Quite frankly, I'm not a fan of cloudy weather. As a redhead with curls, clouds mean dampness and frizz, and something sleek and well defined turns into a nebulous hairball. Something sleek and well defined: now there's a laudable ambition for CIOs struggling with legacy, ERP, CRM, limited bandwidth and capacity. Most enterprises are still struggling to achieve that degree of integration, never mind incorporating Web 2.0, commercialising it effectively, and protecting brand.

Consequently, I have similar reservations about cloud computing, the name given to huge networks of internet-based resources, as embodied by Amazon.com's Elastic Compute Cloud. It obscures the view and lacks transparency. From a CIO perspective, if you don't have definition and measurement, you don't know what you're dealing with. That leads to bad investment decisions.

Analysts at The 451 Group have been working on cloud computing for some time. Where we've got so far is to identify and define some characteristics. Like the concept, the definition will evolve over time but, at the least, clouds require internet accessibility, a shared infrastructure, on-demand characteristics, and highly flexible management. Dead easy, then. Most enterprises have all of those things already... don't they?

Cloud computing for enterprises will be mostly about culture change rather than technology integration. Sharing resources of any kind implies a maturity of cultural and organisational processes and, with the greatest respect for CIOs, this is not the incumbent organisational model of most enterprise IT.

Indeed, for CIOs, moving to a cloud model could involve destroying years of accumulated processes and organisational boundaries. Similarly, the ability to separate opex from capex expenditure means there are other overarching financial implications for buyers.

Cloud computing today is a triumph of vision over actual form. In just six short months it has become a key industry conversation even though software as a service (SaaS) has taken 10 years to reach its current state of maturity, and only recently has become a truly disruptive force for end users, business models and the software industry itself.

Cloud computing is neither SaaS nor internet applications. Salesforce.com is not cloud computing. Neither is Basecamp in on-demand project management. These applications, or ones like them, may be partially built on cloud infrastructure, but that doesn't make them cloud computing.

My tuppeny-worth? When faced with genuinely innovative and disruptive services and models, CIOs have a duty to be risk aware, even if their personal inclination is to innovate like mad. The role of the CIO is to get information to where it can be put to best and most profitable use. The 'I' in CIO stands for information not innovation and, while innovation is critical for any CIO, they want metrics to back up investments.

It's worth repeating. If you don't know what to measure then the basis on which investment decisions are being made is fundamentally undermined. Are CIOs ready to buy into the cloud computing concept, or is it still a vendor-generated phenomenon? Apart from the innumerable people and process issues of moving to a cloud environment and investing in the supporting architectures, the licensing costs remain a mystery.