Google and Apple are two of the most profitable companies on the globe today. They seem to share little in common except that achievement. They took very different paths to the stratosphere.
Google, after all, is less than a decade and a half old, a child of the web with a successful approach to advertising, built around a search engine and many services to enhance the user’s experience. Apple is more than twice as old. Its original product, personal computers, makes up a fraction of its sales today, while its future profitability lies with a mix of software in iTunes and new hardware introduced in the last decade—namely, phones, tablets, and portable music devices.
What economic insight emerges from setting these two firms next to one another? A brief discussion of both of their businesses will reveal something trite and something deep. The trite part is this: Some settings produce lots of market value, and some firms capture large parts of that value, but those rarely happen together. The deep part forms the key insight today: these examples are fabulously profitable because they are unique.
Stretch the frontier
Stretching the technical frontier usually plays a role in generating big profitability. That alone does not predict much, however. Stretching the frontier is merely a step in the right direction, necessary but not sufficient. Google’s experience will illustrate these points.
First, stretching the technical frontier for commercial gain is never an exact science. Stretching the frontier for commercial gain usually requires the right type of scientific discovery, and it can elude many a would-be inventor.
What is the right type of discovery? It combines fundamental insights on the one hand and pragmatic invention on the other. Fundamental insight is the type of knowledge generated by particle accelerators. It is not necessarily useful. Pragmatic invention is the type of insight generated by trial and error, especially the experimental methods made famous by Edison as he tried thousands of different filaments inside the nascent light bulb. That might generate useful knowledge, but not necessarily general lessons with wide application.
Google’s earliest search-engine design is an example of frontier stretching that combines both types of insight. Its first invention—implementing page rank in a search engine—was useful and novel. It earned Larry Page and Sergey Brin attention and lead their young firm to get a little piece of revenue (that is, providing search for Yahoo). After that, Google invested in making the search engine faster and even more relevant, stretching the technical frontier in its server farms and using frontier hardware designs. These investments used general principles from frontier computer science, applied to practical user issues.
Employees at the firm have not stopped adding additional frontier investments, and these take various forms. They show up in enhancements to existing services. For example, the search engine added multiple mediums, maps, news, and other types of information. Improvements to search also leverage Google’s size, using prior users’ experiences to help the next users get the information they seek.
That frontier technical advance, by itself, was insufficient to make Google wildly profitable. The frontier does help, to be sure, but not for the reason one might think. Rather, the scale and size is the key attribute because it reduces effective competition. Only a few firms—such as Microsoft, IBM, or Yahoo—could ever offer a service to match Google’s, because few firms have the budget. Also importantly, this size reduces the likelihood that any startup could quickly ramp up to a size that offers an effective competitor.
Explaining why Google captures so much value requires understanding the role of an additional innovation, one that is not technical on the surface. It is a novel set of routines and practices for allocating costs and collecting revenues. To wit, it is the keyword auction.
Google’s keyword auction is a quality-weighted second-price position auction. It is definitely not an obvious way to do things, and Google did not generate huge profitability until this auction became a part of its offerings. That is because it matched more relevant ads to user requests, allowing for better targeting, making advertisers and users happier.
Designing an appropriate auction for a new setting was actually quite difficult, and required technical innovation as well. For example, developing a useful and relevant algorithm for quality scores for the entire web involved some challenging computer science. It still does, because it involves finding a way to weight the relevance of a website’s landing page and match that to a keyword search.
Here is the point: Google is so fabulously wealthy because of the combination of useful and unique inventions executed at a scale that no other firm comes close to matching.
That last conclusion can be stated more generally. A path to a large payoff often depends on using technical innovations from a range of sources, combining them in a unique way that appeals to many end-market buyers. These combinations of innovations rarely involve a single epiphany. Rather, large value emerges from mixing together a series of complementary inventions that work together, and doing so in a unique combination.
Unique combinations are hard to come by, partly because complementary inventions often do not get invented all at once. To use a historical example, the CD-ROM, by itself, was not a particularly valuable invention. It was an intermediate input into software development for PCs and games. Big commercial value was not created until the disks arrived and many developers began playing with new applications in a range of prototypes. As it turned out, it took a while for developers to learn how to use the new medium, to trade off various design goals and new technical capabilities to generate products many users would pay for.
In the case of the CD-ROM, by the time many firms had learned the key lessons, the lessons were widespread. Hence, the gains also became widespread, and commonality bid down prices. Only a few applications and games were both unique and in demand.
Now consider a rather different example of a product that combines innovations from many sources—Apple’s iPhone. The hardware combines many elements that many engineers already understood—albeit, implemented at a production scale and reliability level rarely ever attempted or executed. In addition, it is combined with an excellent app store with world-class ease of use, and an established distribution network. It also comes from a firm whose many users are loyal because they believe upgrades will face few delays.
Is that combination rare? Pieces of it are not rare at all. After all, Amazon’s storefront is just as good. Motorola’s hardware can be just as slick. Scores of firms can make a musical player that works as reliably. And thousands of firms have access to worldwide distribution and low-cost manufacturing.
Can any single firm put together the same combination? Some have tried and failed. The combination of the iPod and iTunes was quite profitable, in part because Apple outpaced potential substitutes. The firm did not stand still. They redesigned and upgraded the hardware regularly, introducing new combinations to address new needs.
For the time being, Apple’s tablet and phone appear more vulnerable to competition. For example, the Android network gets rather close for many users, with a similar combination of distribution, hardware, and an app store.
That said, it is hard to imagine Apple ever losing a huge part of its market share. The firm has been careful to design its hardware to preserve the value of prior purchases on iTunes. Those customers have something valuable that they do not want to lose, and that they can use on new devices. That motivates many of them to try Apple products first and repeatedly.
What are the generalities?
Here is the broad economic point. In modern economies, attempts to be the unique firm to stretch the frontier are expensive. High expense pushes many innovative efforts toward trading off affordability with less-extensive incremental leaps. That pushes away from radical breakthroughs at any single firm. Hence, uniqueness does not arise from a single invention, but from combining incremental leaps from many sources, if they arise at all.
These examples also hint at why the fortunes of firms and society do not necessarily coincide. Because uniqueness produces large gains, it falls when knowledge becomes widespread and many firms learn similar techniques. Think of it this way: Widely dispersed access to the fundamental inventions keeps the returns lower.
To be sure, many firms have noticed this pattern, so from time to time some firms attempt to capture the gains from discovery by doing all the technical work, intending to keep the knowledge to themselves. While that strategy needs some luck to succeed, it invariably runs into an economic constraint. Few participants in markets can afford to make big bets.
In sum, big technical advances leading to large creation of value happens infrequently, and successful capturing of value requires uncommon luck and skill. Many rare events together yield a very unlikely outcome, so it should not be a surprise that capturing enormous value from technical innovation is very rare, indeed.
There is a silver lining to firms’ failure to capture large profits from big innovations. Society gets a big payoff when competitive forces reduce the gains to any specific firm because none found a unique way to capture value. That spreads the gains around.
Copyright held by IEEE. To view the original essay click here.