IBM recently had the worst trading day in its 114-year history. On July 14, shares cratered 25% in a single session, erasing nearly $69 billion in market value. The cause wasn’t a scandal or a product failure. It was AI ROI, or the lack of a clear story around it.
CEO Arvind Krishna told investors the company simply did not adapt fast enough as clients redirected their budgets toward AI infrastructure, leaving software and consulting deals unclosed. That single admission wiped out a year of stock gains in one morning. If you lead a team, a company or even your own client roster, this is the clearest signal yet that “we’re investing in AI” is no longer a good enough answer.
Why IBM Just Had Its Worst Day in Company History
IBM’s revenue miss itself was small, only about 3.7% below forecast. What triggered the crash was Krishna’s own words. He admitted IBM’s customers were pulling spending away from software and consulting to chase scarce AI infrastructure, and IBM hadn’t repositioned in time to capture that shift.
So what does this mean for you? Investors, and increasingly customers, are no longer rewarding companies for simply participating in the AI boom. They’re punishing the ones who can’t explain where the value actually landed. A vague AI story used to be enough to keep a stock, or a budget, afloat. That’s no longer true.
You’ve probably heard leaders describe this moment as an “AI bubble.” That framing misses what actually happened at IBM. Demand for AI infrastructure is real and growing. What burst wasn’t the technology; it was the assumption that simply having an AI strategy was worth something on its own, regardless of whether it produced a measurable result.
The Split Screen: Silence vs. Receipts
The most revealing comparison isn’t IBM against its competitors. It’s two of the biggest AI spenders in the world reporting earnings the same week with opposite results. When an analyst asked Meta CEO Mark Zuckerberg about return on the company’s $145 billion AI investment, he called it “a very technical question.” The stock dropped 10% overnight, on a quarter where revenue actually rose 33%.
Alphabet told a different story. In April, the company reported Q1 2026 earnings, highlighting a Google Cloud backlog that nearly doubled quarter-over-quarter to more than $460 billion in contracted future revenue. With Google Cloud revenue up 63%, Alphabet’s stock jumped, delivering its best month since going public in 2004.
But it doesn’t stop there. Both companies are spending massively on AI. Both are led by capable executives. The only meaningful difference was whether they could point to a number and say, plainly, this is what the spending bought us.
Zoom out and the pattern holds across the rest of Big Tech’s earnings season too. Companies that paired AI capital expenditures increases with a concrete outcome—a backlog figure, a margin improvement, a specific product metric—generally held up. Companies that asked investors to simply trust the vision generally didn’t. That’s not a coincidence, and it’s not going away next quarter.
What ‘Show Me the ROI’ Really Means for Leaders
Research shows this pattern isn’t isolated to a single earnings season. Across Big Tech, the companies that pair AI spending with specific, measurable outcomes are being rewarded, while the ones offering vision without metrics are getting punished, sometimes within hours of saying so out loud.
The key is to recognize that “AI ROI” doesn’t mean a single tidy number. For Alphabet, it was a backlog figure, contracted revenue investors could actually verify. For your business, it might be hours saved, error rates reduced or revenue tied directly to an AI-assisted process. The specific metric matters less than having one at all.
Industry leaders note that the market’s patience for “trust us” AI narratives is running out fast, and that shift didn’t take years. It took one earnings season.
How to AI-Proof Your Own ROI Story
Start by picking one AI initiative your team or business has already launched and ask a simple question: What changed because of it? Not what it does, but what actually changed in hours, dollars or output.
If the honest answer is “we’re not sure yet,” try this approach: Set a 90-day checkpoint on every active AI initiative, with one metric attached before you spend another dollar on it. That single habit puts you in Alphabet’s category instead of Meta’s the next time someone asks you the technical question out loud.
You don’t need billions invested for this lesson to apply. A solo consultant using AI tools to serve clients faces the exact same test, just at a smaller scale. Can you tell a client specifically what your AI-assisted process saved them or are you also answering with a version of “that’s technical”?
This applies just as directly if you’re the one being pitched. The next time a vendor, agency or new hire tells you they’ll “bring AI” into your business, ask the same question the market just asked IBM and Meta. What’s the number, and when will you have it? If the answer is vague, treat that as information, not a technicality to work around later.
Your Next Move
The market didn’t punish IBM for spending on AI. It punished the company for not being able to explain where that spending went. That distinction is the entire lesson buried in this week’s stock charts.
Before your next AI investment, board update or client pitch, write down the one number you’d point to if someone asked for your ROI. If you can’t fill in that blank yet, that’s this week’s actual priority, not the next tool or model upgrade.
IBM will get another chance to answer that question. So will you, probably sooner than you expect. The leaders who treat that as useful pressure, rather than an inconvenience, are the ones who’ll still be trusted with a bigger budget next year.
Featured image from Mijansk786/Shutterstock
