"I'm not dead!" Those classic words from Monty Python and the Holy Grail kept going through my head last month as the news of Red Hat's 43% revenue increase pushed it into profit. The primary factor in the increase was a 79% jump in subscription revenue for Red Hat Enterprise Linux. Together with this clear indication that the subscription software model is meeting with significant acceptance is the shift to service providers for spam and virus filtering for e-mail. These general trends demonstrate that the declaration of the death of online service delivery as a viable business may have been premature.During the so-called Internet bubble of the late 1990s and early 21st century, service providers were the belles of the ball. Starting with ISPs and then ASPs there eventually was an entire business space often referred to as xSP. Most of the space was vilified following the "bubble burst," and many companies changed their messages to remove the negative connotations of being an xSP. Even this column moved from focusing primarily on ASPs to a focus on the broader marketplace of outsourcing of all kinds.However, as Amazon and eBay have shown us with respect to e-commerce, ASPs still represent not only a viable business, but the most logical approach to delivering a number of technology services. Complex and computationally-intensive services are among those best served by ASPs, and the application of new technology to some of these areas lends itself to the close management, rapid updates, and security considerations possible in a hosted environment.This was brought home to me last month when I met with a new company focused on business intelligence. AnswerOn (https:\/\/www.answeron.com\/) is an early-phase company with a compelling story of business intelligence applied to the problem of customer churn for subscriber-based businesses such as ISPs and wireless telephone service providers. Using an application service model, AnswerOn collects data from CRM applications, performs complex computationally intensive analytics, and returns customer churn data to the CRM to allow their customers to focus retention activities on those customers at highest risk.While an application like this could certainly be offered as a shrink-wrapped software product, offering it as an application service presents a number of advantages that are applicable across a broad range of potential offerings. Among the advantages are the ability to update the software with the latest algorithms and analytics transparently to customers, the ability to upgrade hardware and technology architecture to improve performance, and the control over access to systems and intellectual property brought about by the walled garden of the service provider's infrastructure.These advantages are not negated by the financial markets' skittish response to ASPs. Just as Amazon and eBay have shown, solid business plans and technology advantages can create long-term profitable businesses. There are solid technical and business reasons to provide applications as a service. Furthermore, the introduction and rapid adoption of Web services as an application model creates a market for components as a service. In this model, applications are constructed by linking multiple component services. Each of those components may be offered as a service by one or more service providers, each competing for business by providing the fastest, most reliable version of the component.The ultimate marketplace for providers of software as a service is yet to be seen. However, the benefits clearly show that there remains a viable business for organizations that focus on customer requirements and deliver an excellent product. The marketplace is not dead. "I think I'll go for a walk..."