Does a Company Like Apple Need a Genius Like Steve Jobs?

The old tech adage is that “open beats closed.” In other words, open technological systems, or those that allow interoperability, always beat their closed competitors. This is an article of faith for certain engineers. It’s also the lesson from Windows’ defeat of the Apple Macintosh in the nineteen-nineties, Google’s triumph in the early aughts, and, more broadly, the success of the Internet over its closed rivals (remember AOL?). But is it still true?

The adage has been seriously questioned over the last few years, primarily because of one firm. Apple, ignoring the ideals of engineers and the preaching of tech pundits, steadfastly stuck to a semi-closed strategy—or an “integrated” one, as it likes to say—and defied the rule. Compared to its rivals, Apple is far more structurally integrated. It owns its hardware, software, and retail operations. It also curates and blocks competitors a lot more. Oh, yes, and by doing this, the company became the most valuable one on earth. Last quarter, Apple made more profit than Amazon has made over its entire lifespan.

But now, over the last six months, in ways little and large, Apple has begun to stumble. Accuse me of overreading, but I propose a revision of the old adage: closed can beat open, but you have to be genius. Under normal conditions, in an unpredictable industry, and given regular levels of human error, open still beats closed. Stated a different way, a firm gets to be closed in exact proportion to its vision and design talent.

To explain, I need to first be careful about what I mean by “open” and “closed,” words that are widely used in the tech industry, but with various meanings. The truth is that no company is completely open or completely closed; they exist on a spectrum, somewhat like the one that Alfred Kinsey used to describe human sexuality. Here, I mean it as the combination of three things.

First, “open” and “closed” can refer to how permissive a tech firm is, with respect to who can partner with or interconnect with its products to reach consumers. We say an operating system like Linux is “open” because anyone can design a device that runs Linux. In contrast, Apple is very selective: it would never license iOS to run on a Samsung phone, or sell the Kindle in an Apple store. Second, openness can describe how impartially a tech company treats other firms in comparison to how it treats itself. Firefox, the browser, treats most Web sites about the same. Apple, in contrast, always treats itself better. (Try removing iTunes from your iPhone.) Third, and finally, it describes how open, or transparent, the company is about how its products work, and how to work with them. Open-source products, or those that rely on open standards, make their source code available widely. Meanwhile, a firm like Google might be open in many respects, but it guards things like its search-engine code very carefully. In tech, the standard metaphor to describe this last difference is that of a cathedral versus a bazaar.

No private firm is entirely open, though some non-profits, like Mozilla, come close. Similarly, no firm today can afford to be entirely closed. A platform owner gains from having good apps available (consider, um, the iPhone without Google Maps), and so too much blocking will destroy what makes the product valuable. Even Apple needs to be open enough not to annoy consumers too much. You can’t run Adobe’s Flash on an iPad, but you can plug nearly any kind of earphones into it.

The idea that “open beats closed” is a new one. For most of the twentieth century, integration was widely believed to be the superior form of business organization. Closed or integrated designs come with advantages long recognized and even trumpeted by economists. Coördination is a key advantage: in theory, with one firm coördinating every aspect of a given product’s features, the result can work better than its uncoördinated rival. The economist Joseph Farrell called this the “internalization of complementary efficiencies.” If that doesn’t stick in your ear, consider it the Disney World effect. Disney exercises iron (or near-iron) control, and the amusement park operates more smoothly and impressively than, say, a typical country fair.

Andrew Carnegie relied on logic similar to Apple’s when he integrated ore mining with steel production in U.S. Steel. The old Hollywood studios of the nineteen-thirties and forties integrated acting, writing, production, and theatres into one firm, and successfully chased everyone else out of the industry. I.B.M. was closed, and the old A. T. & T. monopoly was the ultimate closed system: you weren’t allowed to own your own telephone, let alone use one made by someone else.

The conventional wisdom began to change in the nineteen-seventies. In technology markets, from the eighties through the mid-aughts, open systems repeatedly defeated their closed competitors. Microsoft Windows defeated its rivals by being more open: unlike Apple’s operating system, which was technically superior, Windows ran on any hardware, and ran nearly any software. At the same time, Microsoft also defeated a vertically integrated I.B.M. (Remember Warp O.S.?) Google was boldly open in its original design, and sailed past the selective pay-for-placement design of Yahoo. Most of the winner firms in the eighties to the aughts, like Microsoft, Dell, Palm, Google, and Netscape, were open. And the Internet itself, a government-funded project, was both incredibly open and incredibly successful. A movement was born, and with it the rule that “open beats closed.”

The triumph of open systems revealed a major defect in closed designs. As a matter of economic theory, in a state of perfect information, a central designer should be able to produce a better product. But that only follows if the future is predictable, and if you ignore the tendency of humans to make boneheaded mistakes. In a closed system, with one decision-maker, errors are very costly. Stupid decisions, or compromising the product for short-term profit, will make it not just a bit worse but much worse than the open competitor. For example, AOL’s “walled garden” of the nineties tried to guess what users wanted, but AOL made lots of mistakes, and it was ultimately no match for the open Web.

An open product, in contrast, is better buffered against human error, because no one entity makes a decision that can destroy the product. The economists Tim Bresnahan and Shane Greenstein, writing in the nineteen-nineties, described this as “divided technical leadership,” and they meant that as a good thing. The product is the collective result of many, and sometimes thousands, of decision-makers. An open product can also take advantage of collective, voluntary contributions of the masses, a point emphasized by Yochai Benkler. Consequently, an individual Wikipedia entry might be lousy and contain errors, but the entire corpus will remain impressive. In the mid-nineteen nineties, Windows wasn’t as intuitive as Macintosh, but all the accessories and all the applications collectively made it a superior product.

Which brings us to the aughts, and Apple’s great run. For about twelve years, Apple successfully beat the rule. But that’s because it had the best of all possible systems; namely, a dictator with absolute control who was also a genius. Steve Jobs was the corporate version of Plato’s ideal: the philosopher-king more effective than any democracy. The firm was dependent on one centralized mind, but he made very few mistakes. In a world without errors, closed beats open. Consequently, for a while, Apple bested its rivals.

So what’s a technology firm to do? Each faces the open/closed question, and here is how to answer it. First and foremost, there will always be a complex tradeoff between closed and open designs, so there is no use being too religious in either direction. It is easy to underestimate open designs (no one thought Wikipedia would work), but even open systems need some points of control. In the end, the better your vision and design skills, the more closed you can try to be. If you think your product designers can duplicate the nearly error-free performance of Jobs over the past twelve years, go for it. But if mere mortals run your firm, or if you’re facing an extremely unpredictable future, the economics of error suggest an open system is safer. Maybe rely on this test: wake up, look in the mirror, and ask yourself, Am I Steve Jobs?

[#image: /photos/59095293ebe912338a372d62]Infographic: Are open companies more successful?

Illustration by Jon Han.