Samsung apartment cluster in Seoul. It looks like those towers in The Matrix...

The word zaibatsu (財閥) in Japanese, or chaebol (재벌) in Korean, translates most closely to ‘business conglomerate’, but there are some important differences than what Americans or Europeans might understand: zaibatsu are very close-knit to their partner companies, often exchanging shares, emphasis is placed on vertical integration wherever possible, money can slosh around internally much more easily than modern Western law would allow.  While the zaibatsu were broken up after WWII by MacArthur, their shells continue on in Japan: Mitsui, Mitsubishi, Sumitomo. (side note: Mitsui and Mitsubishi each own either side of downtown Tokyo).  In Korea, the Chaebols are even tighter, usually family-run affairs, and pretty much stitch up +80% of the Korean economy.  Perhaps you’ve heard of a few: Hyundai, Smansung, LG, Lotte.

We are now starting to see the rise of the Internet Zaibatsu: Amazon, and Google lead the pack, with Liberty Media, IAC, and other smalle companies following close behind.  These organizations share many of the same qualities of a classic zaibatsu (that sets them apart from just being a conglomerate):

• sympathetic data swapping among properties (either product data or personalized customer profiles) is a central goal in order to maximize economies of scope.  This can bring great seamless integration, or it can dunk you into an opt-in email hell from “partner companies”.
• IT infrastructure is shared wherever possible for economies of scale
• Properties are free to do what they want to meet P&L goals, as long as part of those goals serve the larger good by extending functionality or getting new data sets into the graph.

So far so good, right?  Here’s the thing: zaibatsu are simultaneously tremendous pools of capital and research innovation, but they are also desperately smothering to the environment around them.  The mere hint of a zaibatsu contemplating entry into a market is enough to scare the crap out of the mid-level players.  I am not stating that they are monopolists or oligarchs, just that the zaibatsu structure makes these tough companies to beat.

To a certain extent, the confidence (arrogance) that comes with being a zaibatsu can lead a company into strange directions.  At the extreme edges of the vertical integration strategy, zaibatsu companies will get into businesses where they probably don’t belong.  For example, Samsung specializes in heavy equipment and manufacturing, which lead them to electronics and appliances and all sorts of robots and then data control and professional services.  In the process, Samsung has also bought up huge tracts of land in Korea, which in turn got them into the apartment housing and farming business.  Strange bedfellows.

Google, in a search for ever-more efficient data centers, has gotten itself into the hardware business, data center construction, and overall power management.  Some eyebrows went up, however, when Google started to publicly state it was investigating new ideas toward Fusion power and other nuclear designs.  Is that a natural extension to address their electrical grid needs, or is it just hubris to think they’re so smart they can bring us Mr. Fusion?  Most recently, Google has announced boutiques.com, which may actually put them very close to the line of direct competition with the advertisers that pay the bills.  In terms of competition with Facebook or Amazon, boutiques.com makes perfect sense.  In terms of ’sticking to the core business model’ however, it may stray from the path, IMHO.

Amazon started with books.  Amazon then became such a great early ecommerce store that they branched out to other forms of media.  Soon, Amazon sold everything, including some things that probably don’t make sense, or didn’t really catch fire (like groceries).  That was all horizontal expansion.  The vertical expansion has gone both up the chain and down the layers: Amazon has spent a lot of time and effort to monetize and commoditize as much of their infrastructure as possible.  Their servers, their OS, their digital storage, their shipping are all now services for sale (independent of you being an Amazon merchant).  Shipping, warehousing and business metrics are all available from Amazon.  I’m surprised they haven’t pursued nuclear power like their friends in in Mountain View, yet.

In the end, which model is better?  One school of thought says that the Internet market loves specialization.  As an employee of shoes.com, the day we saw our competitors start stocking watches and trinkets and other chotchkies on their homepages was a good day.  Every day that Zappos moves more toward general merchandise is better for a shoe-only website, right?  The other school (the zaibatsu school) says that it’s all about customer wallet-share.  Anything an organization can do to satisfy more of the needs of their customers is a good thing.

Personally, I am not passing good/bad judgement on either model.  I am just trying to re-introduce a Japanese/Korean term for what we’re seeing lately.