What lies beneath: behind the scenes of a business transaction

In a country in which we have more cellphones than inhabitants, it's safe to say that most South Africans will have filled out an application for a mobile contract at least once during their lifetime. It is also safe to say, judging from the number of complaints lodged on customer experience sites or in the press, that a fair amount of us have experienced some form of delay or error during the process. This brings me to my point: How many of us have thought about what lies beneath a business transaction, of which a cellphone contract is but one example? What path does it follow through the system? What traffic will it encounter? And how secure is it?

Business processes are becoming more and more complex, involve an increasing number of different technologies, platforms and sophisticated applications, subject to an increasing number of requirements - such as security, governance and most importantly, visibility.

Companies spend millions on application performance management (APM). APM can drill down directly into each application and reveal what's going on behind the scenes, but it provides little insight into the end-to-end transactional behaviour.

Unable to locate the root of the problem

We recently dealt with a cellphone service provider that had, at any given moment, thousands of stuck orders. Their entire application process, from the time a customer registered an account to when it became live, involved a dozen different applications, each comprising several transactions running on multiple servers. Although they could pore over the log files and investigate each application individually, they were unable to locate the root of the problem and determine where the transaction had become stuck - at the expense of valuable customers.

This problem has become endemic in South Africa, where true adoption of a Service-Orientated Architecture (SOA) from a tech perspective is embryonic at best. Although companies are eager to adopt sophisticated applications, they have yet to master the art of seamlessly bringing these applications together. As a result, when something goes awry within the dependent services in their network, they are faced with an unexpected performance crunch.

By making use of a console that communicates with monitors and provides an intuitive, graphical representation of how transactions are progressing across the system (regardless of the technology), companies are able to take preventative measures when necessary, capitalising on revenue opportunities and responding to impending risks before the customer or sale is affected.

Most companies are aghast when they come "face-to-face" with the number of rogue services within their supposedly well-behaved operating environment. By merely making use of available SOA technology, their enterprise risks are reduced and customer service delivery is improved.

Just as the network world went from simple single vendor to complex unmanaged single vendor, to complex unmanaged multivendor to complex managed multivendor environments, the enterprise software world needs to realise that if they want best-of-breed, they must have a system that can manage it, from start to finish.