History teaches us that a challenge to a superior power can succeed only by changing the rules under which any confrontation takes place. Head-on attacks on strong, entrenched positions are easily repelled. That's why RCA, Honeywell, Xerox, General Electric, Burroughs, Univac, Amdahl and many others failed when they tried to compete with IBM. That's why the computing landscape over the past 15 years has become littered with so many gravestones of firms overcome by Microsoft. Now a new Microsoft challenger is emerging on the horizon: Google. What makes Google different is that it follows innovative rules as it carves out a share of the IT business. Google's emerging strategy may give us a clue as to who may be bidding for IT leadership in our uncertain industry by the end of this decade.
Microsoft's power is based on selling customers software that becomes a sequence of increasingly sticky entanglements. Once you've installed Microsoft operating software on a desktop, laptop or cell phone, the steadily increasing inclusiveness of features will raise your costs for choosing any alternative. Google, on the other hand, relies on a generic browser to gain access to a rapidly growing menu of services.
Microsoft's business model is based on mass distribution of software, where economies of scale generate high profit margins. Revenue growth depends on extending software functionality, and the customer pays for each upgrade. Google's business model is based on offering added services based on economies of scale in operations and on innovation. Revenue growth results from extending the economic utility of services that the customer can add at low cost.
Microsoft's superior profits are based on passing on just about all the costs - including those of support labour and capital - to the customer.
When you adopt the Microsoft environment, you can't ever budget for the total costs of ownership in the long run. Google is a network services provider. Google owns and operates perhaps the world's largest computer complex, consisting of a rapidly growing population of 100,000 servers that manage petabytes of data with unprecedented reliability and response time. Google pays for all of its labour and capital out of its pocket. You pay only for what you get.
Microsoft can expand its business by leaving the job of tending to most of the burdensome tasks of integrity, interoperability, reliability, security and maintainability to the customer. These tasks now absorb perhaps as much as 75 per cent of the total cost of user ownership. This overhead baggage is slowing down the implementation of new applications. Google removes the customer overhead altogether by delivering defined services, with a predictable performance.
Microsoft leaves the management of knowledge capital, which is represented by the data and the files spinning on disks, largely in the hands of users. Enterprise knowledge management is now almost entirely dominated by IBM and Oracle (and, arguably, by SAP). Microsoft has recognised that this is the source of its greatest competitive vulnerability. It has now targeted the extension of its dominance from desktops (where it has a near monopoly) to dominance over servers (where it is behind). This would be achieved by making server and desktop software functionally inseparable.
Google's view is that what matters is not data but the content that clusters of information arrayed with intelligence may convey. Google is committed to extracting meaning (not data) by means of non-proprietary standards from files that may be scattered but are nevertheless accessible. The transfer of data from paper to computers has characterised the first 50 years of computing. The next era will deal with the transfer of intelligence from computers to people. The proliferation of data and the inability to extract information from it is the problem we must overcome. Consider the inefficiencies in the prevailing practices. My first computer used a 160KB floppy disk to store everything. Now I need two 160GB disks just to keep up. That scales up by a factor of a million, and I see that headed for a billion. Yet everything I put on my computers prior to 1997 has been lost as I migrated hardware and software configurations. At the enterprise level, the losses of institutional memory are even more severe. This sort of memory failure has become a characteristic of all computer-based organisations, as millions of disk drives fill up with redundant, and unretrievable, information stashed away in proprietary formats.
As we evolve toward datacentric architectures, the management of systems for collaboration and for knowledge sharing will be central to profitable investments in IT. Judging from Google's research and its current marketing probes, it appears to be focused on competing for this opportunity. How Microsoft proposes to address this challenge is yet to be seen.