# The Five Years Exxon Tried to Beat IBM — Notes on an Asianometry Episode

> Asianometry's 2026-09-17 episode tells how an oil giant walked into the office equipment business and squared off against IBM. These are listening notes on separating a diagnosis from its prescription, how resources turn into ropes, and one question for deciding whether something is worth your time. Educational, not investment advice.

Published: 2026-09-18
Locale: en
Tags: Asianometry, diversification, tech history, Zilog, investing mindset


![An office floor at dusk, an early-1980s electronic typewriter on a desk in the foreground, a corridor running deep toward a window where refinery flare stacks glow](/covers/asianometry-2026-09-17-exxon-s-office-fling-cover.png)

> The long is never in excess, the short is never lacking. A duck's legs are short, yet stretch them and it grieves; a crane's legs are long, yet cut them and it mourns.
>
> —— Zhuangzi, "Webbed Toes" (Warring States period; translated by the author)

## What the Episode Covers

On 2026-09-17, Asianometry posted an episode called "Exxon's Office Fling," about the late 1970s, when Exxon declared war on IBM.

Not as a metaphor. Exxon folded fifteen acquired technology companies into a division called Exxon Information Systems — word processors, typewriters, fax machines — aiming to become a major player in office equipment within three to five years. One of those companies was Zilog, maker of the Z80 processor.

My first reaction was a laugh. My second was a chill, because every step of this adventure had a solid reason behind it at the time.

## The Points Worth Keeping

**One: the diagnosis was well written.** In 1960, Standard Oil of New Jersey asked finance executive David Jones to produce a report on investment policy. It concluded that oil and gas had stopped being a growth business and that future opportunities there would show ever-worsening returns. The report was right about the setup: posted crude stayed stuck at two to three dollars a barrel until 1970. In 1958 the company sold more oil than ever and earned a third of what it had at the start of the decade.

**Two: the first diversification was chemicals, and it bought its lessons with losses.** The board lifted the chemical subsidiary's annual investment budget from an average of 53 million dollars in 1961–63 to 257 million in 1964. Executives took the money around the world — Central America, the Caribbean, the Philippines, Malaysia, Pakistan, Lebanon, Greece, Spain, the Netherlands, Canada. CEO Rathbone told Forbes at the time: "Everybody and his brother seems to be making polyethylene. But this dashing into chemicals is going to lose a lot of money for some companies … And it won't be us." By 1965 the chemical arm returned 4% on net assets. He had already complained in late 1963 about being asked to approve new plants in Cyprus and Colombia that he had never heard of.

**Three: the second diversification scattered seeds, and one of them was Zilog.** Jersey Enterprises, founded in 1963, picked ten categories on one criterion: could this birth a hundred-million-dollar company? Industrial cryogenics was bought and dropped within four years. A venture with Nestlé's Guatemalan affiliate grew bacteria on hydrocarbons to make protein powder; chicks fed on it performed "poorly." Petroleum residues bound into building blocks cost as much as cement and could not be shown to last. A few categories survived the winnowing, one of them information technology. In 1975 Exxon became Zilog's sole investor. Cofounder Ungermann's pitch to customers: "the money of Exxon and the brains of Intel."

**Four: the size gap made every success irrelevant.** Word processor maker Vydec did roughly 24 million dollars of revenue in 1977, a respectable figure. Exxon's oil business did 48.6 billion that year. In 1980 one analyst forecast the office systems business would eventually rival or surpass the oil business. The other side's arithmetic was shorter: grow the division to the size of Xerox and you add seven billion in revenue.

**Five: merging fifteen small companies produced a forty-page org chart.** Exxon Information Systems opened with 6,000 employees. Wang Labs EVP John Cunningham: "If God himself came down in a junior businessman's uniform, worked very hard for three years, and got very lucky, he'd still have a tough time pulling those companies together."

**Six: the calendar stood on the other side.** Apple introduced the Apple II in April 1977, and VisiCalc turned it into a business machine. The IBM PC arrived in 1981. Exxon had bought into old-line office equipment on the eve of the PC revolution. The second-generation QWIP fax took six years to develop and arrived dead — the paper feeder jammed. A Qyx "intelligent" typewriter caught fire inside Exxon's own New York headquarters. Exxon's Houston marketing division said outright they preferred Wang, were forced to take company equipment, and sent it back a year later.

**Seven: the identity itself carried a cost.** Because Exxon was now a competitor, IBM was never going to pick Zilog's Z80 for the PC. Faggin later confirmed this weighed in the choice of the Intel 8088. In 1984 Exxon admitted what everyone else could see, sold what it could and closed the rest; hundreds of leftover QWIP machines went to a dump and were bulldozed. Zilog left in a late-1980s management buyout backed by Warburg Pincus at about four times free cash flow, went public in 1991, and was acquired in 1997 for 527 million dollars — twelve times the private equity firm's 17 million dollar investment.

## Thinking It Through

### A company announces it is entering a hot new field — how should I look at that

I used to start by checking the size and growth rate of the new market. Listening to this episode, I think that order is backwards.

Start with the denominator. Exxon's office division turned 100 million in 1978 and 200 million in 1979. Graph it and you get a steep line. Drop it into the parent's income statement and you get a rounding error against 84 billion of oil revenue.

The harder question fixes the success case first: suppose this works as well as it possibly could — how much does the parent gain? That is exactly how the skeptical analysts ran it in 1980. Xerox-sized, plus seven billion. The probability of getting there is a separate discussion. When something cannot move the needle even in its best case, the quarterly progress updates in between are not worth reading.

The same arithmetic explains the irony in Zilog's ending. Zilog eventually returned twelve times for the private equity firm that bought it. That result meant nothing to Exxon and meant a career-defining outcome to a 17 million dollar position. Same company, same performance, two different worlds depending on the denominator.

### The claim can be right while the plan is wrong

The Jones Report's diagnosis holds up. Oil prices stayed low for a decade and the chemical opportunity was there. The break comes at the step from diagnosis to prescription: "oil and gas has stopped growing" does not get you to "therefore we should build office equipment."

What gets skipped is capability. What was the oil business good at? Executing decade-long, multi-billion-dollar projects on every continent, through discipline, hierarchy and prudence. Analysts in 1980 cited that discipline as a reason Exxon could win. Faggin described the same discipline from inside information technology: "anytime there was a decision to be made, the lower guy on the totem pole would look up, the other guy would look up at the next, and Ben Sykes who was the senior guy would say, 'Yes,' or 'No.'"

I use this as a checkpoint. When a judgment sounds persuasive, split it in two and score the halves separately: the description of the world (oil will stay cheap) and the action derived from it (so we should diversify into X). The first half can be checked against evidence. The second half hides a pile of unstated assumptions. The holes I have fallen into were mostly in the second half.

The expiry date matters too. The report assumed the glut would continue, and then 1973 arrived. The premise behind the transformation expired before the transformation finished.

### Is a deep-pocketed backer good news for the company receiving the money

My favourite stretch of the episode is how the Zilog–Exxon relationship went from honeymoon to leash.

Faggin has acknowledged Zilog would never have gotten off the ground without Exxon. Total technology investment across all of 1975 came to about ten million dollars; the venture market had dried up after the early-1970s bubble, and Exxon Enterprises was the only game in town. Exxon money later funded their fab.

Then the same resource turned. Former employee Bernard Pueto's account: the oil giant gave them too much money and too many directions, pushed Zilog into things it should not have done, and the young managers lacked the experience to know when to push back. Before long Faggin spent more time in New York keeping Exxon out of the way than in California working with customers. The friction over Exxon frayed his partnership with Ungermann, who left.

So when a company lands a large backer, alongside "is the money enough" I want to ask whose agenda rides along with it, and which doors that identity closes. The Z80 was locked out of the IBM PC while being good enough that the TRS-80 ran on it and people still write about it today. Zilog had no vote on that.

## Where to Read More

- Asianometry, "Exxon's Office Fling," 2026-09-17, the starting point for this piece
- The Computer History Museum's oral history with Federico Faggin, first-hand on the Zilog–Exxon relationship
- Material on the 1911 Standard Oil breakup, for why Jersey Standard held no crude production of its own
- Exxon's annual reports from the early 1980s, to see how a division describes itself in the two years before it closes

## One Thing to Take With You

Whether something is worth your investment depends on how much it changes you on the day it succeeds, not on how fast it is growing now.

Here is a practice I have tried. Take a sheet of paper and write down the three things that consumed the most of your time this month — a project at work, a skill you are learning, a relationship you are maintaining all count. Next to each, write one sentence in this form: "If this goes as well as it possibly can, the concrete difference in my life is ______." Concrete enough that another person could picture it. "I'll feel more fulfilled" does not count. "I stop working late every Friday" counts.

Whichever one you cannot fill in, cut its time in half this month. The item I cut when I last ran this was something I had been talking about since January and feeling guilty about every few weeks.