It's a while since I've written a serious piece in this blog about business or technology strategy, so I thought I would make amends below.  After this dry piece, the few remaining readers of this blog will probably wish that I go back to soccer and sauna – at least subjects that I know something about..

First, the punchline.  I think one of the most difficult decisions managers need to make in any business that relies upon a significant amount of technology (pretty much every business these days) is when to do it yourself and when to partner or rely on third parties.  Dangers abound on either side of these narrow straits.  If you choose to do everything yourself you may very well please your in-house developers, but you'll soon find that you suffer from a terrible case of "not invented here" syndrome; you will in all likelihood lack sufficient scale; and  you will find that your costs become uncompetitive.  However, if you steer to the opposite shore and do nothing yourself, you will have no differentiation, no proprietary advantage, and others will be able to readily duplicate your efforts. Scylla and Charybdis indeed.

To me, the course to set to avoid these perils lies in truly understanding the intersection between what your organization can do better than anyone else and how these technologies can advance your business objectives.  In the old days you could consider this buy vs.build strategy once every several years, or at worst, every budget year.  Given the exponential development of technology, I believe that managers now need to think through these issues on a rolling quarterly basis.

To start with, no one should approach technology as a single undifferentiated mass any more than a manager should assume that all his customers and markets are the same.  Not even the very best technology companies are good at every aspect of every technology.  The key is deciding where to pick one's spots and how to understand and measure what these specific technologies can do for your business.  Too many people fall in love with the leading technology of their era or become infatuated with the plumbing or gadgetry,  rather than seeing on a more abstract basis what technology can do for a business.

Let me give you an example from the history of Thomson Reuters.  At various points in our development, pigeons and rowboats represented state-of-the-art technologies which conferred significant competitive advantage.  The pigeon story is perhaps the better known.  During the mid-19th century Paul Julius Reuter used carrier pigeons to bridge the gap that then existed in the telegraph system between Aachen and Brussels, and thereby beat the next best technology (couriers on horseback) in delivering the news as fast as possible.  I sometimes joke within the company that old Baron Reuters also invented redundant packet transmission (a feature of some modern routing networks) because he sent two pigeons with identical news scrolls attached to their legs. 

Some years later, the Baron again introduced a great technological advance in low latency transmission when he engaged an intrepid Irish rower to intercept mail boats coming from North America and row back to shore to telegraph the news to London, thereby beating other news services that needed to wait until the mail boats docked in England.  It was in this manner that Reuters scooped the London market by hours with the news of the death of Abraham Lincoln in 1865.  

The serious point I'm trying to illustrate with these quaint stories is that we are often trapped in and by our current generation's biases as to what constitutes "technology.”  Thus, in the mid-19th century pigeons and rowboats represented signficant technological innovations compared to the then state-of-the-art.  The key is to focus on the business benefit and value to the customer ("deliver the news faster than the other guy") rather than obsessing over the current methodology. 

I believe we can see a modern equivalent of pigeons vs. horses being played out today in the media business.  Serious grown-ups are having a hard time making the distinction between quality journalism and the printing of news on paper created from wood pulp.  Again, the key is to focus on the customer benefit (provide me with information concerning what's going on in the world that I either need to do my job or enjoy as a form of entertainment) rather than the specific publication technology of the era. 

It turns out that publishing news on processed wood pulp paper has actually been a very good and long-serving medium.  It has been relatively cost efficient; it provides high contrast, highly legible output; and it is easily portable.  People forget, however, that before wood pulp, stretched animal skins, papyrus and even cave walls once constituted "paper".  It should not be too far a stretch of the ordinary imagination to foresee a day when printing with electronic ink on a highly flexible and reusable, and probably plastic-based sheaf of "paper,” will be seen as the obvious state-of-the-art technology.  In fact, I'd go as far as to say it will be seen in retrospect that the felling of trees in a Canadian or Finnish forest, their transportation using fossil fuels to a paper pulp mill, their processing into "paper" via a noxious chemical process, their re-shipment in large rolls to newspaper plants around the world, their printing with only a single day’s worth of news, and their third shipment to your front door or newsstand as a modern “newspaper” will seem like the most counter-intuitive, environmentally unfriendly and uneconomic production chain of all time. 

So the moral of this story is that companies and their managers should stay as agile as possible with respect to the current generation of technology; they should focus on the customer and business benefit and not the plumbing; and they should be prepared to re-invent the underpinnings of their enterprise.  We have too many examples of the mainframe makers who failed to transition to mini-computers; the mini-computer makers who failed to become PC makers, and the PC makers who may yet fail to transition into net appliance makers interacting with the cloud.  Even the mighty and much hyped cloud will one day pass as well.