Building Fast, Living Slow

Chad Davis, CPA, co-founder of LiveCA and founder of AutomationTown, built an entire community around using technology to automate accounting work. His goal? Creating capacity for more leisurely walks with his family—not more client accounts.

It’s 10 o’clock in the morning in Spain, and Chad Davis is on a walk with his wife. Nobody in North America is awake yet, and they won’t be for another three hours or so. That leaves enough time for breakfast and errands—and no particular reason to hurry through either. He calls this window a “cheat code.”

Davis started LiveCA, Canada’s first fully virtual accounting firm, in 2012. He is also the founder of AutomationTown, a paid community where accountants learn to build with AI. 

Across the accounting world, Davis is known as one of the more technically fluent people in a profession that’s often pegged as moving too slow. But despite his reputation as a tech guru, he argues that “slow” doesn’t always equate to “bad.” In fact, speed is rarely the priority with anything he builds.

I don’t use automation to be faster,” Davis says. “I use it to strengthen the parts of the job that allow me to enjoy the slow part.”

Chad Davis

Old Fiverr Receipts

LiveCA was born out of an idea Davis and his wife dreamt up while living in the Cayman Islands for nearly five years. Eventually, they decided they missed the snow, so they moved back to Canada and got to work building the firm. About a year later, Davis met Josh Zweig, the firm’s co-founder and CEO. Though LiveCA is based in Toronto, it has been fully remote since 2012.

The firm has always been tech-forward (kind of a requirement for a 100% virtual operation that existed long before remote work was the norm), with Davis constantly trying to automate things before the right infrastructure existed. In fact, a decade-plus later, there are still emails sitting in his inbox from people he hired on Fiverr to write small pieces of code—a Google Apps Script here, a web scraper there, an API call he could (sort of) describe but definitely could not build.

There are receipts in my email from me hiring people on Fiverr and sitting there watching them build Google Apps Scripts for things that I really wanted to automate, but I couldn’t do [myself] prior to 2022,” he says.

Chad Davis

At that time, the only alternative was teaching yourself—which typically meant hours of tumbling through Google search results, YouTube videos, Reddit posts, and Stack Overflow, desperately trying to make sense of it all.

“It was a slog,” Davis recalls, “But it’s how we learned.”

All of that changed when generative AI exploded onto the scene. Suddenly, terms like “prompting” and “LLM output” were everywhere. For Davis, that flattened the learning curve substantially. Figuring out the “how” no longer meant hours and hours of research and self-education. The new sticking point was choosing the “what.”

Interestingly, though, while Davis views AI as a tool that can reduce the time it takes to learn something new, Financial Cents’ recently released State of AI in Bookkeeping & Accounting Report revealed that many accounting professionals consider the time required to learn AI itself as the biggest barrier to taking full advantage of this still-emerging technology.

Davis doesn’t disagree that time is scarce. But he also wonders how much of the problem is time versus expectation. Are accounting professionals simply abandoning AI when it doesn’t show its value instantly (or at least as quickly as they thought it would)? To explain his point, Davis drew a comparison between a brand-new AI tool and a brand-new employee.

If you’ve hired somebody, and they come in on Monday and by Wednesday they’re not doing client work, are you gonna be mad?” he argues. “No. You recognize that there’s going to be an investment upfront to get this working. And in some firms…it might take three months before they’re super helpful. And I think that’s what the issue at hand is [with AI]. How much effort do you want to put into this to be able to see a return?”

Chad Davis

The way Davis sees it, getting value out of AI is no different. It requires patience and commitment. Accounting leaders know that their investment in onboarding a new team member will eventually pay off in the form of increased capacity, higher-quality work, and better morale across the firm—and they must apply that same mindset to their investment in AI (or any other new technology, for that matter).

“Information is not going to be the problem anymore,” Davis says. “It’s what do you prioritize? How much effort do you want to put into it? And how curious are you to be able to get to a point where you feel good building something for your team?”

A Car with Bad Power Steering

If committing time to AI is step one, choosing the right tools is a very crucial step two. Davis sees this mistake play out constantly. Someone decides to take AI seriously, so they spend ten-or-so hours on it—mainly interacting with a chatbot. And not surprisingly, by the end of those ten hours, they have very little to show for their investment.

“You could easily at the end of those ten hours determine that was not useful, completely useless,” he says. “And then that might be what you carry with you when you associate AI [with value to your firm].”

Again offering a metaphor to illustrate his point, Davis likened a new AI user’s first tool to a new driver’s first car.

It’s kind of like getting a driver’s license when you’re young, and you get a really bad car,” Davis says. “And it has bad power steering, and you’re like, ‘No, I don’t want to drive anymore because that was a horrible experience.’ Meanwhile, the car that has better power steering and is nicer to drive and has a better suspension could be the most enjoyable thing you’ve ever experienced on the road. That’s really tough to swallow because you don’t know sometimes that the other car exists.”

Chad Davis

Beyond expanding their horizons on tooling, Davis encourages accounting professionals to think bigger about what they can do with the tools. This requires a mental shift to a builder’s mindset. The true value of AI doesn’t usually come from raw, prompt-to-answer output, Davis argues. Instead, it comes from using the tool as a means rather than an end—to create something that runs on its own, forever, without you.

“You could use this tool to build something that doesn’t use AI at all,” Davis explains.

That could be a script, a webhook, or a connector between two systems that historically haven’t been able to share context. Anything that removes a manual to-do from your task list is worth exploring.

Of course, the other piece of a builder’s mindset is that building requires testing, and sometimes tests fail. Not every idea will pan out the way you envisioned, and that’s okay. In fact, as Davis explains, those “lessons learned” are an important part of the process. And even as a tech and automation expert, he’s been learning those lessons for years.

Failing Forward

Back in 2019, for example, LiveCA hired a developer and put four or five employees on a project Davis had been thinking about for a long time: taking every spreadsheet process at the firm and consolidating them all into a single app. It was, he says twice in the same breath, “very expensive.”

In the end, they threw out the whole thing.

We ultimately scrapped it because it would cost so much money to get incremental gains,” Davis says. “The timing was off. We didn’t have AI, and the quality wasn’t there.”

Chad Davis

But even though they failed at building the app, Davis says the endeavor was still worthwhile, because it illuminated a much larger foundational issue—one that could have caused a lot of downstream damage had it gone unaddressed: they couldn’t trust their own data.

“If you’ve been in an accounting firm for a long time, the whole question of is that right—it comes up way more often than it really should,” he says.

And during this particular build process, it was coming up a lot—enough that fixing it became a top priority.

One of the most important aspects that we were missing was centralized data that was the source of truth. Full stop,” Davis says. “So we spent years making that source of truth believable.”

Chad Davis

The tedious work of making their data dependable enough that something could be built on top of it was critical to all of their future automation endeavors. But again, it required a major commitment of both time and effort. The long-term payoff was well worth it, though, and that’s exactly why Davis hesitates to push anyone toward a “speed-above-all-else” philosophy.

The Mythical AI Hump

As the leader of AutomationTown—described on its website as a “community for accountants who build with AI (or want to)”—Davis encourages members to work on problems that deserve a substantial time investment.

One member, for example, came in with his entire audit workpaper process living in a spreadsheet with an unmanageable number of tabs. His guiding question was how he could make the spreadsheet more efficient. With that goal in mind, he started building.

Where the tool’s at now is that it’s a fully fledged, hosted application with databases and users and security,” Davis says. “And it does the work. There’s no more spreadsheets.”

Chad Davis

Other impressive member achievements include rebuilding Xero Projects across a firm, creating a successful lead magnet, and taking a reporting pack for a group of companies from three days down to a couple of hours.

Davis says his most successful community members are not necessarily the most tech-savvy, but those who picked a problem big enough to justify the time spent solving it. If something eats 15 hours a month and takes 15 hours to fix, he points out, you are square after four weeks and ahead forever.

But beyond honing in on a worthy problem, how can folks in the accounting world push themselves over the “hump” of becoming AI-forward?

For his part, Davis rejects the idea of a “hump.”

“I actually believe you don’t have to strive to get over the hump,” he says. “Because there’s always going to be another thing to get over.”

Furthermore, Davis believes there’s no real end point to learning about technology, especially in this new age of AI. Perhaps that’s also why he takes such an unhurried approach to his own AI education.

People should feel safe learning at whatever speed and degree that they want to learn,” he says.

Chad Davis

This slow-and-steady mindset drives LiveCA’s general approach to all things AI, from governance to formal training. Davis prefers a more flexible approach to adoption and learning.

“We’ve actually had discussions around [whether we] should do more trainings,” he explains. “And we’ve ended up at no. We’re not leaning into more documentation. We’re not leaning into more sessions of training. We’re not leaning into hard-and-fast rules of governance.”

Instead, the team adheres to common-sense guidelines about what they can and can’t do with AI. The most important rule of all: whenever someone is unsure, they should ask. Simple as that.

Davis and the leadership team also maintain an open-door policy for anyone who wants to meet and walk through their specific AI questions.

[They] tend to get more out of a safe one-on-one fixing a specific problem, clicking the buttons together, than they do in a group setting where it’s kind of passive,” Davis explains.

Chad Davis

Why AI Access and Authority Must Be Separate 

As things like AI agents become more mainstream, a common question that tends to arise is which AI actions should run on an autonomous basis—and which should be monitored or approved by humans in the loop?

It’s a particularly important distinction when AI plugs into other systems, especially those housing sensitive client data.

“Just because we can give the AI access to Financial Cents, it doesn’t mean that we should give it the authority to change the stuff in Financial Cents,” Davis says, using his firm’s practice management software as an example.

But while access and authority should be separate decisions, Davis says small firms often make them at the same time. This is part of why he recommends a slow-ramp approach to offloading a firm’s most complex processes (or those involving sensitive information) to tech-powered automations. He suggests running the automated version alongside the manual one for a while, watching where the edge cases surface, and fine-tuning over time until you reach a point where you feel comfortable making the full switch-over.

“There’s this risk that people perceive speed with competence,” he says.

With AI, he cautions, it’s easier and faster than ever before to design and implement new tech solutions. But hasty work often leads to shoddy work, and the cost of a bad build can quickly spiral.

You’re just that crazy person that wants to automate everything and it just breaks and you cause more work than it’s actually saving,” Davis says. “It’s a brutal cycle.”

Chad Davis

As for the autonomy question, the biggest line LiveCA currently draws with respect to what AI can and can’t handle on its own is internal versus external:

• Anything client-facing runs through a person.
• Anything internal is fair game for AI automation.

He also admits to having no idea how long that line will hold.

“I could almost guarantee you that if you ask me this in six months, that line is blurred,” Davis says, emphasizing that it’s only a matter of time before clients have their own agents that they want to connect with the firm’s.

For now, LiveCA’s current approach is generating meaningful results. One clear example is a bookkeeping file that used to consume ten hours a month and now takes two or less.

For Davis, that increased capacity is proof that AI is having the intended effect at LiveCA. Unlike some firm owners, though, he’s not interested in filling that capacity with more clients. Instead, the extra hours go back into “creating the space for people to have better conversations, to hit their deadlines, to spend a little bit of extra time with their customers.”

It’s a choice that has some historical precedent behind it. LiveCA grew to around 120 employees during the hypergrowth years of 2021 through 2023; today, the team is down to 60. The contraction was intentional, but that didn’t make it any less painful. 

“Becoming 60 people from 120 was the best thing that’s ever happened to us, even though it was really hard and it hurt a lot of people,” he explains.

The downsizing was necessary, he says, to build a firm that could hold to its priorities: calmness, work delivered on time, and not chasing growth for the sake of growth. He’s not interested in making the same mistake twice, and automation is a huge part of what keeps the smaller, more stable version of LiveCA running smoothly with fewer hands on deck.

Peace and Purpose

After spending years living in different places around the world—including eight years on the road in an RV with two kids in tow—Davis knows a thing or two about paring down to the bare essentials. There is peace in the simplicity, he says, and that’s ultimately what he’s after—personally and professionally.

He and his family currently reside in Spain, with no plans of uprooting themselves anytime soon. They’ve all embraced the less-hurried speed of daily life, siestas and all.

Meanwhile, Davis’s co-founder, Zweig, is on a worldwide quest to conquer mountains. He just summited Mount Elbrus in Russia, and he also recently climbed Mount Vinson in Antarctica. Davis mentions it as further evidence of the culture they’ve built at LiveCA: one that views technology and automation as a means to creating more space and balance inside and outside of the firm.

I get peace from learning new skills so that I can apply them later and have more peace,” he says. “It compounds over time.”

Chad Davis

He’s not interested in moving faster in the name of maximizing growth or optimizing margins. He knows other firms out there might take that route, and he’s at peace with that too. After all, he didn’t start a fully remote firm in 2012 with visions of following any type of conventional path.

When asked whether the firms of today will look radically different in five years, he answers with a shrug, as if to suggest he’s not qualified to make that kind of projection: “I don’t know. Who am I?”

Davis argues you could have asked the same question in 2011, 2016, and 2021—and knowing what we know now, the answer would have been a resounding “yes” every time. The pandemic, in particular, came with a sweeping and largely unexpected shift, making remote work ordinary for firms that never planned to operate outside of a physical office. LiveCA, remote since day one, didn’t have to change much, but that didn’t mean they had things 100% optimized.

“There were some firms that probably operated way better than we did and had no experience,” Davis admits.

Instead of spinning your wheels on forecasting how things might unfold long-term—or comparing what you’re doing to what everyone else is doing—he encourages his fellow accounting professionals to think incrementally. At the end of planning any build, process change, or workflow initiative, ask yourself one question: How else could we do this using different technology?

“You might be surprised at the path that it shows you,” he advises, adding that he’ll never stop trying to free up more breathing room for himself and his team.

Our mantra at our firm, the whole purpose we exist, is so that people can have meaningful work and a meaningful life outside of it,” he says. “And without the two, why even bother?”

Chad Davis

For Davis, that meaning comes through strongest every morning he wakes up in Spain and eases into a precious few hours of unhurried time. It’s time he can be fully present, fully calm, and fully focused on what matters—time bought with years of unglamorous building and protected with tools that continue giving him back more hours rather than creating more work. He learns something new, it saves him more time, and he spends the time learning the next thing.

It compounds.

Chad Davis is the co-founder LiveCA and the founder of AutomationTown. Learn more at liveca.ca and automationtown.io.

Be among the first to know

Enter your email address to receive new issues of the magazine straight to your inbox.