The batch, the whole batch, and nothing but the batch
The economics of application development when business systems were first introduced in the 1960s and 70s was driven by the high cost of mainframe computers. Complex data encoding algorithms, satellite date entry and quality control operations, and the whole concept of batch processing, revolved around maximizing the utilization of the computer.
Online, real time, airline
As computer manufacturing costs came down, and processing capabilities improved, the scarcest resource became application developers not computers. Despite the development of specialized business programming languages such as Cobol and Natural, the organization of data in Database Management Systems, and the development of online systems, the ever growing complexity of business data processing continued to demand more trained developers than were available.
PC or not PC
With the advent of mass produced, and massively deployed, personal computers it became impossible to continue to develop business automation with custom programs. Software factories, notably Microsoft, introduced shrink-wrapped software packages that let non-programmers build and deploy complex business systems using spreadsheets, word processing, email, and groupware tools. The economics of application development shifted as business users took on much of the development work of software specialists. The new bottleneck became systems integration, caused by limited network connectivity, middleware, and specialized support staff.
A tangled Web
Partly as a result of the mass deployment of PC's, and partly as a natural outgrowth of advancements in networking technology, the Internet removed the barriers of integration. Now any customer, at any web browser, on any platform could participate in an eBay auction. In addition, trained development staff could be put to work, around the clock, in every corner of the world - it seemed as though the only limitation was capital, and at the height of the dot com boom investors were happy to provide capital.
Marlowe's ghost
As the technology economy emerges from the dot com aftermath, and Web 2.0 is touted as the future direction for business applications - even for the Internet itself - it's worth considering where the application constraints are today. Although web services promise seamless integration of heterogeneous applications, and the semantic web promises to provide data in self describing chunks, the most challenging problem confronting business application providers today is history. Each advance in computing has increased the quantity and importance of business applications - indeed many organizations could not function today without an entire ecology of applications.
The Grim Keeper
Unfortunately, unlike the natural world, there is no built-in death cycle for applications, merely an endlessly prolonged life, where programs execute on the virtual ghosts of previous platforms. Each successive wave of technology has stranded different generations of applications at different points on the beach, and the IT budgets of most large organizations are consumed merely in raking the sand to try to keep the beach clean.
The relative slow down in technology innovation from traditional software vendors seems to be driven by the same constraint. Just as IBM was unable to break free from the legacy of success they enjoyed from the OS/360 operating system that dominated mainframe computing, so the computing world today is suffocating under the burden of Microsoft's Windows PC monopoly.
The end of history
The promise of On Demand computing has the potential to overcome this inertia, and free organizations to innovate and introduce new applications without regard for their historical environment. Freed from the tyranny of existing infrastructure, and with an expectation of reliable, available, customizable solutions delivered when and where needed, organizations can focus their IT resources not on finding lower cost programmers to add yet more lines of code to be maintained, or building one more layer of internal workflow, but instead focus on leveraging the emerging wealth of On Demand services.
Small is beautiful
The organizations best able to take advantage of this economic opportunity may not be large companies with sizeable IT budgets and heavy IT legacies, but smaller organizations with the ability to move quickly. Indeed the initial adoption of On Demand solutions by small and mid-size companies is not so much a reflection of the limited capability of On Demand solutions to handle complex business problems, as it is of the limited capacity of large organizations to embrace IT innovation.