What kind of AI do accounting software buyers want?

Just about every accounting software company, it seems, now touts its AI capabilities. Many make bold claims about automating workflows and freeing accountants to spend more time solving problems or working with clients, and a good number use the term “agents.”

But what are the folks who buy this software actually looking for—especially those who might not work for big firms or have the largest budgets? A survey by software review platform Capterra sheds some light on what accounting software purchasers want most in terms of AI.

Capterra surveyed 3,385 software buyers across a variety of sectors in 11 countries. 139 of the respondents were in the accounting industry, and of this group, around 72% were from businesses employing fewer than 250 people, Capterra Senior Analyst David Jani told CFO Brew.

Accounting software buyers are intensely interested in AI, the survey found. The vast majority (94%) said they were adopting AI, and 24% said they were doing so aggressively. And more than a third (36%) said that adding functionality such as AI is the top reason they were spending more this year, suggesting that the technology might be driving purchases.

Though agents are all the rage, accounting software buyers seem to need less flashy forms of AI. When asked which AI or machine learning application would most benefit their business, their top choice was generative AI (42% named it), followed by predictive analytics (39%). The third most-chosen option was “computer vision” (37%), or the ability to capture and analyze real-world data (such as scanning receipts to verify expense claims).

Buyers’ remorse? Though you’d expect accountants to be as careful with their software purchases as they are with their spreadsheets, that’s not always the case. Accounting software buyers were less likely than buyers in other industries to take steps like “defining their budget and must-have features, risk assessment, defining the desired outcomes, and [doing] market research and vendor shortlisting,” Jani said.

That less-than-methodical approach can lead to regret. Nearly half (44%) of accounting software buyers said they’d been disappointed in a purchase. And 36% of the time, the reason for their dissatisfaction was that they later found a product that was a better fit.

To prevent buyers’ remorse, Jani recommends taking a “a more process-driven approach” to software purchases.

“Especially in this time of AI functionality coming into the mix, you need to really try and pinpoint what the software is going to do for you,” he said. Buyers should think through “how is that functionality going to map onto your current situation?”