Mid-Tier Accounting Firms Tread Slowly in Gen AI Transformation
Six of the largest mid-tier CPA firms in the US haven’t been in a rush to dive into generative AI, opting for a more cautious approach than their much larger Big Four counterparts to embrace technology transforming how their accountants and consultants work.
Still, the mid-tier firms began to roll out productivity applications that rely on generative artificial intelligence to help craft emails and compile meeting minutes over the past year. And they are testing dozens of other uses of a technology that promises to free up hours, even days of work for accountants and business consultants so—the companies say—they can spend more time focused on their clients.
“We want our teams to play with the technology, to experiment with AI, to see how it could make their life easier,” said Jen Leary, CEO of CliftonLarsonAllen LLP, one of the 10 largest US accounting firms.
Unlike the Big Four firms, these mid-sized firms don’t have billions to invest in the latest generation of AI that can summarize reams of tax regulations or contracts into clear, concise paragraphs. But their small- and mid-market clients are lining up, seeking help to adapt the technology for their own businesses—promising a stream of new revenue.
Table Stakes
CLA, which works with main street businesses and local governments, has set aside $500 million to advance its technology capabilities—with the bulk of that funding for AI, Leary said.
That investment represents roughly a quarter of the firm’s annual revenue. CLA’s competitors among the top mid-sized firms in the country declined to say how much they are spending.
Those firms range from RSM US LLP, which reported $3.7 billion in revenue in 2023, to Baker Tilly Advisory Group LP, rounding out the 10 largest firms with $1.6 billion in revenue last year.
Over the past year, CLA launched its own chatbot, acquired UK-technology company Engine B, and said it will roll out a Microsoft productivity tool called Copilot that runs on generative AI to its workforce.
Chatbots and virtual assistants have become table stakes for the firms and their clients, offering a gateway to more advanced uses of generative AI.
Grant Thornton Advisors LLC has challenged staff to free up four hours of work each week with the help of tools like Copilot. Enzo Santilli, chief transformation officer at the firm, wants his Grant Thornton colleagues to be comfortable using the technology, he said.
“So that when we start to advance some of the more commercial use cases, there will be better familiarity. It won’t scare people,” Santilli said.
Mid-tier firms have focused on piloting early applications, crafting guardrails for their use and providing training. They continue to roll out productivity tools like Copilot or internal assistants that answer IT or benefits questions.
“We weren’t in a hurry to get it out there. We wanted to make sure that we were fully informed before releasing it,” said Denny Ard, innovation director at Forvis Mazars LLP, which released Copilot to its more than 7,000 professionals this spring.
Boundaries
The mid-tiers say they have already set clear boundaries on how accountants and consultants can use products with gen AI baked in, writing formal safeguards and requiring training before staff could use any new capabilities.
Generative AI tools at Grant Thornton, for example, can’t be used for direct client work like audit opinions. However firms have provided specific examples to encourage employees to tap generative AI for tightening an email before sending it or writing a first draft of a report, for instance.
“None of this gives you the final answer,” said Sergio de la Fe, enterprise digital leader for RSM. “It takes the bite out of the massive writing and detailed work that can really help you get faster.”
Tech Race
In sharp contrast to the mid-tier firms and their tests, the Big Four have been quick to jump into generative AI, broadcasting their plans to integrate it into their DNA.
Their spending spree has pushed smaller firms to carve out funding in their own budgets needed to adopt the tech, said Isaac Heller, CEO of Trullion, a platform for accounting and finance teams, which works with more than 100 US accounting firms.
Those commitments add up to more than any individual firm’s position in a competitive professional services market. Cutting-edge technology could streamline audit and tax work, and also could entice a new generation to join a shrinking accounting workforce.
“The industry needs to buy in and get together around adoption because you’re facing a major labor shortage and a general reputational challenge,” Heller said.
A Leap Forward
Efforts by mid-sized firms to scale up generative AI also serve as a sort of how-to guide for clients who could use the technology to leapfrog their own competition.
Demand for generative AI, however, has led to tough conversations with clients about the state of their in-house data and whether they are prepared to take advantage of the tools.
Mid-market companies don’t have the deep pockets or staff resources of larger businesses in the Fortune 500. Some haven’t invested in cloud-based data platforms or safeguards to protect the integrity of the information.
“We want this thing that we can’t have because we don’t have good quality data,” said Kirstie Tiernan, digital practice leader for BDO USA P.C., of the quandary facing some of the firm’s clients.
Other clients are squeamish about the risks posed by unreliable and sometime irrational results from generative AI tools, known as hallucinations. That sets a high bar to move from tackling routine tasks to uses that directly benefit clients.
“They want very high assurance that the end result is going to be the same or better than what we are producing and how we are producing it today,” said Ethan Bach, managing principal for strategy and innovation at Baker Tilly Advisory Group LP.
Such concerns mean some clients will have to wait a little longer than they may have anticipated for generative AI to take hold.
“We don’t want to blow the world up; we want to be deliberate,” RSM’s enterprise digital leader de la Fe said.
link