AI ROI: Why 56% of CEOs See Zero Return

Sixty percent of executives now use AI regularly to support business decisions, according to Deloitte’s 2026 Global Human Capital Trends. But a Forbes analysis of CEO survey data found that 56% of CEOs report neither increased revenue nor decreased costs from those investments. Only 12% achieved both.

The problem isn’t the technology. It’s what companies choose to measure.

Usage Is Not Value

Most organizations track AI the same way they track any new software: how many people are using it, how often, and how many departments have adopted it. These are input metrics. They tell you about activity, not impact.

A company might report that 80% of its team uses an AI analytics tool weekly. That sounds like success. But if nobody changed a decision, caught an error, or saved meaningful time because of those interactions, the tool is generating reports that feed other reports. Activity without outcome.

Forbes called this the capability overhang: the distance between what AI can technically do and what a business actually captures from it. The wider that gap, the harder it becomes to justify the next investment.

What Do Companies Seeing AI ROI Actually Measure?

The 12% of companies reporting real returns share a pattern. They don’t measure AI as a technology initiative. They measure it as a business capability, tied to specific workflows and specific outcomes.

Time recovered, not time saved. “Saving time” is vague. Companies seeing ROI track what happens with recovered capacity. If AI cut three hours from weekly reporting, did someone use those hours to close a deal, resolve an escalation, or analyze a new market? Recovered time that returns to meetings and email is invisible on the balance sheet.

Decision speed and quality. How long does it take to answer a question about margins, overdue receivables, or customer profitability? Before AI, the answer might have been “two days and a data analyst.” After, it might be “two minutes and a conversation.” The ROI isn’t the speed. It’s whether faster answers led to different actions.

Error and rework reduction. This is the most straightforward metric and the one businesses most often forget to track. If AI-assisted data validation cut invoice discrepancies from 4% to 1%, the dollar value of that reduction is real and auditable. In our experience working with mid-size businesses, this is usually where the clearest, fastest ROI appears.

Why Mid-Size Companies Have an Advantage

Enterprise AI projects fail at high rates partly because of their scale. When 58% of C-suite leaders report no clear ownership of AI and 75% lack governance programs, the measurement problem compounds across thousands of users and dozens of departments.

Mid-size companies can move faster because they have shorter feedback loops. The CEO asking “did this help us?” sits closer to the people using the tool. Ownership is clearer. The path from AI output to business action is shorter.

Companies reporting the strongest results tend to start with one measurable workflow, establish a baseline, run for 30 to 90 days, and calculate both hard savings and time redeployed. That discipline matters more than the sophistication of the AI itself.

Frequently Asked Questions

How do you measure ROI on AI?

Start with a specific workflow, not the whole business. Measure the baseline (time, errors, cost) before AI, then track the same metrics for 30 to 90 days after. Include all tool costs, and attribute gains to business outcomes like recovered capacity, fewer errors, or faster decisions.

Why do most AI projects fail to deliver ROI?

Most organizations measure adoption (how many people use the tool) instead of outcomes (what changed because of it). Without tying AI to specific business metrics, there’s no way to separate activity from impact. Governance gaps and unclear ownership compound the problem.

What is a good AI ROI timeline for small businesses?

Focused implementations on a single workflow often show measurable returns within three to six months. Broader deployments take longer. The key factor isn’t the timeline but whether you’re tracking outcomes that connect to revenue, cost, or capacity.

How Pluto Makes AI ROI Visible

Measurement gets simpler when AI connects directly to the data your business already runs on. Pluto sits on top of your existing ERP and lets you ask business questions in plain language: margin by customer, overdue invoices, cost trends by period.

Because every question maps to real data in your system, the path from question to decision is auditable. You can see what people asked, what they learned, and whether it led to action. That separates tracking usage from tracking outcomes.

See how it works or talk to our team.

What This Means in Practice

The 56% of CEOs seeing zero AI return aren’t necessarily using the wrong tools. They’re measuring the wrong things. Track what AI changes in your business, not how many people opened it this week.