Google’s vertical integration

By pushing down transaction costs, many business seers have argued, the internet will drive the final nail into the coffin of vertical integration as a business strategy. Companies will become highly focused modules, doing the narrow tasks they’re best suited for and connecting with other specialized modules to form flexible, superefficient supply chains.

So what the heck are we to make of Google? Here we have the ultimate internet firm, and it’s pursuing a vertical integration strategy so expansive that it would make Henry Ford dance with delight. It builds its own server computers as well as designing some of their components, writes its own operating systems, builds and operates its own data centers, owns much of the fiber-optic cabling through which its data travels, codes and delivers a suite of applications ranging from search engines to business productivity programs, sells and delivers ads over various media, operates its own checkout and payments service, writes much of its internal operating software, generates some of the electricity for its buildings, and operates the bus system that brings its employees to and from work. And now we hear of its apparent plans not only to operate a wireless network and make a smartphone but to lay a communication cable across the floor of the Pacific Ocean.

When it comes to the demise of vertical integration and the triumph of specialization, is Google the exception that proves the rule or the exception that proves the rule wrong?

7 thoughts on “Google’s vertical integration

  1. dubdub

    The best companies routinely violate the “don’t reinvent the wheel” rule in the (excellent) list you published awhile back. If they succeed, they get rich (and a halo!); if they fail, they become a cautionary tail.

    Since success is hard, there are always more successes than failures, which leads to the rule in the first place. But the rule isn’t causal.

  2. Ciaran

    The google buses that run up and down my street at all hours of the day and night are run by a charter coach company. But mayber they’re just a front for google…

  3. patfitzsimmons

    I think a couple of things are going on.

    1) Google made so much money from search, that the laws of economics do not apply to their other activities. They can lose money on dozens of other side projects and not even notice it. The founders like building stuff more than they like making money. Google could have made PayPal bid for the right to be the payment system of choice for Google ads. This likely would have netted more profit than building their own. But it would not be as fun.

    2) Engineers are motivated by a lot more than money. Google creates the dream environment for engineers, and the result is that people just build things. Part of the secret to their success is Google gets more for the engineering dollar than any other company.

    3) It is now becoming possible to build a search engine on top of Amazon’s EC2 using open source versions of Google’s tools ( Hadoop, HBase, Nutch). It will be very interesting to see if a small startup using these tools will be to build a search engine that takes a bite out of Google’s market share. PowerSet and Yahoo! are trying, and there will be more.

    4) While transaction costs may be vanishing, lock in via network effects is still very profitable. Ad networks right now have huge networks effects and economies of scale.

  4. Phil

    A Pedant Writes: the exception ‘proves’ the rule in the archaic sense of testing it – in other words, it’s by looking at the exception that we can tell whether the rule is a rule or not. The example of Google certainly suggests there’s something a bit wobbly about this particular rule.

  5. niraj j

    I think you probably need to look at Vertical integration in different light.

    I call it Aspirational Core Competency based Vertical integration – which is something like – You treat area’s you want to master as competencies and build that over time.

    Check out

    Google could have got into making its own hardware. But did not. Probably because it does have aspirations to be in the area which could be as a result of

    1. There is already a lot of innovation in Hardware to add further value.

    2. Size of market

    It got into Energy – which it thinks does not have the right innovation happening.

Comments are closed.