Three major reports landed this quarter. Deloitte, PwC, and MIT Sloan all looked at the same question: is AI actually delivering?
The honest answer is... not yet. And that might be exactly right.
74% of companies say they expect AI to drive revenue growth. Only 20% report that it actually has. That's a 54-point gap. But before you read that as failure, consider what else those reports found: the companies seeing returns aren't the ones who spent the most. They're the ones who started with a clear problem and gave the investment time to compound.
This isn't a new pattern. Every major technology shift works this way.
Cloud computing took years to show ROI. So did ERP systems, e-commerce, even electricity in factories. The return was real. It just lagged the investment.
Microsoft, Google, Amazon, and Meta are collectively spending over $200 billion a year on AI infrastructure right now. Not because the revenue is there today. Because they know the cost of waiting is higher than the cost of being early. Sundar Pichai calls it "investing ahead of revenue."
These aren't reckless bets. They're strategic ones with a known lag time built into the thesis.
PwC pointed to where things go wrong: "Crowdsourcing AI efforts can create impressive adoption numbers, but it seldom produces meaningful business outcomes."
MIT Sloan found the same thing. Most companies are still using AI at the individual level. One person drafts emails faster. Another summarizes meeting notes. Useful, sure. But scattered adoption doesn't add up to business-level results. There's no way to connect the tool to the outcome.
It's like putting money into a retirement account and checking the balance every week. The math works. It just doesn't look impressive on a short timeline. The companies investing in AI right now are making deposits. The compounding hasn't kicked in yet. And it won't, if every deposit goes into a different account with no strategy behind it.
That's not an AI problem. That's a clarity problem.
Here's what this means for businesses.
The investment is needed now. The return will follow, but only if you start with intention.
The companies in those reports that can't find ROI aren't proof that AI doesn't work. They're proof that spending without a defined outcome doesn't work. And that's a mistake smaller businesses can actually avoid, because there are fewer layers between the question and the decision.
You don't need a seven-figure AI budget. You need to know which problem you're solving, how you'll measure progress, and the patience to let the investment mature. A 15-person company with that clarity will outperform a 5,000-person company spreading bets across every department.
The ROI is real. It's just a lagging indicator. The companies that understand that, and invest with discipline anyway, are the ones building something that compounds.