Escolar Documentos
Profissional Documentos
Cultura Documentos
The FT’s Undercover Economist on the real reasons that corporate innovation dies
Tim Harford
In 1990, a young economist named Rebecca Henderson published an article with her supervisor Kim Clark that
presented a different view of why it is hard to do new things in old organisations. The relevant word is “organisations”.
Dominant organisations are prone to stumble when the new technology requires a new organisational structure. An
innovation might be radical but, if it fits the structure that already existed, an incumbent firm has a good chance of
carrying its lead from the old world to the new.
Consider, for example, IBM — the giant of mainframe computing. IBM is a survivor. It predates the digital computer by
more than three decades. While the performance of computers was being revolutionised by the semiconductor, the
integrated circuit, the hard drive and the compiler, IBM maintained a dominant position without breaking stride. This
was because the organisational challenge of making and selling a sophisticated mainframe computer to a bank in the
1970s was not greatly different from the organisational challenge of making and selling a mechanical tabulating
machine to a bank in the 1930s. Change was constant but manageable.
When computers started to be bought by small businesses, hobbyists and even parents, IBM faced a very different
challenge. It did build a successful business in PCs, but was unable to maintain its old dominance, or bring to bear its
historical strengths. In fact, the PC division prospered only as long as it was able to snub the rest of the organisation,
often partnering with component suppliers and distributors that directly competed with IBM divisions.
Ideas that got stuck: Sony's Memory Stick Walkman; a Kodak SLR digital camera (Alamy); a Medium-D tank
Internal politics soon asserted itself. A case study co-authored by Henderson describes the PC division as “smothered
by support from the parent company”. Eventually, the IBM PC business was sold off to a Chinese company, Lenovo.
What had flummoxed IBM was not the pace of technological change — it had long coped with that — but the fact that
its old organisational structures had ceased to be an advantage.
Rather than talk of radical or disruptive innovations, Henderson and Clark used the term “architectural innovation”.
“An architectural innovation is an innovation that changes the relationship between the pieces of the problem,”
Henderson tells me. “It can be hard to perceive, because many of the pieces remain the same. But they fit together
differently.”
An architectural innovation challenges an old organisation because it demands that the organisation remake itself.
And who wants to do that?
The armies of the late 19th century were organised — as armies had long been — around cavalry and infantry.
Cavalry units offered mobility. Infantry offered strength in numbers and the ability to dig in defensively.
Three technologies emerged to define the first world war: artillery, barbed wire and the machine gun. They profoundly
shaped the battlefield, but also slipped easily into the existing decision-making structures. Barbed wire and machine
guns were used to reinforce infantry positions. Artillery could support either cavalry or infantry from a distance.
Tanks, however, were different. In some ways they were like cavalry, since their strength lay partly in their ability to
move quickly. In other ways, they fitted with the infantry, fighting alongside foot soldiers. Or perhaps tanks were a new
kind of military capability entirely; this was the view taken by J F C Fuller.
These discussions might seem philosophical — but in the light of Henderson’s ideas, they are intensely practical. “You
have to find an organisation that will accept the new bit of technology,” says Andrew Mackay. Mackay runs an
advisory firm, Complexas, but was also the commander of British and coalition forces in Helmand, Afghanistan, in
2008. “The organisational question is deeply unsexy, but it’s fundamental.”
A more recent example: is the helicopter basically a kind of aeroplane, and therefore an asset of the Royal Air Force?
Or something quite different? Who should be in charge of drones today?
So it was with the tank. If it was to prosper, it needed an organisational home. Someone would have to argue for it,
someone would have to pay for it, and someone would have to make it all work, technologically and tactically.
Perhaps the two most obvious places to put the tank were as a standalone unit (since it offered quite new capabilities)
or in cavalry regiments (since it was highly mobile and the horse was becoming obsolete).
There were traps along either route: the established regiments would resist a standalone structure for tanks, which
would compete for resources while the postwar army was shrinking. A new tank regiment would lack both allies and
the heft of historical tradition.
After various twists and turns, it was the cavalry that ended up as the organisational home of the tank. And cavalry
officers certainly understand a highly mobile strike capability. But they were never really organised around the concept
of “mobility”. They were organised around horses. The cavalry officer loved his horse and rode it with skill. His
regiment was devoted to feeding and caring for the horses. Would he not resist the usurper tank with every fibre of his
being?
Xerox Parc developed or assembled most of the features of a user-friendly personal computer, but Xerox itself did
not have the organisational architecture to manufacture and market it. Xerox Parc did develop the laser printer, a
product that matched the company’s expertise nicely. As Gladwell pointed out, this easily paid for the entire Parc
project. The laser printer was like artillery or the machine gun for Xerox: it was an exciting new technology, but it was
not a challenge to the organisation’s architecture. The personal computer was like the tank.
The same is true for Sony and the Memory Stick Walkman. As Sony expanded, it produced radios and televisions,
video recorders and camcorders, computers, game consoles and even acquired a film and music empire. But to keep
this sprawl manageable, Sony’s leaders divided it into silos. As Gillian Tett explains in The Silo Effect, the silo that
produced the PlayStation had almost nothing to do with the silo that produced portable CD players.
The Memory Stick Walkman was like the tank: it didn’t fit neatly into any category. To be a success, the silos that had
been designed to work separately would have to work together. That required an architectural change that Sony tried
but failed to achieve.
The tank was like the personal computer: it required a different organisational architecture
And for IBM, the shift from a mechanical tabulator to a mainframe digital computer was like the shift from rifles to the
machine gun: an awesome step up in firepower, but a modest adjustment to organisational capacity. The tank was
like the personal computer: it may have been a logical step forward given the technology available, but it required a
different organisational architecture — one that bypassed and threatened the existing power centres of Big Blue. That
was the problem.
The politics of organisational change are never easy. In the case of the tank, they were brutal. The British public never
wanted to fight another war in mainland Europe, and the tank represented an admission that they might have to. The
armed forces were starved of cash in the 1920s and 1930s. In 1932, the British army ordered just nine tanks —
delicate four-tonners. The total weight of this entire force was less than a single German Tiger tank. But at a time of
declining budgets, who could justify buying more?
It did not help that the tank enthusiasts were often politically naive. Since an architectural innovation requires an
organisational overhaul, it is a task requiring skilful diplomacy.
Fuller was no diplomat. His essays and books were dotted with spiky critiques of senior military officers. After a while,
even the junior officers who admired his wit began to tire of his “needlessly offensive” lecturing.
Despite alienating the army top brass, Fuller was handed a unique opportunity to advance the cause of tanks in the
British army: he was offered the command of a new experimental mechanised force in December 1926.
There was just one problem: he would have to step away from his single-minded focus on the tank, also taking
command of an infantry brigade and a garrison. In short, Fuller would have to get into the organisational headaches
that surround any architectural innovation. He baulked, and wrote to the head of the army demanding that these other
duties be carried out by someone else, eventually threatening to resign. The position was awarded to another officer,
and Fuller’s career never recovered. His petulance cost him — and the British army — dearly. Architectural
innovations can seem too much like hard work, even for those most committed to seeing them succeed.
Within academia, Rebecca Henderson’s ideas about architectural innovation are widely cited, and she is one of
only two academics at Harvard Business School to hold the rank of university professor. The casual observer of
business theories, however, is far more likely to have heard of Clayton Christensen, one of the most famous
management gurus on the planet.
That may be because Christensen has a single clear theory of how disruption happens — and a solution, too: disrupt
yourself before you are disrupted by someone else. That elegance is something we tend to find appealing.
The reality of disruption is less elegant — and harder to solve. Kodak’s position may well have been impossible, no
matter what managers had done. If so, the most profitable response would have been to vanish gracefully.
“There are multiple points of failure,” says Henderson. “There’s the problem of reorganisation. There’s the question of
whether the new idea will be profitable. There are cognitive filters. There is more than one kind of denial. To navigate
successfully through, an incumbent organisation has to overcome every one of these obstacles.”
In an email, she added that the innovators — like Fuller — are often difficult people. “The people who bug large
organisations to do new things are socially awkward, slightly fanatical and politically often hopelessly naive.” Another
point of failure.
The message of Henderson’s work with Kim Clark and others is that when companies or institutions are faced with an
organisationally disruptive innovation, there is no simple solution. There may be no solution at all. “I’m sorry it’s not
more management guru-ish,” she tells me, laughing. “But anybody who’s really any good at this will tell you that this is
hard.”
Almost a decade after resigning from a senior position in the British army, Andrew Mackay agrees: “I’d love to think
that there could be a solution, but I don’t think there is.”
If I had to bet on the most significant disruption occurring today, I would point to the energy industry. Chris Goodall is
a longtime observer of the renewable energy scene and author of The Switch, a book about breakthroughs in solar
panel technology. Goodall points out that solar photovoltaics have enjoyed a dramatic fall in costs, one that shows no
sign of abating.
Solar PV electricity is now cheaper than electricity generated by gas or coal in the sunny climes where most of the
planet’s population live. A few more years and that advantage will seem overwhelming, which is great news for the
planet and terrible news for incumbents.
Consider General Electric, which this year disappeared from the Dow Jones Industrial Average. In little more than a
year, the old industrial titan’s share price had halved. One of the key culprits for its woes was a precipitous collapse in
the demand for large gas turbines, that, in turn, was the result of a fall in the cost of solar power cells that had been
relentless, predictable and ignored.
This possibility has been clear to the fossil fuel industry for a while. I know: I used to work in long-range scenario
planning for Shell International. Back in 2001, my Shell colleagues and I were discussing thin solar films that could be
printed cheaply and applied to windows or hung as wallpaper.
We could see the threat of exponentially cheaper solar power — but recall what Joshua Gans said about Kodak and
Blockbuster: “They knew what the future looked like. They didn’t know later than everybody else, they knew ahead of
everybody else.”
They knew. But they could not act. Because what is an oil company to do in a world of abundant, cheap solar energy?
Offshore wind farms play to some oil-company strengths; they know a lot about large metal structures in the North
Sea. But solar energy is an architectural innovation. The pieces just don’t fit together like an oil rig or a refinery. As a
mass-market, manufactured product it is closer to the skill set of Ikea than Exxon.
The implication of Christensen’s theory is that oil companies should have set up solar subsidiaries decades ago. Many
of them did, without much success. The implication of Henderson’s theory is that the oil companies are in big trouble.
Chris Goodall thinks the oil companies should rescue what they can — for example, by developing synthetic
hydrocarbons derived from water, atmospheric carbon dioxide and solar energy. Such products would play to oil-
company strengths. But for most of their business lines, Goodall says, “The best strategy for the oil companies is
almost certainly gradual self-liquidation.”
Or as BP’s chief executive Bob Dudley candidly admitted to the Washington Post recently, “If someone said, ‘Here’s
$10bn to invest in renewables,’ we wouldn’t know how to do it.”
Despite all the obstacles, the British army continued to develop both tanks and tank tactics throughout the 1920s
and 1930s. Yet the internal politics proved toxic. The Germans, meanwhile, watched and learnt. If the British were
hamstrung by their inability to reorganise what was, after all, a victorious army in the first world war, the Germans had
the opposite problem: they had barely any army, and no status quo to defend. There was no organisational
architecture to get in the way.
When Adolf Hitler came to power in 1933 and began to expand the German army and invest in tanks, he encountered
a German military that had been watching, thinking and experimenting for 14 years. On his 50th birthday in 1939,
Hitler celebrated with a parade of Germany’s newly reconstructed army through Berlin. “For three hours,” wrote one
witness, “a completely mechanised and motorised army roared past the Führer.” This witness was a guest of honour
at the celebrations. His name: J F C Fuller.
After quitting the British army in frustration, he had thrown his lot in with the British fascists of Oswald Mosley. He
wrote vitriolic attacks on Jews. Some observers wondered whether this was simply an attempt to win favour with the
world’s tank superpower, Nazi Germany. One of Fuller’s biographers, Mark Urban, doubts this: “The facility with which
Fuller made anti-Jewish jibes in letters and books suggests pleasure rather than duty.”
Nobody doubts, however, that Fuller was obsessed by German tanks. After all, there was one army that had really
understood and embraced his ideas: that of Adolf Hitler.
After the parade, Major General Fuller met Hitler himself in a receiving line at the Chancellery. The Führer grasped
Fuller’s hand and asked, “I hope you were pleased with your children?”
“Your excellency,” Fuller replied, “They have grown up so quickly that I no longer recognise them.”
Illustrations by Janne Iivonen Tim Harford is an FT columnist. His latest book, ‘Fifty Things that Made the Modern
Economy’, is available in paperback