Tom Ellsworth says IBM has 135,000 employees in India and only 40,000 in the U.S.

Tom Ellsworth says IBM has 135,000 employees in India and only 40,000 in the United States. That was the real grenade inside the PBD Podcast segment on IBM’s ugly earnings reaction, because it turned a stock-chart faceplant into a question about whether Big Blue has become something closer to Big Bangalore.

The headline number was brutal enough. The panel said IBM lost $72 billion in market cap in two days after pre-announcing earnings below Wall Street expectations, with management blaming, in part, customers shifting capital spending toward servers, storage, and memory to get ahead of supply constraints. Fine. Markets panic. Analysts overreact. CEOs say “buying patterns” and everyone pretends this is English.

But Tom Ellsworth wasn’t buying the idea that this was just a guidance problem. His claim was sharper and more combustible: IBM’s earnings wobble is sitting on top of a workforce and morale problem, with American employees watching jobs and internal gravity move overseas.

Its own people are referring to it as Indian business machines rather than International Business Machines.

The offshoring claim is the part IBM won’t want clipped

Ellsworth’s numbers were the kind that get screenshotted in Slack channels and fought over on Reddit. He said one side of IBM has 40,000 employees and the other has 135,000, then let the panel guess which was which. The answer, according to him: India is the 135,000. The U.S. is the 40,000.

IBM USA only has 40,000 employees. It’s been shifting. It’s been shifting on H-1Bs.

He then added a second data point: current openings. Ellsworth said he checked that morning and found 2,840 IBM job openings in India versus 376 in the U.S. That doesn’t prove a one-for-one replacement machine. Job postings are a snapshot, not a confession. But as a morale story, it’s potent. If you’re an American IBM employee training colleagues abroad while the company posts far more roles there than here, you don’t need a McKinsey deck to understand the vibe.

This is where the segment got more interesting than the usual “old tech company missed earnings” cable-news slop. The panel wasn’t just asking whether IBM had a bad quarter. It was asking whether IBM still knows what it is. That’s a mean question to ask about a company that used to be the future. It’s also fair.

there’s more to it than just this earnings report. IBM has got some issues inside

Patrick Bet-David thinks this is an Andy Grove moment

Patrick Bet-David’s read was less inflammatory but probably more useful. He kept pressing the table on a basic question that sounded almost rude: What is IBM’s flagship product today? Not what did it used to dominate. Not what does it sell to banks and government agencies. What does IBM own in the public imagination now?

The answers came back as a stack of enterprise terms: mainframes, Red Hat OpenShift, Watsonx, consulting, cloud, security. All real businesses. All serious money. Also, none of them exactly has “iPhone launch line around the block” energy. IBM is still huge, but huge is not the same as legible.

Bet-David reached for the Andy Grove and Gordon Moore story at Intel, the one about walking out the door, coming back in as new management, and deciding to exit memory chips. The point wasn’t nostalgia. It was that a giant company sometimes needs to fire its old self before the market does it with a flamethrower.

What’s next? What what business do they need to get out of? What business do they need to get in that they’re not?

That’s the stronger argument. The outsourcing claim will travel because it is specific, angry, and easy to Google. The reinvention claim is the one IBM’s board should care about. Enterprise customers may still pay IBM enormous sums for mainframes, consulting, hybrid cloud, and security, but investors are not grading tech companies on survival anymore. They’re grading them on whether they look like winners in artificial intelligence.

Verdict without the corporate fog machine

Ellsworth’s offshoring argument is plausible and pointed, but he overstates what the evidence can prove. Employee counts and job openings can show where a company is placing its bets. They don’t, by themselves, prove that IBM’s earnings miss came from morale collapse or that American workers are being systematically swapped out in every case.

Still, the segment found the pressure point. IBM’s problem is not that nobody knows the company. It’s that too many people know it as a monument. A monument can still be worth billions. It can still have government contracts, bank clients, and a respectable consulting arm. But when a tech company has to explain why it matters, the earnings call has already gone sideways.

If Ellsworth’s numbers are right, the next question for IBM is not just whether it can recover the market cap. It’s whether the company that helped define American computing can convince American workers, and investors, that they’re still part of the plan.