
By Chuck Gallagher — Business Ethics Keynote Speaker and Trainer
TL;DR: Chuck Gallagher, AI ethics speaker and author, argues that small businesses should judge AI by customer results rather than output volume, counting correction time, customer confusion, and ongoing costs before claiming any productivity gain.
Picture a family-owned pool supply store heading into its busy season. The owner’s daughter runs marketing, and this spring she starts using AI. Last year the store sent two emails a month. Now it sends eight. There’s a new blog post every week and a social post every day. On paper, marketing output has more than doubled.
Then the phone starts ringing. One email promises 20% off chemicals. Another, sent two days later, offers a bundle discount that can’t be combined with the first. A blog post recommends a product the store stopped carrying. Customers walk in holding their phones, asking which deal is real. The counter staff spend half their morning explaining. Unsubscribes climb. Sales stay flat.
More got done. Less got accomplished. That’s the gap a lot of small businesses are about to fall into.
As an AI ethics speaker and author, I hear owners measure AI success the same way, by how much more they’re producing. Atlassian’s State of Product 2027 report suggests that’s the wrong scoreboard. Based on a Wakefield Research survey of 1,000 product professionals in the United States, Germany, and France, the report found that 80% say AI is helping them ship faster, but customers aren’t seeing value any sooner. And 69% agree that more effort has gone into producing faster while decision-making hasn’t sped up, even with AI.
Why Doesn’t Faster Work Automatically Create Customer Value?
Because customers don’t buy output. They buy results. The Atlassian report puts it in terms any shop owner would recognize: when production speeds up, the bottleneck moves somewhere else. Customers can only absorb so much change, and the people who sell, support, and maintain what you produce aren’t unlimited. In the same survey, 86% agreed that as AI makes producing easier, deciding what to build matters more, because customer attention is limited and every new feature adds maintenance and technical debt.
A small business feels that even faster. Eight emails a month don’t create four times the customer attention. They split the same attention four ways, and sometimes they wear it out entirely.
What Hidden Costs Do Businesses Miss When Measuring AI Productivity?
Speed gets counted. Cleanup usually doesn’t. The Atlassian survey asked what people have had to do with the time AI saved them. Forty-eight percent said they fix AI-generated work so it doesn’t look or feel machine-made. Thirty-nine percent said they manually verify AI outputs. And 61% of product managers said AI has created more technical debt, the kind of buildup that has to be paid down later, while fewer teams have set aside time to address it.
None of that shows up when someone announces that the team produced twice as much this quarter. It shows up later, in overtime, in confused customers, and in systems nobody has time to maintain. Don’t get me wrong. AI can save real time. The same report found 53% of respondents now have more time for strategic planning thanks to AI. But time saved and value created are two different numbers, and too many businesses only write down the first one.
When Does an AI Productivity Claim Become an Ethical Problem?
Let me be clear. Counting the output and ignoring the rework isn’t just sloppy math. When it’s reported to a boss, a board, a lender, or a business partner as proof that AI is paying off, it becomes a misleading claim. Maybe nobody meant to deceive anybody. But a number that leaves out half the story is still telling a story.
There’s a second ethical layer, and it’s the one owners forget. When AI output creates confusion, somebody absorbs the cost. Sometimes it’s the employee stuck fixing the work. Sometimes it’s the counter clerk explaining three conflicting promotions. And sometimes it’s the customer, who gave you their time and trust and got a muddle in return. Now, the way I see it, a productivity gain that’s quietly paid for by your customers isn’t a gain. It’s a transfer. And they’ll notice before your spreadsheet does.
How Can a Small Business Measure the Real ROI of AI?
You don’t need a consultant or a dashboard. You need four honest numbers, tracked together for at least a month on any task you’ve handed to AI.
The first is time saved, meaning how long the task used to take compared with how long it takes now. The second is correction time, meaning how many minutes go into checking, fixing, and redoing what the AI produced, including time spent by people other than the one who used the tool. Subtract the second from the first, and you have your real time savings. Sometimes that number shrinks. Sometimes it goes negative.
The third is customer outcomes. Pick something your customers actually feel, like sales, repeat business, response rates, complaints, or unsubscribes, and watch whether it moves. The fourth is ongoing cost, which covers subscriptions, setup, the time someone spends supervising the tool, and anything you’ll have to fix or maintain later. Only when all four point the right way can you honestly say AI is paying off.
Go back to the pool supply store. By the output measure, marketing doubled. By these four measures, the store spent more staff time, confused its customers, and sold nothing extra. The fix wasn’t to stop using AI. It was to send fewer, clearer emails and check every offer before it went out.
As an AI ethics speaker and author, I’ll leave owners with one test. Before you celebrate how much AI helped you produce, ask your customers whether they noticed. If they didn’t, you haven’t gained much. If they noticed for the wrong reasons, you’ve lost something.
Frequently Asked Questions
Does AI actually improve small business productivity?
AI can save real time, but faster output doesn’t automatically produce better results. In Atlassian’s State of Product 2027 survey of 1,000 product professionals, 80% said AI helps them ship faster, yet customers aren’t seeing value any sooner. Productivity should be measured by customer outcomes, not volume alone.
What hidden costs come with using AI at work?
Common hidden costs include time spent checking and correcting AI output, supervising AI tools, and maintaining what AI produces. In the Atlassian survey, 48% said they fix AI work so it doesn’t look machine-made, 39% manually verify AI outputs, and 61% of product managers said AI has created more technical debt. These costs rarely appear in simple productivity claims.
How do you measure the ROI of AI in a small business?
Track four numbers together for any AI-assisted task: time saved, correction time, customer outcomes, and ongoing costs. Subtract correction time from time saved to find real savings, then check whether sales, complaints, or repeat business actually changed. Count subscriptions, oversight, and maintenance as part of the cost.
Is it misleading to report AI productivity gains based on output volume?
It can be. Chuck Gallagher, an AI ethics speaker and author, argues that reporting increased output while leaving out rework, customer confusion, and ongoing costs tells only half the story. When that half-story is presented to a boss, board, lender, or partner as proof of results, it becomes a misleading claim.
Why isn’t faster work leading to better business results?
Speed moves the bottleneck rather than removing it. Customer attention, sales capacity, and maintenance time are limited, so producing more can overwhelm customers and staff instead of helping them. Atlassian’s report found 86% agree that deciding what to build matters more as AI makes building easier.
Bring This Conversation to Your Organization
If your business is producing more with AI but you’re not sure it’s producing better, that’s the right question to be asking. I help leadership teams, associations, and conference audiences look past the hype and measure what AI is really delivering, honestly, including the costs nobody wants to count. To bring a keynote or workshop on AI ethics and responsible adoption to your organization, connect with me, Chuck Gallagher, at ChuckGallagher.com.
Research source: Atlassian, The State of Product 2027, based on a survey of 1,000 product professionals conducted by Wakefield Research, August 13–26, 2026.
Five Questions for Reflection
1. How does your business currently measure whether AI is working, and does that measure include any customer outcome?
2. For one task you’ve handed to AI, how much time goes into checking and correcting the result, and who spends it?
3. Have your customers experienced more communication from you recently? Has it made their lives easier or harder?
4. If you reported your AI results to a lender or partner today, what costs might you be leaving out?
5. When does doing more stop being a sign of success and start being a sign that nobody decided what mattered?
