There are certain times when a tipping point is reached – that has happened with cloud computing, or software as a service (SaaS), as users prefer to refer to it in South Africa, writes Richard Firth, CEO and chairman, MIP.

Companies of all sizes are moving to adopt it, and with good reason. IT managers and their boardrooms are tired of paying vast amounts of money up-front for software applications and the supporting infrastructure, with no guarantee that it will work. In some cases, as with the City of Johannesburg, not only does it not work, it severely compromises the city and its ratepayers.
Here are the five key requirements to make SaaS, or cloud computing, work:
* Alignment with business requirements – don’t think about going this route it if the application is even marginally out of line with business strategy. Functionality always trumps all other issues.
* Availability and security – it should go without saying that SaaS should always be available and secure, but it bears repeating. Denial of service and incompetence at the back-end are realities.
* Legacy protection – no company will rip and replace existing applications, so those which have been around for some time need to be future-proofed. This can be done through basic SOA (service-oriented architecture) principles, which allow long-standing applications and their functionality to be made available to new-generation systems.
* Pay as you grow – there should never, ever again be a situation where a company acquires a software application without tying its purchase implementation to business success. As an example, the hundreds of millions spent at City of Johannesburg would not have been incurred had this principle been followed. Capex for software acquisition should be all but outlawed.
* Ease of management – visibility, transparency, data ownership, systems management, database management, fine-tuning, forward growth paths: these are just some of the issues which should be subsumed in a SaaS environment.
In the South African context, SaaS works. Companies who care about reducing their cost of computing while aligning it with business strategy will ultimately gravitate towards this model.