THE THETA LENS · ONE COMPANY, SEVEN ERAS, SEVENTY YEARS

IBM had world-class talent the whole time. What changed was where each leader pointed it.

No company reports its Core, Edge, and Beyond split, because until now the map didn't exist. So what follows is the first time IBM's history has been read this way: not by revenue or stock price, but by where leadership placed its bets on the innovation map, and what happened a decade later.

Core — the proven engine
Edge — the next S-curve
Beyond — the wild bet
How to read this. The bars estimate where each CEO placed investment and attention, not revenue — revenue always lags a decade behind the bet. Figures are directional, built from R&D mix, acquisitions, and strategic focus. The pattern matters more than any single percentage.
Thomas Watson Jr.
1956–1971
Bet Beyond from strength

DRIVER · personal conviction, from a position of strength Inherited a healthy business and risked it anyway. The System/360 was a ~$5B bet — more than IBM's annual revenue — to unify computing into one compatible family when the industry said it couldn't be done. Nobody forced the gamble.

Impact: Built an empire. The 360 became the mainframe standard and locked in enterprise computing for a generation. The Beyond bet became the Core that funded everything for the next 25 years.

Cary & Opel
1973–1985
Milked Core, leaked the Edge

DRIVER · dominance & complacency IBM was printing money from mainframes, so leadership optimized the Core. The one Edge move, the IBM PC (1981), happened despite the company — a skunkworks team in Boca Raton, built fast and open, outside IBM's process.

Impact: A won Edge, given away. Outsourcing the OS to Microsoft and the chip to Intel created the two companies that would eat IBM's lunch. The value leaked to partners because a Core-optimized company didn't know how to own an Edge win.

John Akers
1985–1993
Defended the Core into near-death

DRIVER · denial · a culture that couldn't imagine losing Doubled down on mainframes and proprietary systems exactly as the market shifted to PCs, open architectures, and client-server. Not a bet from strength like Watson Jr. — a reflex, defending the Core out of institutional habit while the ground moved.

Impact: An $8B loss in 1993 — the largest in US corporate history at the time. The stock collapsed, the board weighed breaking IBM apart, and Akers was forced out. The framework's central warning, at maximum scale: died from overconfidence in the Core.

Lou Gerstner
1993–2002
Redefined the Core to survive

DRIVER · crisis · outsider clarity First non-IBM CEO ever, hired to save or dismantle the company. Famously said "the last thing IBM needs right now is a vision" — deliberately low on Beyond because the company was bleeding out. The move was a Core redefinition: stop selling hardware boxes, start selling services and integration.

Impact: Stopped the dying. Killed the breakup, restored profitability, rebuilt IBM around services. But the recovery started here; it didn't finish here.

The correction most people get wrong: disruption of a Core business is never instantaneous. Gerstner is remembered as the man who "saved IBM in the '90s," but changing a Core business and the culture around it is a generational project. The pivot he began in 1993 took roughly three decades to fully pay off — arguably only firing on all cylinders now.
Sam Palmisano
2002–2012
Good numbers, thinning engine

DRIVER · strategy, bent by financial engineering Sold off commoditizing businesses (PCs to Lenovo, 2005) and pushed "Smarter Planet" as Edge. Real discipline — but increasingly bent toward the "2015 Roadmap" and its $20 EPS promise, so allocation began serving the number rather than the future.

Impact: Strong on paper (EPS up ~5×, $100B+ shareholder value) — but much of the growth was bought with buybacks, not earned in new markets. Good-looking numbers masking a thinning engine.

Ginni Rometty
2012–2020
Judged a Beyond bet like Core

DRIVER · right allocation on paper, wrong execution Intended a big Edge/Beyond push (Watson AI, cloud), but the numbers didn't translate to shipping. The one move that mattered: buying Red Hat for $34B (2019) — a genuine Core-rebuild that seeded the recovery, though the harvest came later.

Impact: 22 consecutive quarters of declining revenue. The stock badly underperformed. The "moonshots live in PowerPoint decks" era.

The Watson mistake, in one line: Watson was a Beyond bet — speculative AI, a 7–15 year payoff — but it was marketed and measured like a Core product, expected to book revenue and prove itself quarter by quarter. Judged by Core metrics, a Beyond bet always looks like failure. Watson didn't fail because the technology was worthless. It was strangled by expectations that belonged to a different zone.
Arvind Krishna
2020–present
Rebalanced all three

DRIVER · deliberate rebalancing, clear-eyed on the situation Spun off the dead-weight infrastructure business (Kyndryl, 2021), made software the Core engine, pushed Watsonx as real Edge, and turned quantum into a serious roadmapped Beyond with a $10B commitment. The most balanced allocation IBM has had since Watson Jr. — the first time in decades all three zones are genuinely funded and moving.

Impact: Turnaround in progress. Core growing on software, the Edge shipping without chasing the AI frontier, and the industry's most credible quantum roadmap. The market rewarded the story in 2025 and punished a soft quarter in 2026 — but underneath both reactions, the direction is genuinely reversed for the first time in a generation.

The 30-year Core redefinition. Read across the timeline and one slow current runs under everything: the services-and-software Core that Gerstner began in 1993 only truly comes good under Krishna, thirty years and four CEOs later. Changing what a company is at its center is not a quarter's work or a CEO's work. It is generational.

The variable was never talent. It was allocation.

IBM had world-class R&D and brilliant people in every one of these eras. What changed, decade to decade, was where leadership pointed them on the Core / Edge / Beyond map — and the zone mix predicts the outcome a decade later better than almost anything else.

Watson Jr. bet Beyond and built an empire. Akers defended the Core and nearly killed it. Gerstner redefined the Core and bought survival. Rometty let the bets drift and delivered 22 quarters of decline. Krishna rebalanced, and the market turned.

Note the nuance, because it keeps the framework honest: this isn't "Beyond good, Core bad." Gerstner's near-zero Beyond was the correct move — for a company that was dying, the right call was redefining the Core, not chasing a moonshot. The fatal error is narrower and more specific: defending the Core out of habit when the world has already moved. That error has a name in this story. It was called 1993.