Amazon is the most important software company in the world that does not think like a traditional software company, particularly of the "enterprise" or "business" software variety. I am talking about the Amazon Web Services platform. In any rational business, as scale goes up, per-unit costs go down. Competitive market dynamics ensure that prices go down. Sure enough, Amazon recently loweredprices of their services, as they gain scale, and lower their cost of operations. Compare that with what has happened in the business/enterprise software market.
In 1995, when I bought my first personal computer (a whopping 16 MB of RAM!) in San Diego, it cost me almost $3,000 and that was from a local white-box assembler who specialized in cheap. Windows 95 added about $50 to the cost. Microsoft Office came in around $250. Fast forward 13 years, and a perfectly serviceable PC can be had for $300-400. Vista probably will set me back about $80 assuming OEM pricing by the PC vendor. Microsoft Office will cost more than the PC itself, if I am a business user. Even a "generously" discounted student edition is about $120. As hardware costs plummet, software prices have actually gone up, so outfitting a PC with software now costs more than the hardware cost of the PC itself.
Contrary to what software sales and marketing types would have you believe, there is nothing rational about any of this. Throw away arguments like "Software is getting more complex, cost of development only goes up" and so on, because all of those things are true for hardware as well - a semiconductor chip is really hardwired-software. It is all too easy to blame sticky prices on "monopoly taxes" and while I believe there is some truth to the monopoly argument, there is something more fundamental going on. After all, Oracle is in a fairly competitive market, and still no one even expects Oracle to ever drop prices - perish the thought, that never happens. The clearest evidence of such an ingrained expectation is that no Internet company today would even consider Oracle for its database needs. On the flip side, Internet Explorer got itself into a near monopoly after the demise of Netscape, and yet, Firefox has stormed its way up to 20% market-share and climbing.
So the real explanation is elsewhere, and the Firefox example is illuminating: it is the friction of switching costs that keeps prices high. Switching from IE to Firefox is fairly easy - in fact, just last week, at the Web 2.0 Expo trade show floor, we discovered that the demo machines we had rented came only with IE, and the first thing we did was to download and install Firefox. That was almost an instinctive act, done without even much thinking. And a perfect illustration of almost zero switching costs.
That is also why cloud computing is so important to business and enterprise customers, even for those that don't use cloud services. Switching from one web service to another is nowhere as complex as switching your operating system or even your desktop software. In any event, the vendor does most of the work. As one illustration, we frequently migrate Salesforce cusotmers to Zoho CRM, and the process is nearly fully automated, with human intervention mostly confined to perform quality assurance checks to ensure there are no glitches for the customer. As cloud services compete in a relatively low-friction market place and therefore prices fall along with unit costs, it is going to drive prices down even for conventional software.
Amazon is a perfect illustration of it. A lot of the code behind Zoho services was written before Amazon web services came on the scene, yet we have found that it is fairly easy to run our applications on the Amazon infrastructure - it took just a few days to test this out.
Lowered switching costs is going to dramatically alter the dynamics of the software industry. For all too long, vendors have hidden huge inefficiencies in their business model under the comforting cushion of high switching costs for customers. Not anymore. Watch for falling software prices ahead!