The Economist gets e-commerce all wrong.
�...The web is the most selfish environment in the world,� says Daniel Rosensweig, chief operating officer of Yahoo! �People want to use the internet whenever they want, how they want and for whatever they want.�
Look, Yahoo is my favorite victim these days - but any decent strategist has got to have figured out there's a lot more to the Net and business than switching costs. In fact, the Net offers unparalleled opportunity to
raise switching costs exponentially - take a look at Meetup, eBay, Amazon, Blogger, etc.
I've argued before that 'ecommerce' was always kind of a silly notion that was doomed to fail - because the economics of the Net support an entirely kind of business model. The Net is not just a 'channel' through which to sell pet food or DVD players - in fact, the most powerful business model for the Net is the switchboard model, which exploits the fact that the Net is really a massive infrastructure for rare direct network externalities (which raise switching costs), in a number of different ways.
The term 'network effects' has been bandied around enough to gloss this over - most 'ecommerce' business models ignore the very special economics of the Net entirely and simply recreate the same old model with a different facade - ie iTunes. These 'ecommerce' models are doomed to die because the Net has massive information problems compared to the real world - 'ecommerce' is an infrastructure for massive adverse selection, because of huge information asymmetries between buyers and sellers. Do you really know whether you're buying a lemon?
Almost certainly never. This is why the 'transaction costs' argument put forth by the Economist is not valid (and is very 1998) - anyone doing business online knows now that 'transaction costs' - more accurately, negotiation or enforcements costs - can be much, much higher than in the real world. It's only search costs that are, by definition, lower on the Net - search economies are great, but they're not a strategic advantage by themselves, because their gain can be offset by the potential loss of adverse selection. Succesful Net business model use these search economies to deliver guarantees against information problems via network externalities - we see this in playlists, in Amazon's review system, in eBay's reputation system, etc, etc.
The point is, at the end of the day, obvious. Before the MBA's and analysts invaded, the Net was good at, well, connecting people - not pushing products (or spam, or fly-through interstitial mind-control ads). The most successful business models for the Net don't worry about 'ecommerce' - they know it's a failure; they focus instead on creating properties that exploit the Net itself - like switchboards and markets.