The Future of Application Service Provision

During the internet bubble many of us became enthusiastic advocates of the Application Service Provider (ASP) model – a way to gain access to advanced business applications through the internet for a monthly fee. Bad news for the in-house IT guys, but good news for the bottom line and the new economy. In those days we talked of the virtual organisation – a company made up of best in class building blocks provided by geographically dispersed suppliers, seamlessly integrated through digital networks and unified under the only thing – that really unique thing about any company – the brand.

In the virtual organisation, why own, operate and manage your own IT infrastructure? Software vendors were falling over themselves to make their applications available on the web – the ASP model promised them regular recurring revenues while centralising maintenance and version control. There were start-ups and IPOs in abundance, anticipating massive demand from corporates. You could rent access to ERP packages, TMS packages…the dominance of Microsoft Office was even challenged by the possibility of accessing word processing, spreadsheet and presentation applications through a web browser. Sun’s vision of the ‘network as the computer’ seemed close at hand.

When the bubble burst and the plug was pulled on the new economy, lots of babies went out with the bathwater. ASP looked as if it would go the way of the original boo.com and other dot.gones. It wasn’t just a case of bad timing – there were also substantive practical issues that hindered the development of the model. Some of the issues were:

• Integration – taking the treasury environment as an example, a TMS system needs to be integrated with other applications such as ERP and banking software. Increased complexity of integration with other applications was at least a perceived weakness of the ASP model.
• Trust and security – many companies had valid concerns about entrusting vital company data to relatively unknown ASP providers, and the security of data passing over the open internet.
• Standardisation vs. flexibility – companies do have their own ways of doing things and often like to tinker with software, perhaps more than is healthy. By its very nature, the ASP model provided a standardised set of centrally managed services. Many companies want standardisation, but are in practice loathe to give up the flexibility to do things their own way.

Fast forward to the present and we can see that takeup of ASP services has been much more modest that market research indicated in 1999, although there have been clear success stories such as salesforce.com. In a way the bad timing for ASP continues because the story is drowned out by wider outsourcing trends focused on the physical displacement of labour to low cost countries.

ASP is more of a logical or virtual displacement of the business functionality ensconced in software applications. It is perhaps unsurprising that labour outsourcing has taken centre stage given the apparently ‘no brainer’ business benefits of moving work to countries where staff costs are a percentage of the existing operations. Costs of running ones own applications are either lower in absolute terms, or perhaps just less visible to senior management. It can be easier to count heads than accurately weigh up the fully loaded costs of the company’s IT architecture.

Labour outsourcing and ASP are of course both manifestations of the same wider phenomenon. The economy is re-organising itself to concentrate functions in the most efficient centres, with scant regard for traditional company boundaries.

This itself is a continuation of an older trend – from the early days of Ford Motor Company onwards, companies had to specialise functions within themselves to increase efficiency. Now market forces are ripping functions right out of the corporate body and concentrating them into new centres. We are living through a progressive process whereby the means of production flow inexorably to the most efficient locations:

• Capital flows to the most desirable place in terms of risk/return
• Labour migrates to the best location in terms of cost/quality
• ….and as a continuation of this trend, it is somehow inevitable that software applications will be concentrated with specialised providers who will make these available to users across the internet.

Each these developments are only possible in the presence of networks – financial flows are facilitated by the banking infrastructure, labour outsourcing could not take place in the absence of telecommunications networks and ASP depends on the internet. All of the pieces do seem to be in place for a major wave of applications outsourcing – the networks are in place and further development of the ASP model seems like a historical inevitability. We mentioned some of the factors that blocked takeup in the first wave. Here too, there are also a number of positive forces at work:

• Web Services – there are a new range of tools in the marketplace that allow deeper and more seamless integration between applications over the internet. Developments such as .NET may enable ASP providers to overcome previous integration obstacles.
• Service Oriented Architecture – many companies are following an IT strategy that aims to expose the functionality of its applications over the internal network and in some cases beyond. As more companies deploy Service Oriented Architectures, the mindset of IT managers will change – it will not matter where a service is provided from, or whether it originates from inside or outside the organisation.
• Message Standardisation – integration will be further facilitated by efforts to standardise communications between parties, such as TWIST and RosettaNet.

In banking, the ASP model will be facilitated greatly by the introduction of the SWIFTNet network, which provides a secure real time communication medium between financial institutions. In banking at least there will be a kind of ‘PartnerGrid’ that will see institutions sharing capabilities across the real time SWIFTNet infrastructure. Banks are facing a growing realisation that it is uneconomic for every institution to develop and support the same functionality within the boundaries of the organisation.

Banks and corporates alike have a shared interest in the development of the ASP model. Just as the means of production tend to pool in the most desirable places, so the means of solution to common business problems will clarify and concentrate. One of the criticisms that corporates have of banks is that they too often provide proprietary solutions to what are in essence common problems. This approach can create insurmountable barriers to standardisation and efficient processing.

ASP can be seen as part of the solution. Corporates and banks will begin to make more use of centralised applications utilities which are by their very nature standardised services. The common processing functions will concentrate, leaving only specialised, custom built, highly value added functionality in-house. ASP is therefore not only part of a wider concentration of the means of production, but also the wider trend towards common solutions for common problems.

The ASP model has had a difficult infancy and the industry as a whole remains in a fairly precarious state, with a very small minority of companies using these kinds of services. However, the fundamental concept of the virtual organisation remains a compelling strategic vision, and while attention has been focused recently on the incredible wave of labour outsourcing, ASP is a natural and indispensable part of a bigger, longer-term trend that sees the means of production concentrating in the most efficient places. With this in mind, the ASP providers should take heart – their time is close at hand.

Comments are closed.