BambooHR's New Survey Says AI Is Burning 20 Workdays a Year
Small businesses keep hearing that AI will save time.
A new BambooHR survey says a lot of that time is being spent getting AI to behave in the first place.
The study, released Tuesday, found that workers spend the equivalent of 47 workdays a year using AI on average, but about 20 of those days are lost to troubleshooting errors and prompt iteration rather than productive output. BambooHR also said 63% of organizations have already raised their AI budgets in response to employee usage, even though the gains are still hard to prove. BambooHR press release
That matters for small and midsized businesses because they have less room to absorb wasted time. BambooHR is a people intelligence platform aimed at SMBs, so this is not a giant-enterprise edge case. It is a warning sign for the businesses most likely to adopt AI quickly and with lean teams.
The research surveyed more than 1,600 full-time salaried U.S. workers, including a subgroup of 520 HR professionals. Workers said they spend 87 minutes a day using AI on average. That adds up fast, but the report suggests the bigger problem is friction: fixing mistakes, rewriting prompts, and checking whether the output is usable before anyone can trust it.
That is the real story here. AI is no longer just a question of access. It is a question of operational drag.
The hidden cost is supervision
The survey found that 65% of workers feel confident and enthusiastic about using AI at work, and 58% say saving time is their main motivation. Those numbers do not point to resistance. They point to overuse before systems are ready.
For owners, that creates a familiar SMB problem. A tool can look cheap on paper, then get expensive once you count the cleanup. If the output needs a human to rewrite, verify, or repair it, the business is still paying for labor. It has just moved some of that labor into the review queue.
That is why this kind of data matters more than the usual AI hype cycle. The pitch is not "AI does nothing." The pitch is that AI works, but only after you build enough process around it to make the output reliable.
What owners should take from it
The BambooHR findings are not a reason to abandon AI.
They are a reason to stop pretending adoption is the same thing as value.
If a small business is going to use AI well, it probably needs three things:
- a narrow use case
- a clear owner
- a hard rule for when human review is mandatory
That is dull advice, but the survey makes the dull stuff look smart. The businesses that win with AI are likely to be the ones that treat it like a workflow problem, not a novelty purchase.
The bigger risk now is budget creep. If 63% of organizations are already increasing AI spend because employees are using the tools, leaders need to know whether the extra cost is buying better output or just buying more cleanup.
For small businesses, that answer will decide whether AI becomes a genuine edge or just another monthly subscription that asks for babysitting.