Does any of this sound familiar or reminiscent of the hype of the dotcom era?
In essence, Web services are a sound idea. They are a means of integrating applications, of making them available to application builders, of streamlining automated processes, and making them available to partners, suppliers and customers. The idea that Web services can provide easy integration to already automated business processes inside an orchestrated business process management system is appealing. But do we really expect to see Acmewidget. com providing Web services to form part of a mission-critical system? I think not.
There's no doubt that Web services will be useful for integrating existing automated processes within an enterprise, or where companies collaborate on a business project. In other words, they will find most use where the commercial requirements for security and service levels exist already. They will also find a role in non-mission critical areas such as Web content or internal applications. So why don't I buy the dream?
Imagine a banking application: a true mission critical system. Is it likely that this bank will happily plug together hundreds of Web services to produce the application? What about security and availability? Would the bank be happy to delegate the responsibility to Acmewidget.com? Or would it look for comprehensive service level agreements and stringent security assurance?
To overcome these issues the bank would have to administer hundreds of service contracts with Web service providers, including not just those with the primary service provider but the back-up service providers as well.
It's true that in some cases banks co-operate with other companies which could, conceivably, become Web service providers to them. Credit checks have long since been provided by third parties. The point is that these relationships are tightly controlled and monitored and are relatively few in numbers. The banks expect a service level and have vetted their partners to ensure they do. Somehow, "trust me I'm a Web service" just doesn't do the trick.
The adoption of Web services in mission critical applications is like jumping off a cliff and hoping that the ground is soft and not very far away. Ask your board members if they wouldn't mind outsourcing their main business system to a large number of dotcom businesses. They will politely admire your enthusiasm for technology and show you the door.
In organisations that are already in a commercial agreement such as collaborating construction companies, key customer relationships, or different agencies in local or national government, Web services could be the key to simplifying the integration of the required inter-company processes. Web services cannot stand on their own, however, they logically fit as a sub-set to a wider view of business process management.
I mention government agencies as an example of where Web services could be of great benefit. These agencies, in theory at least, are all trusted parties and so the service, security and desired result are more likely to be reliable and predictable.
We are on the up-ramp of marketing hype and will soon sink into a trough of disillusionment. Following this, however, there will be a period of clarity, where Web services will find its appropriate place in the market. That place will be one of easy integration with internal or trusted external systems supporting a business process management platform.
Web services are not an economic revolution, just an easy way of integrating components. Trust will still be the key issue, so non-mission critical applications will benefit from global providers of Web service components, along with carefully controlled applications which link with trusted and reliable sources from partners, suppliers and customers. Will your business fail if you don't use Web services? Definitely not.
Richard Wall is sales and marketing director at BSoftB