard Nilay Patel, Editor-in-Chief and Co-Founder of The Verge, talk about this at the Thomson Reuters Institute AI event in Chicago. He pointed out that AI can attend a city council meeting, transcribe everything that happened and write an article about it.

As the mom of a journalist, I immediately understood the distinction he was making.

A good reporter isn’t sitting in a meeting just to record what everybody says. They walk in with years of experience processing situations and determining what actually matters enough to include. The journalist in the room is there to figure out:

• Who has a vested interest

• What isn’t being said

• Which piece of context changes the story

• What questions still remain

Two reporters can sit through the same meeting and walk away with completely different stories because they noticed different things.

Accounting is like that too.

The same numbers with different insights

Two accountants can look at the same set of numbers and come away with completely different ideas about what matters. One might see a tax opportunity; another sees a cash flow problem coming six months from now. Someone who knows the client well might see a number that looks completely normal on paper and immediately know something has changed.

That’s the part of professional expertise that gets lost when we talk about all of the things AI can produce for us.

AI is getting really good at producing answers. It already has access to more information than any individual accountant could ever know, and it can process that information faster than we can.

But having all of the information doesn’t automatically tell you which piece of it matters right now!

That requires context, strong relationships with your clients and an actual human brain with years of experience understanding the work.

What will you do with the time AI gives you?

Still, AI can give us something incredibly valuable. Time!

We’ve spent so many hours producing work that we haven’t always had enough time to think about what the work is telling us, but that doesn’t have to be the case anymore.

We don’t have to painstakingly input numbers and stress about making sure the return is perfect. Now, we can spend less time getting the answer and more time figuring out what to do with it.

If AI can prepare the return in five minutes:

• What are you going to notice when you look at it?

• What do you know about the client that changes how you interpret it?

• What question are you going to ask because of what you see?

That’s the opportunity I see for accountants. We get to spend more of our time doing the thinking that comes after the answer. The finished product will keep getting easier to produce. Our value is everything we know to do with it once it’s in front of us.

If the return is all you’re selling, what happens when the client knows AI can do it?

If the Conversation Gap is what’s holding your firm back, here’s the good news. It’s not permanent, and closing it doesn’t require finding some rare personality gem. It takes a system, and systems can be built.

Most firms simply never built one. Conversation skills are treated as something you’re born with or something you eventually pick up through experience. That’s always been a real cost, and it’s about to become a much bigger one.

As AI keeps closing the distance on technical work, the basis of competition starts to change. Technical expertise will always matter, but high-level conversation skills are becoming a unique competitive advantage because they’re one of the few things AI can’t replicate.

So, how do you build conversation skills across the firm?

Focus on Continuous Conversation Improvement

Think about how an athlete improves their craft. Nobody hands a quarterback a playbook once a year and calls it coaching. He practices, watches film, gets specific feedback on what worked and what didn’t, and adjusts before the next game. He gets better because the system around him is built for it.

The same principle applies to client conversations in accounting firms.

It’s called Continuous Conversation Improvement (CCI). The idea is that every important conversation becomes an opportunity to prepare, learn, receive feedback and improve before the next one.

Walk through that cycle consistently, and something starts to change. Confidence, communication, trust and follow-through stop being qualities a few people happen to possess. They become capabilities your firm develops intentionally.

The result is better conversations and better business outcomes.

Improvement at Every Level

When CCI becomes part of your firm’s culture, improvement starts to happen at every level.

• Individual Level. Most people rely on one of two things: instinct that hasn’t been tested or running everything past a senior manager or partner before they feel comfortable moving forward. Both are slow. Neither builds lasting confidence.

  Take a staff member we know who found it genuinely intimidating to call a client about a past-due invoice. Instead of winging it or waiting for someone else to weigh in, she prepared, thinking through how she’d open the conversation, what language felt respectful but clear, and how she’d respond if the client pushed back. She then made the call and the client agreed to pay. But the big win was how she proved to herself that she could handle a conversation she’d been avoiding. The next time a similar situation comes up, she won’t be starting from zero.

• Team Level. This is where coaching starts to scale. As specific, actionable feedback becomes part of your team’s everyday work, people improve faster because they learn from one another’s conversations, not just their own.

  A tax manager we know experienced this firsthand after a file corrupted just before an important internal meeting. Someone reviewing the conversation pointed out that he’d immediately explained what happened, what information was missing and what the team could expect next. Nobody had ever identified that strength before. Now he does it intentionally, and he’s started giving his own team that same kind of specific feedback.

• Firm Level. At the firm level, CCI becomes part of how your firm develops its people. Instead of relying on a handful of partners who naturally excel with clients, you begin building that capability across the entire firm. You start to see where opportunities consistently move forward, where conversations tend to stall and where additional coaching will have the greatest impact. Over time, coaching becomes more targeted, improvement becomes measurable and developing stronger professionals becomes a repeatable process instead of something you hope happens through experience alone.

Over time, your firm develops high-level conversation skills the same way it develops technical expertise, through continuous improvement.

How Navi Supports Continuous Improvement

Most firms that want to coach this way eventually run into the same limitation. Partners and firm leaders only have so many hours in a week, and they can’t personally review and coach every important conversation happening across a growing team.

That’s where AI can make a meaningful difference. Most AI is designed to help you complete today’s task. CCI asks how you help your people become better at tomorrow’s conversation.

Instead of replacing human judgment, Navi strengthens it. It analyzes conversations, identifies patterns that are easy to miss and introduces personalized coaching at a scale partners simply can’t provide on their own. It’s a repeatable system for developing a team that consistently capitalizes on growth opportunities.

Closing the Gap, Growing the Firm

Every firm wants stronger client relationships and more revenue. Very few have a deliberate system for building the skills that produce both.

The Conversation Gap isn’t closed by luck or personality. It’s closed by helping your people get a little better, conversation by conversation.

That’s how AI opportunities become client action. And that’s how your firm builds a distinctly human competitive advantage.

When you talk about where something is headed, people always think you’re being dramatic for a while. Then, the market catches up and suddenly everyone is talking about the thing you’ve been saying for years.

I’m watching that happen right now with the changing structure of accounting firms.

The Firm I Saw Coming

In April 2024, I wrote an article for Accounting Today called “Protect the profession as traditional firms fade away.” At the time, I said we were witnessing the evolution of the “new firm” — professional services and technology merging with significant investment from venture capital and other outside capital while traditional firms merged up with private equity.

I was questioning what the CPA firm would look like by 2030.

Two years later, look around. Private equity has become a major force in accounting. Firms are consolidating, outside capital is pouring into the profession and entirely new business models are emerging.

The conversation has shifted from: “Could private equity change accounting?” to “What happens now that it has?”

I’m not just saying, “Told you so.” Okay, fine, maybe I’m kind of saying that! But I really want to let you in on why I could see it coming.

I Keep Coming Back to the Same Question

I’ve spent my career looking at where the market is going instead of where it happens to be today. I was early on cloud accounting. I was early on automation. I was early on the changing structure of firms. And underneath all of those predictions has really been the same question:

What is going to make an accountant valuable when the way we do the work changes?

That was actually the bigger point of my 2024 article. Private equity itself wasn’t the thing to fear. I wanted us to separate the CPA profession from the traditional CPA firm.

If accounting services can increasingly be delivered by businesses that look nothing like the firms we grew up with, then protecting the old firm model can’t be our strategy for preserving the value of the CPA.

That distinction matters even more today. Technology is changing how work gets done. Capital is changing where the work gets done and who owns the companies doing it. AI is automating and accelerating work we once considered central to being an accountant. That doesn’t make accountants less valuable. It changes what makes us valuable.

The Next Advantage is Human

That’s why I keep talking about coaching, communication and the human side of this profession. I wrote in 2024 that as technology and bots became more embedded in our work, we needed to focus on the human element and serving customers. I believe that even more strongly today.

Technical competency matters, but it won’t differentiate you when technology can do more of the technical work faster and cheaper.

The accountants who become more valuable will be the ones who can take all that technology, data and knowledge and do something distinctly human with it. They’ll be able to:

• Ask better questions

• Communicate complicated ideas clearly

• Understand what a client is really struggling with

• Coach that client toward a better decision

• Turn numbers and insights into a clear answer to “What should we do next?”

They won’t just deliver the numbers. They’ll help people figure out what to do next.

Don’t Wait Until the Shift Is Obvious

So when I talk about accountants developing coaching skills and becoming better communicators, maybe that sounds a little radical today. Maybe it feels unrelated to what you were taught an accountant was supposed to be.

But I’ve been early before.

The firm I wrote about in 2024 isn’t theoretical anymore. It’s being built around us. So maybe we shouldn’t wait until the next shift becomes obvious before we start preparing for it.

Seeing the future can be a little lonely sometimes, but I’d still rather be building tomorrow than protecting yesterday.

The common assumption is that firms are underpricing deliverables, underestimating hours or struggling to keep up with the shift from billable hours to value pricing. But the deeper issue is that firms are still pricing the output while quietly giving away the experience surrounding it.

That matters because clients don’t just buy a tax return, an audit or a financial statement. They buy access, responsiveness, interpretation, reassurance and confidence. They buy the relationship, even if the invoice is centered on deliverables.

Not All Time Carries the Same Value

One of the biggest flaws in legacy pricing models is the assumption that all time is created equal. It’s not!

Fifteen minutes spent formatting a file is not the same as 15 minutes spent calming a nervous client, spotting a risk before it becomes a problem or helping a business owner make a better decision. But firms often treat both the same because traditional systems are built to count time, not understand the true value of the service itself.

This takes us back to an inherently flawed system, and the fact that firms have been trying to capture time in an environment that makes accurate time capture next to impossible.

That blind spot becomes even more obvious when communication enters the picture. In fact, Laurel’s State of Work Report found that:

49.4% of the business day is spent on communication and coordination work

That means nearly half the day is going toward the conversations, check-ins, emails, meetings and touchpoints that move work forward and shape the client experience. Yet, a lot of that time is underweighted, unpriced or ignored altogether.

The Relationship Work is Real Work

Client-facing time often carries enormous value, but because it doesn’t always look like production, it gets treated as overhead. That’s a mistake, especially when you consider that Laurel found that:

50% percent of meetings involve external client interactions, and those meetings consume more time per session than internal meetings do

In other words, a meaningful share of professional time is being spent directly with clients, but firms still tend to anchor pricing around the thing produced, not the trust built around it. A major flaw in overall system design.

It’s important to understand how much time is actually client contact versus behind-the-scenes work because good client meetings have a bigger impact than simply creating the deliverable.

This is a critical pricing insight. If the relationship is what drives retention, trust, upsell opportunities and perceived value, then it can’t keep being treated like free labor wrapped around the “real” work.

Automation Changes the Math, But Not the Value

AI makes this even harder to ignore. As automation reduces the time it takes to complete certain technical tasks, firms can no longer rely on effort alone to justify price. Some firm leaders still instinctively compare this year’s work to last year’s hours, but that logic starts to fall apart fast in an AI-enabled environment.

Historical time is becoming less relevant because AI is changing the nature of the work and the workflows that surround it. But faster delivery doesn’t automatically mean lower value. AI-supported work can produce a more sophisticated deliverable, meaning that the hour itself may actually be worth more, not less.

AI improves quality, not just velocity, which means firms need a more nuanced way to think about pricing than simply tying fees to time spent. They need pricing based on judgment, access, context, communication and confidence because this is the part that’s being underpriced today.

What Firms Need to Price Now

Once firms can actually see how work happens, pricing starts to look different. Leaders can begin to distinguish between commodity execution and high-value interaction. They can see where relationship-heavy work is happening, where communication is driving outcomes and where advisory value is being delivered even when it doesn’t show up as a neat line item on an invoice.

And that’s the shift. Firms aren’t just underpricing services; they’re underpricing what clients actually value most. Until that changes, margins will continue to leak through all the invisible, high-impact work that keeps clients loyal, confident and coming back.

That’s a pretty fundamental change to how we develop people in this profession.

Preparation Was Never Just Preparation

AI agents are starting to do more of the work that traditionally taught young accountants how accounting actually works. The journal entries, reconciliations, returns and other preparation work weren’t just tasks we needed someone to complete. They were how people learned:

• What good work looked like

• What tended to go wrong

• What should tie

• What didn’t look quite right

• When something just didn’t make sense

Now the agent can do that work faster, but the human still has to make sure it’s right. And reviewing something is a different skill than preparing it.

“Check the AI’s Work” Isn’t a Training Plan

We can’t hand someone an AI agent and tell them to “check its work.” What does that actually mean? What should they check every time? What should tie? What mistakes should they anticipate? When should they trust the output, and when should they start asking questions?

Those are review skills, and we’ve historically developed them after someone had years of experience doing the work themselves. Now we may need to figure out how to develop them much earlier.

The Learning Curve Is Flipping

Instead of spending most of our time producing the work, we will soon be spending more of it:

• Reviewing

• Questioning

• Testing

• Interpreting

• Explaining

From the worker’s perspective, that could actually be harder. Imagine an agent constantly handing you completed work and knowing you’re responsible for finding the thing it got wrong.

That requires a different kind of confidence and judgment than simply knowing how to follow a process.

We Need to Teach People How to Supervise AI

That means firms need actual procedures for reviewing AI work. We need to teach young accountants:

• What they should review every time

• Where AI has failed before

• What common errors look like

• How to test whether an answer makes sense

• How to question the AI when they know something is wrong but don’t immediately know why

And they need opportunities to practice finding those errors before the stakes are high. You can’t just teach people to how to use the agent, you have to teach them how to supervise it.

Technical Knowledge Matters, But The Goal Changes

Young accountants still need technical knowledge. They still need to understand why something should tie, what a correct return looks like and what the numbers actually mean. But maybe we need to teach those things with a different goal in mind.

Instead of only teaching someone how to produce the answer, we need to teach them how to interrogate one.

That’s a significant shift.

AI Could Create More Room for Judgment

There’s an opportunity in that shift, too. We spent so much time doing the work that we barely had time to think about the work. If AI handles more of the preparation, accountants can spend more time:

• Understanding what the numbers mean

• Explaining them to clients

• Asking better questions

• Identifying what doesn’t fit

• Developing judgment

Those are exactly the capabilities firms say they want their people to develop sooner. But that development won’t happen automatically just because we gave them an AI agent.

The Profession Has to Redesign How People Learn

AI isn’t going back in the box. The young accountants of the world are going to have to learn how to work with it, just like I had to learn how to use the internet and Google.

The difference is that AI isn’t just changing the tool they use to do the work. It’s changing the work they need to learn first. Our job is to make sure they’re ready.

Nearly three-quarters said the greatest value would come from judgment, interpretation and helping clients make decisions. Only about one-quarter chose technical knowledge or getting the work done accurately.

That tells me accountants know what is coming, but the bigger question is whether firms are actually preparing for it.

We Know Technical Work Won’t Be Enough

During a recent webinar, I polled more than 100 accounting professionals about AI, client value and the capabilities accountants will need in the future.

When I asked what clients would be most willing to pay accountants for in 2030:

• 42% said judgment and interpretation

• 32% said confidence in making decisions

• 14% said technical knowledge

• 12% said getting the work done accurately

Put another way, 74% believe clients will place the greatest value on what happens after the technical answer is produced. So accountants realize their role will be having more conversations around:

• What does this mean?

• What should we do about it?

• What are we missing?

• What happens if we choose one path instead of another?

Those are very different questions from, “Can you prepare this correctly?”

For decades, technical knowledge and accurate execution have been the foundation of accounting. They still matter because nobody wants an accountant who gets the answer wrong. But being necessary does not automatically make something valuable enough for clients to pay a premium for it.

AI is steadily making it faster and easier to produce, research, review and analyze technical work. As that happens, value moves somewhere else. Apparently, most accountants already recognize that.

We Also Know Human Skills Matter More

I asked another question: As AI gets better at technical work, what happens to the value of human communication skills?

Eighty-nine percent said those skills become more valuable. Even more striking, 80% said they become much more valuable. That is an overwhelming consensus.

Accountants increasingly understand that the future of the profession depends on capabilities such as listening, questioning, interpreting, challenging assumptions, communicating insight and helping another human make a decision. But there was a contradiction hiding inside the responses. Among the people who said communication skills would become much more valuable, roughly one-quarter still said clients would be most willing to pay for technical knowledge or accurate execution. Think about that.

We can believe human skills will become dramatically more important while still believing the economic model will revolve around the work we have always done. That tension may be one of the biggest challenges facing firms right now. We intellectually understand the future but haven’t fully let go of the past.

Trying AI Isn’t the Same as Changing How You Work

When I asked participants where they were in their own use of AI:

• 29% said they use it every day

• 41% said they are experimenting but inconsistent

• 20% are mostly watching and learning

• 10% haven’t really started

There’s good news in those numbers. Accountants are no longer ignoring AI, with roughly 70% using it daily or actively experimenting. But experimentation is not transformation. The largest group is still in the middle, where they are trying tools, testing prompts and figuring out where AI fits. It’s an important stage, but firms can spend a very long time there.

You can use AI to summarize a document, draft an email or research a question without changing anything fundamental about your job. You simply perform the same job faster. The bigger shift comes when firms start asking a different question:

If AI can do more of the work accountants have traditionally been paid to do, what should our people be getting better at instead?

That is where this stops being a technology conversation and becomes a talent conversation.

Who Teaches Judgment When AI Does the Repetition?

One attendee essentially asked: If AI replaces much of the data entry and verification work people traditionally learned from, how do we teach judgment?

Exactly!

The profession has spent decades developing accountants through repetition. You prepare the work, then you prepare more of it. Eventually you review someone else’s work. Over time, you see enough situations to recognize patterns, identify problems and build judgment. You start sitting in more client meetings. You see how experienced partners respond. Eventually, you become the person expected to advise the client.

AI threatens to compress that development path. That can be a tremendous opportunity, though. I don’t think anyone needs to spend years doing repetitive work simply because previous generations had to. But we cannot remove the repetitions that developed judgment and assume judgment will somehow appear on its own.

If firms want professionals who can interpret information, ask better questions and give clients confidence in their decisions, they have to intentionally develop those skills. Earlier. And that may require redesigning career paths that have existed for decades.

The Profession’s Beliefs Are Ahead of Its Systems

Accountants believe human communication will matter more and judgment and decision support will create more client value. What I’m less convinced of is that firms’ development models and career paths have caught up with those beliefs.

Many firms still promote people based heavily on technical execution. They still wait years to give younger professionals meaningful client exposure. They still treat communication, coaching, curiosity and strategic thinking as “soft skills” rather than core professional capabilities. And they are still structured around a model in which producing the work creates the value. That is the disconnect we need to address.

The Firm of 2030 will not be created when you buy enough AI tools. It will be created when you redesign the work around what humans need to do once AI can do more of everything else.

The encouraging part is that accountants already seem to know where value is going. Now we have to build firms that reflect it.

However, that doesn’t mean we don’t use AI. I think we just have to get better at questioning it.

I was reviewing a tax return that an AI created the other day when I noticed something didn’t look right. There was a small discrepancy I wouldn’t have caught if I didn’t have decades of experience reviewing returns.

Here’s the funny part. I used the AI agent to help me find its own mistake.

I didn’t necessarily know what was wrong at first. I knew enough to see that something didn’t make sense. Then I started asking the agent questions.

• What looks off?

• Why doesn’t this tie?

• Where is this number coming from?

• What should have happened here?

Eventually, we got to the problem.

Trust Doesn’t Have to Be Binary

It made me think about the way accountants talk about trust and AI. We tend to make it very binary. Either we trust AI and let it do the work, or we don’t trust it and shouldn’t use it.

There’s a lot of space in between.

Think about a junior accountant. You wouldn’t give a brand-new staff accountant a return and say, “Great, you did it! Send it to the client.” You’d review their work. You’d ask questions. You’d teach them what they missed and why. Over time, as they demonstrated that they knew what they were doing, the way you reviewed their work would change.

I think we need to approach AI with some of that same skepticism.

The agent is kind of like a newbie. It can do an incredible amount of work, sometimes ridiculously fast, but that doesn’t mean you blindly trust the output. You still need:

• A review process

• Controls

• Someone who knows what good looks like

• And, most importantly, judgment

But Judgment Is Changing

We’ve always assumed that the person reviewing accounting work needs to know how to do the work themselves. And there’s obviously truth to that. You need accounting knowledge to know when something doesn’t make sense.

But preparing something and reviewing something aren’t exactly the same skill.

I can write something. That doesn’t automatically make me a great editor of someone else’s writing. The same thing happens in accounting. Knowing how to prepare a return and knowing how to look at a completed return, identify what seems wrong and ask the questions that uncover the problem are related skills, but they aren’t identical.

AI makes that distinction much more important.

The skill might not always be knowing exactly what’s wrong the second you look at something. It might be recognizing that something is wrong and knowing how to interrogate the AI until you understand why.

That can be taught.

AI Doesn’t Automatically Understand Your Job

We need to remember that most AI tools accountants are using weren’t built by accountants who have spent 15 years closing books, preparing returns or sitting across from clients. They were built by engineers.

That doesn’t make them bad tools, but software doesn’t automatically understand your job just because it can perform pieces of it.

So now we have to figure out the layer between the technology and the final work product:

• Who teaches the AI what good looks like?

• Who reviews what it produces?

• What does the preparer need to know?

• What does the manager need to know?

• What questions should each person be asking before something goes out the door?

Those are new skills, and we need to start training people on them.

Humans Get Things Wrong, Too

Because I worry that when accountants hear, “AI gets things wrong,” we use that as proof that AI can’t be trusted and go right back to doing everything ourselves.

Humans get things wrong too. That’s why we have review procedures in the first place.

We can’t expect AI to be perfect, but we need to make sure that our people and our processes can catch its mistakes.

Put Your Skepticism to Work

So question your AI. Push back on it. Ask it to show its work. Tell it something doesn’t make sense even when you can’t immediately articulate why. Make it explain itself. Ask the same question a different way.

When I tell you to implement AI in your firms, I’m not saying to trust it blindly. In fact, one of accountants’ greatest advantages could be our natural skepticism of new technology.

If we can learn to channel that skepticism into asking better questions, challenging the output and knowing what needs a second look, it could actually put us ahead.

You introduce a new AI tool inside your firm to transcribe meetings and capture action items. And then, in one of your first meetings using it, a client says: “I’m not comfortable being recorded.”

It’s usually said politely. Sometimes even apologetically. But the reaction inside the firm can be immediate and intense.

Partners start wondering whether they’ve moved too fast. Someone asks if recording is even allowed. Another person worries that clients will think the firm is monitoring them or collecting data in ways they didn’t expect. The technology that felt like progress suddenly feels like risk.

Pause … Right… There.

This is not about recording. It’s about communication!

Let’s Start with the Rules

In the United States, call and meeting recording laws generally fall into two categories: one-party consent and all-party consent. In one-party consent states, only one participant in the conversation needs to agree to the recording. In all-party consent states, everyone on the call must consent before the conversation can legally be recorded.

Accounting firms rarely operate inside a single state boundary. A partner may be sitting in Ohio while the client is in California. Another team member may be dialing in from Texas. When conversations cross state lines, you have to default to the stricter standard. That means ensuring that everyone on the call knows the conversation is being recorded and has the opportunity to consent.

The good news is that transparent consent isn’t a new concept. Tools like Zoom and Microsoft Teams both display an on-screen notification the moment recording or transcription begins. Participants can see it, and they can talk about it with you if they have concerns. The notification protocol on these platforms helps ensure the recording process is designed to be transparent from the moment the meeting begins.

Why Firms Are Beginning to Record Client Conversations

Client conversations typically disappear the moment they end. Often, someone leaves a meeting with a few handwritten notes and maybe sends a follow-up email with whatever details they happened to remember later. If another team member joins the engagement later, they have to reconstruct the context from fragmented notes or quick hallway explanations. Details get missed. Follow-ups get delayed. Institutional knowledge walks out the door when a team member changes roles.

Recording and transcription tools allow firms to preserve the substance of important discussions in a way that handwritten notes never could. Instead of trying to type while a client is explaining a complex situation, an advisor can focus fully on the conversation. Decisions, commitments and next steps are captured automatically and can be referenced later.

The First “No” Often Comes from Inside

More often than not, the biggest resistance to recording comes from you and your team and not from the clients.

There is concern that asking to record will feel intrusive and prompt negative reactions. As a result, someone in your firm is choosing not to start the recording because they don’t want to create an awkward moment. But clients are highly attuned to tone and confidence. If you apologize for using AI to help, it will be treated as something risky. If you speak about it confidently as a service enhancement, it will be viewed as professional evolution.

This is why firms that successfully adopt conversation intelligence tools focus first on internal alignment.

Introducing Recording to Clients

Don’t wait until the meeting has started to introduce recording. Start before the meeting ever begins. Send a short note explaining that your firm has implemented a transcription tool to improve follow-up and reduce note-taking distractions. Make it clear that its purpose is accuracy and service quality, and that clients can opt out if they prefer.

This does two important things. It removes surprise, and it removes pressure. If resistance still surfaces during the call, stay steady. You don’t need to defend the technology. You simply need to clarify intent.

Explain that the recording allows you to stay fully present, ensures commitments are captured correctly, and it helps your team deliver better continuity between meetings. If a client remains uncomfortable, offer flexibility. Ask if it would help if you paused recording during sensitive segments or provided transcription (no video) only. And if they still say no? Respect it immediately.

A calm response, such as “That’s no problem, I’ll just take notes manually today,” reassures the client that their preferences matter. Consider it a trust-building move and effective communication.

And one “no” is not a “no” forever. Once clients experience clearer summaries and stronger follow-through, their resistance softens naturally. Proof builds comfort.

Increase Client Comfort with Recording

In many cases, a client’s hesitation has less to do with the recording itself and more to do with uncertainty. Easing those uncertainties will help obtain approval. Here are a few considerations:

1. Clarify how the information is used and protected. Recordings should be treated in the same way as other client information, with strict access controls and clear retention policies. Simply explaining that the recording is used internally to document discussions and improve follow-up can alleviate many concerns.

2. Language matters. Some clients may react strongly to the word recording, while they are comfortable with the idea of transcription or meeting notes generated automatically. Emphasizing purpose rather than technology helps to alleviate uncertainty.

3. Connect the practice to the client’s own experience. Comparing the practice to an experience the client has already had, such as a Teams Meeting recordings or Fireflies notetaker, can introduce familiarity and reduce pushback.

The key is to demonstrate that recording is about improving communication and client satisfaction.

The Bigger Shift Happening

Recording client conversations is really about creating value. As AI takes over more routine technical work, the differentiator shifts. Clients aren’t paying for data entry, calculations or compliance alone. They’re paying for judgment, insight and the quality of the conversations that help them make better decisions.

That’s why accountants are paying more attention to conversation intelligence tools. These tools do more than create transcripts. They create visibility into moments that have historically been difficult to capture and even harder to scale.

Where your opportunity lies is in what you do differently now that you use them. You can stay present in a conversation instead of taking notes. You and your teams can follow through more consistently. New team members can understand client history faster. And you can learn from your best client conversations and share those practices more broadly.

Recording is not a radical departure from professional norms. It’s simply another form of documentation. The difference is that today’s technology makes that documentation more complete, more accessible and more useful.

At its core, recordings are about clarity. And as you shift toward advisory services, clarity becomes one of the most valuable assets you can create.

• You have to tell a client their fees are going up

• You need to explain why their bill was higher than expected

• You have to push back on their investment strategy

• You have to remind someone, once again, that you really do need their documents by the deadline

You know you need to make the call. You probably even know what you need to say. But that doesn’t mean you’re excited to pick up the phone.

So what if you could practice it first?

Using an AI Coach

Imagine telling an AI coach: “I need to tell Fred our fees are going up 12% next year, and I need to hold the line if he pushes back on the value.”

Then you tell it who Fred is. Maybe he’s skeptical or price sensitive. Maybe he’s warm but distracted. You can even tell the AI what you’re trying to get better at, like handling objections, asking better questions, explaining your value or actually setting next steps before the call ends.

Then you have the conversation. No, not with the real Fred. (Not yet, at least.) You practice with the AI playing Fred.

And maybe AI-Fred pushes back: “Twelve percent? What exactly am I getting for that?”

Now you have to answer the question instead of just thinking about how you might answer it. You have to explain the value you delivered this year. You have to listen instead of immediately getting defensive. You have to get comfortable sitting in that uncomfortable moment.

Practice, Get Feedback & Try Again.

And when the practice call is over, AI can tell you what actually happened.

Maybe it’s something like:

• You talked 71% of the time

• You never clearly explained the value

• You handled the price objection really well

• You completely forgot to establish the next step

So you practice again and it goes a little better this time.

It may still be a bit uncomfortable when you call Fred, but you’re more prepared, more confident and a lot less likely to panic when he pushes back.

AI Should Make You Better, Not Replace You

That’s a different way of thinking about AI. The goal isn’t to have AI call Fred for you. It isn’t even necessarily to have AI write a perfect script that you read to him. The goal is to use AI to make you better at having the conversation.

We don’t practice this stuff in accounting. We’re really into technical skills and CPE. Most of us learn communication skills on real clients, in real time, with real money and relationships on the line. No wonder we’re nervous!

A Safe Place to Be Bad at Something

AI can give us a safe space to be bad at something before we really need to be good at it. And unlike outsourcing the conversation altogether, practicing it actually builds your skill. You learn what an objection sounds like. You get more comfortable answering difficult questions. You hear yourself overexplaining. You learn to stop talking and listen.

Eventually, you don’t need the nudge telling you what to say next because you’ve developed the muscle yourself.

Practice Before the Stakes Are Real

AI gives us something accountants have never really had before: a place to practice before the stakes are real. So use it.

Before the next uncomfortable client conversation, don’t just rehearse it in your head. Don’t ask AI to write you a script. Have the conversation, let yourself stumble and get the feedback. Then do it again.

You don’t have to focus solely on getting better at using AI, but you do to need to get better at the part of the job your clients will still need you to do.

The traditional timesheet assumes that work happens in clean blocks, but the actual workday does not roll out that way. In fact, according to Laurel’s State of Work Report:

Professionals have, at most, 15-minute periods of uninterrupted focus during business hours before communication requires attention again.

Additionally, a typical day includes more than 110 email interactions, each adding another break in attention and another data fragment.

That fragmentation isn’t just annoying; it literally rewires how work gets done. Deep work, that high-value analysis and problem-solving clients actually pay for, does not thrive in the spaces between pings, chats, meetings and hallway-level requests. It gets pushed out of the core business day and into the edges.​

The report shows exactly that as:

• Nearly half of weekly deep work happens outside standard business hours

• 28.2% of all work takes place outside of 9 to 5

This is all a blaring signal that the official workday is dominated by communication and coordination, while focused value creation gets displaced to early mornings, evenings and weekends.

And to be clear, the big issue is not that professionals suddenly became bad at timekeeping. It’s that firms have spent decades asking people to produce precise data from an imprecise environment, then acting like the output was good enough to run the business.

That’s the elephant in the room. Many leaders don’t want to admit that they’ve managed on incomplete data for a very long time. But now we know better because we have the data to back the claim.

So, the real question then becomes whether firms are willing to first admit the traditional workflow is flawed and then lead differently to correct it.

Interruption Becomes the Default Operating System

From an operational perspective, firms are still managing by timesheets instead of actual work being done, and by the time the reports are assembled, the information is often already stale.

This describes a current broken system where leaders are often relying on lagging indicators and manually pulled reports rather than trusted, timely visibility into what is happening across projects and people.​

That delay matters because the modern workday is dynamic. Fire drills happen. Senior people jump into tasks they should not be doing. Staff get blocked waiting on client information. Quick consults, out-of-scope asks and scattered billable moments happen in between other activities. And none of this fits neatly into the tidy narrative structure the old timekeeping model expects.​

So, when firms ask people to keep accurate timesheets, they aren’t asking for discipline but, rather, for a retroactive reconstruction of a day built on interruption.

That’s why the data is consistently incomplete. And this is not the result of professionals being careless. The problem is that the operating environment makes reliable manual capture structurally unrealistic.

This also explains why so many firms misread the symptoms. They see low realization, uneven utilization or margin pressure, and they assume the root cause is productivity. But if the workday itself is broken, then the numbers flowing out of it will be broken, too.

You can’t build sound economics on top of fractured attention and incomplete reporting.​

What This All Says About Value

There’s a second layer here that firms should not ignore, and that’s what the pattern says about value. If the work requiring the most concentration naturally migrates outside business hours, then the issue is not merely time capture. it’s operating design.

Firms have organized the day around responsiveness while leaving concentrated expertise to fend for itself.​ And while this model may have worked when work was easier to bucket and easier to remember, it doesn’t work so well today.

In a world of constant communication, AI-enabled acceleration and rising pressure on margins, firms need a better way to see how work actually flows. They need data that reflects reality as it happens, not as someone tries to summarize it later.

Until that happens, firms will keep diagnosing the wrong problem. They will keep treating missing time as a compliance issue instead of a design issue. And they will keep asking professionals to report work in a way that no longer matches how professional work is actually performed. The simple fact is that you can’t fix time tracking without fixing how work happens.

Firms can no longer treat missing time like a people problem. When the day is built around interruptions, meetings and reactive communication, of course the data that comes out of it is skewed.

So, no more playing the “blame game.” You can’t pin inaccurate reporting on your team when your current structure makes accurate reporting almost impossible.

There was a small discrepancy in the numbers. Mistakes happen. A human could’ve made the exact same error. But it made me think about the processes that we have to rework when we introduce AI into our firms.

You can’t just add AI to the process you already have and expect everything to work the same way. And I think a lot of firms are going to start learning that the hard way.

When a Human Makes a Mistake, We Ask Why

Think about what happens when a person on your team makes a mistake. You don’t just fix it and move on. You ask what happened. Was something missed during the close? Were the accounts reconciled? What was reviewed before the work went out the door? Most importantly, what needs to change so you don’t have the same problem next month?

That’s basic accounting. But something weird happens when AI enters the equation. We start thinking that because the agent did the work, the work is done. It’s not!

If you’re signing your name to something, you still need to know it’s right. That means someone needs to understand what the AI did, what it didn’t do and how its work was checked before it got to you. And that’s bigger than adding “review AI output” to your existing checklist.

AI Changes the Assumptions Behind the Workflow

Our accounting processes were built around humans. Someone prepares the work. Someone reviews it. We know what each person is responsible for, what they should check and when something needs to be escalated. Over time, we train people to develop enough judgment to recognize when something just doesn’t look right.

Now we’re putting AI agents into those workflows, and that changes the assumptions the workflows were built on. So the process has to change, too.

Firms need to start asking:

• Where does the AI belong?

• What should it be responsible for?

• Where does a human need to step in?

• What does that person need to review?

• How do they know when the AI is wrong?

• If something goes wrong, how do we figure out whether the problem came from the AI, the human or the process itself?

Those aren’t technology questions. They’re workflow and control questions.

Using AI Is Not the Same as Reviewing AI

There’s also a training issue here that I don’t think firms are talking about enough. You can’t teach a young accountant how to use an AI agent and assume you’ve trained them to review its work. Those are completely different skills.

If AI is doing more of the preparation, we have to be even more intentional about teaching people what good work looks like. They need to understand why something should tie out, what should be reconciled and what questions to ask the AI agent when the numbers don’t make sense.

Otherwise, how are they ever going to know when the AI is wrong?

Don’t Drop AI Into Step Four

This is why I think AI implementation is going to require firms to rethink a lot more than their technology stack. We need to rethink our workflows, our review procedures, our training and even what we consider a finished work product.

Don’t take a 20-year-old workflow, drop an AI agent into step four and call yourself an AI-enabled firm. Start over. Look at what the technology can do, where it needs human judgment and what controls need to exist between the two. Then build the process around that reality.

Because whether a mistake comes from a human or an AI almost doesn’t matter. If bad work can make it all the way through your workflow without being caught, you don’t have an AI problem. You have a process problem.

AI didn’t create it. It just made it impossible to ignore.

We’ve spent the last few years teaching accountants how to use AI. That matters. People need to understand the technology, know how to use it responsibly and get comfortable incorporating it into their work. But that cannot be the end goal.

The real question is what accountants need to become because AI exists.

Where Does the Value Move?

As AI takes on more of the work accountants have historically performed, the value doesn’t disappear. It moves. If AI can:

• Summarize the data, the accountant needs to explain what it means

• Generate possible answers, the accountant needs the judgment to know which one makes sense

• Automate more compliance work, the accountant needs to have better conversations about what the client should do next

That means skills we have sometimes treated as “soft” become incredibly important: curiosity, communication, empathy, critical thinking, professional judgment and the ability to challenge someone constructively. Those aren’t nice extras in the Firm of 2030. They’re core accounting skills.

We Have to Rethink How People Learn

Historically, accountants developed judgment by doing the work. They had thousands of repetitions, saw different situations and learned from more experienced people along the way. If AI removes some of those repetitions, we need to be intentional about replacing the learning that came with them.

We can’t automate the work and then wonder why younger professionals aren’t developing experience fast enough. That means rethinking training, giving people client exposure earlier and creating more opportunities to practice judgment, communication and problem solving. AI changes the work and our talent model has to change with it.

That’s the Conversation I Want to Have

That’s why I was excited to help ADP shape its AI Forum. I’ll start the afternoon with Heather Sperduto talking about what the Firm of 2030 needs to become, then join Laci Buzzelli and Tal Goldhamer for a conversation about how we prepare our people for an AI-powered future. Because figuring out which AI tool to buy next is probably the easy part.

Figuring out what kind of firm — and what kind of accountant — we’re building is much harder and much more important.

Register now for the ADP Accountant Connect AI Forum on September 29 from 1–3 p.m. ET, with up to two free CPE credits available.

Join us and be part of the conversation.

Lately, I’ve been trying to be a little more discerning. I think the better question is: What should I give it?

There are things AI can do for you that save time, eliminate work you hate and make you better at your job. Great. Give it those things! But there are also things you could hand over to AI that you probably shouldn’t, especially if they’re the skills that make you valuable.

Think about it like outsourcing. If I’m really good at talking to clients, but terrible at organizing my notes afterward, why would I outsource the client conversation? Let AI organize the notes! Let it summarize the meeting, create the follow-up list and remind me what I promised to send.

I don’t need to become the world’s greatest meeting-note taker, but I absolutely want to keep getting better at the conversation.

Efficiency Isn’t the Only Thing That Matters

That distinction is important. We’ve spent the last few years talking about AI primarily in terms of efficiency. What can it automate? How many hours can it save? How much more work can we get done?

Those are accounting questions, right? Of course, we want to measure the efficiency. But I think there’s another side of the equation we’re missing: What skills are we losing when we automate them?

If you use AI to write every email, eventually you might become a worse writer. If you ask it to analyze every complicated situation before you’ve thought about it yourself, maybe your critical thinking gets a little rusty. If you let it handle every difficult client interaction, how are you going to get better at having difficult conversations?

Don’t Automate Away Your Advantage

AI makes it incredibly easy to avoid practicing things that are hard. And sometimes that’s exactly what we want! I have no desire to practice a skill I don’t need, don’t enjoy and don’t believe is important to my future value. But if it’s something that differentiates me? I want to be very careful about giving that away.

This is especially important for accountants because we tend to start from the technical side. Maybe you’re incredible technically. You can spot something in a tax return nobody else sees. You understand the numbers at a level that makes you invaluable to your clients.

If that’s your superpower, don’t hand the entire thinking process to AI just because you can. Use AI to challenge your thinking. Ask it what you might have missed. Let it be your second set of eyes. Use it as a coach instead of a replacement.

On the other hand, maybe the technical stuff comes naturally, but you struggle to connect with clients.

That’s probably not the skill you can actually outsource, but it’s a skill that AI can help you practice. Role-play the conversation. Ask it to challenge your explanation. Have it tell you where you’re using too much jargon. Then go have the actual conversation yourself.

Decide What Belongs to You

That’s why I don’t think there’s one “right” AI stack for every accountant or every firm. Before you choose the tool, you need to understand what you’re trying to protect.

Ask yourself:

• What am I uniquely good at?

• What do I want to become better at?

• What makes clients want me involved?

• What work do I genuinely not care about mastering?

Then decide what belongs to you and what belongs to AI.

The accountants who win with AI won’t necessarily be the ones who automate the most. They’ll be the ones who know what not to automate.

AI should make you better at your job, not slowly make you less capable of doing it. Use it everywhere you can. Just don’t outsource the parts of you that your clients are actually paying for.