HP Spent 76 Years Chasing the Next Business. The Split Was an Admission It Caught Too Many.

It started with $538 and an audio oscillator. In a partnership formalized on January 1, 1939, two Stanford men built a test instrument so good that Disney bought eight of a modified version to mix the sound for Fantasia.⁷ That oscillator was a precise tool for a narrow job, and the company that grew around it was, for decades, a precise tool for a narrow job: build the best test-and-measurement gear in the world, and let the engineers run it. The trouble began when HP decided it should be more than that — and then decided it again, and again, for seventy-six years, until in 2015 it gave up and cut itself in half.¹

The official story of the 2015 split is a story of focus: a sprawling giant gracefully unbundling into two sharper companies, each free to pursue its own destiny. The real story is closer to a surrender. HP didn't split because two focused companies were a brilliant idea. It split because it had spent decades bolting on adjacent businesses it couldn't digest — and the seams finally gave way.

Each pivot looked like growth. Each one quietly leaked value.

Here is the pattern, and it is remarkably consistent. HP's core competence was hardware engineering — and its most successful product ever, the LaserJet launched in 1984, was exactly that: a beautifully engineered machine HP could dominate.⁸ But every time HP reached one rung up the value chain toward something stickier and higher-margin, it reached for a business that operated on logic its engineering culture didn't speak. Its early personal computers, the HP-85 in 1980 and the HP-150, both failed in the market — the HP-85 wasn't even compatible with the IBM PC standard that was busy becoming the whole game.⁸ HP could build a flawless instrument. It could not, it turned out, build a platform people had already standardized around someone else.

So when organic entry failed, HP started buying its way into adjacencies instead. In 2001 it agreed to absorb Compaq — an all-stock deal at 0.6325 HP shares per Compaq share, creating a colossus projected to do roughly $87 billion in revenue.⁴ The pitch was that scale in PCs and enterprise services would make HP a one-stop technology supplier. What it actually bought was a low-margin commodity PC business stapled to the side of a company that prized margins and engineering pride. The logic of the move was the logic of the conglomerate: more revenue, more end-markets, more of everything. The cost of the move was coherence.

Then came the most expensive reach of all. In 2011 HP paid $11.1 billion for Autonomy, a British enterprise-software firm, at a premium of roughly 58 to 64 percent over its pre-announcement price — for a company that had never booked more than $1 billion in annual revenue.⁵ This was HP trying, finally, to buy its way out of hardware and into the recurring, high-margin economics of software. It paid a software premium for a software dream. Then, on November 20, 2012, barely a year later, it wrote off $8.8 billion of the value.⁶

The number everyone repeats is wrong — and the wrong number hides the real lesson

You will hear, endlessly, that HP 'paid $11 billion for Autonomy and wrote it all off.' That is two errors in one sentence. The $11.1 billion was the price; the write-off was $8.8 billion.⁵⁶ And the $8.8 billion wasn't all the accounting scandal that made headlines. HP's own CFO, on the earnings call, attributed roughly $5 billion of it to alleged accounting improprieties and about $3.8 billion to the drop in HP's own stock price.⁶ The independent valuation expert Aswath Damodaran went further, arguing the fraud explanation could account for at most around $2.45 billion of the gap.⁶

Strip away the fraud allegations and a plainer fact remains: HP paid a software multiple for a sub-billion-dollar revenue company because it desperately needed to be something other than a hardware company. That is not a due-diligence failure. It is a strategy failure wearing a due-diligence costume. The premium was the symptom. The disease was a company that had spent decades convinced its next identity was always one acquisition away.

Why the split was an admission, not a strategy

By 2015 HP was carrying the residue of every reach it had ever made: PCs and printers on one side, servers, services, and the wreckage of its software ambitions on the other. The board approved the separation on September 30, 2015, and on November 1 it distributed 100 percent of the enterprise shares to stockholders — one HPE share for every HP share held.¹² The filed rationale is almost touchingly honest once you read it the right way: simplify a complex business structure so each company can 'concentrate its financial resources solely on its own operations' and create 'two distinct and targeted investment opportunities.'³

Read that as a press release and it sounds like vision. Read it as a confession and it sounds like this: we put these businesses together over decades, and they were never one company, and now we are paying to separate what we paid to combine. 'Simplify a complex structure' is corporate for 'we made it too complicated to run.' The split didn't create focus. It restored a focus the adjacency-chasing had spent thirty years dismantling — and shareholders had to pay the separation costs to get back something the company had owned, almost for free, when it was just making the best oscillator in the world.

The honest case that the split actually worked

The fair objection is that this reads too neatly as failure, because the split was, on its own terms, the right call. Two focused companies could indeed allocate capital better than one sprawling one — that's not spin, it's basic corporate finance, and plenty of conglomerate breakups have unlocked real value. By that logic, every pivot was just HP searching, and the split was HP finally finding. There's truth in it. A company that never reaches never grows, and some of HP's reaches — the printing franchise that grew out of the LaserJet — were genuine triumphs.⁸

But notice what the steelman concedes. If splitting was the right call in 2015, then combining was the wrong call somewhere along the way — and the value 'unlocked' by separation is, definitionally, value that the combination had been suppressing. You don't get credit for curing a disease you gave yourself. The Compaq dilution and the Autonomy write-off weren't the price of healthy exploration; they were the price of a hardware company that kept trying to escape being a hardware company by buying businesses it couldn't operate. The split was the bill arriving for all of it.

HP began as a tool that did one thing superbly, and it spent its life convinced that doing one thing superbly wasn't enough. Each reach up the value chain — into PCs, into services, into software — was an attempt to become a grander kind of company, and each one quietly subtracted from the thing it already was. The 2015 split is remembered as a bold act of strategic clarity. It was something humbler and more instructive: the moment HP stopped trying to be everything it had bought, and admitted it was worth more as the two things it had actually become. The oscillator knew what it was for. It took the company seventy-six years and several billion dollars to relearn the lesson.