You feel like AI is saving you time. But is it? Or are you just busier in a different way?
QuickBooks asked 34,000 small businesses that question in their 2026 AI Impact Report, conducted with the University of Chicago. The headline numbers look great: 77% use AI regularly, and 41% report revenue increases. A separate Goldman Sachs survey of 10,000 small businesses found 93% report positive impact from AI — but only 14% have fully integrated it into core operations.
Then they asked how those businesses were measuring those gains. More than half said they had “a general feeling” that things were better. Less than half tracked any specific metrics. The productivity figure — 74% improvement — was based entirely on self-reporting, not time studies.
Gartner estimates global AI spending will hit $2.52 trillion in 2026. A significant portion of that is being spent without any method to measure the return. Forrester predicts 55% of AI projects this year won’t meet their intended goals — not because the AI failed, but because nobody defined what success looked like before they started.
Why Measurement Breaks Down
Two problems kill AI ROI tracking before it starts.
The first is missing baseline data. If you don’t know how long a task took before AI, you can’t calculate what the AI saved. Most businesses adopt a tool, feel faster, and call it a win. “Feels faster” isn’t a number you can act on.
The second is attribution. If you started using an AI email tool and hired a new sales rep in the same quarter and revenue went up — what caused it? You can’t know. Both changed at once. The AI gets credit it may not deserve, or gets blamed for results the market drove.
Five Numbers That Actually Tell You
These metrics come from TerDawn DeBoe of the Institute for Business AI. Each one can be tracked with a stopwatch and a spreadsheet.
1. Time per task. Pick three tasks you regularly use AI for. Time them with AI. Then time the same tasks done manually. Multiply the difference by your hourly rate. If a 90-minute task now takes 20 minutes and your time is worth $100/hour, that’s $116 in savings per use — a real number you can defend.
2. Output quality. Track how much editing your AI-generated content needs. If 80% of it ships with minor tweaks, the AI is delivering value. If 80% needs a full rewrite, you’ve changed your workflow without gaining productivity. Be honest about this one.
3. Revenue per AI-supported activity. Run a controlled comparison: one campaign drafted by AI, one drafted by hand, similar lists, similar offers, same time period. The revenue difference is your AI marketing ROI. If they’re equal, the AI saved you drafting time — useful, but different from driving revenue.
4. Error rate. For tasks like data entry, bookkeeping, or report generation, track how often you have to correct AI output versus how often you corrected your own work before. Fewer corrections is a measurable improvement. More corrections means the AI is creating a new category of work.
5. Cost vs. value. Add up what you’re paying for AI subscriptions monthly. Stack that against the dollar value from metric 1 (time saved multiplied by your hourly rate). If the math doesn’t work, you’re paying for the feeling of productivity, not the reality of it.
Start With One
You don’t need all five. Pick the one that matches how you most commonly use AI right now, and start timing things this week. Within a month you’ll have actual data instead of a feeling.
The businesses getting real ROI from AI aren’t the ones using the most tools. They’re the ones who defined what “working” meant before they started.
Data source: QuickBooks 2026 AI Impact Report (34,000 U.S. small businesses, in partnership with University of Chicago). Analysis via Forbes contributor TerDawn DeBoe, Institute for Business AI.