Cloudy with a little sunshine

The IT industry has spent so much energy over the last couple of years talking about clouds you could be forgiven for thinking it might have usurped the role of the Bureau of Meteorology. Mind you given the Bureau’s track record for forecasting accuracy many in IT would probably feel quite at home there. Moreover, it seems there are just as many varieties of ICT clouds as there are meteorological ones.

A recent review of Cloud Computing by the UK based analyst firm Freeform Dynamics posed the question just what is cloud. Not surprisingly for the IT industry it reached the conclusion that the definition depended on just who was trying to sell you something. This answer though did highlight the versatility of the cloud approach. To some it is about delivering services over the Internet as and when they are needed. To others it is about utilising the Internet to harness external infrastructure such as servers and storage to compliment your own. This is extended by other proponents of cloud like Amazon who argue that it provides an opportunity for a CIO to provision infrastructure for short term or spasmodic needs such as application development. Then there are those who advocate that Cloud represents Software as a Service or SaaS which entails accessing external applications like Salesforce.com on a pay for use basis rather than licensing these applications internally. Finally, there are the Cloud supporters who argue that this represents a move to the IT data centre architecture of the future where no one owns anything.

In the uncertain post-GFC business world of today many of these arguments clearly resonate for many businesses, especially from a financial perspective. With IT typically the third biggest expenditure item in most corporate budgets, after staff and real-estate, it is a conspicuous cost that many executives probably feel needs to be better contained. In moving IT investment away from big ticket capital expenditure towards an incremental, pay as you go, approach Cloud clearly has appeal. This use of operating expenditure to recoup ICT costs also appears to offer a better way to attribute IT costs against the business units that are actually using the ICT services.

To many ICT executives this promised greater transparency in accounting for the IT budget clearly has appeal. However, many also recognise it has the opportunity to free many staff in the IT department from operational housekeeping so they can make a greater contribution to the strategic aspirations of their organisations. A Telstra/MIS Magazine study in 2008 of Software as a Service revealed that 45% of IT executive respondents were attracted to this approach as it freed them from the need to install and maintain software. Moreover, 37% of respondents to the survey also saw that SaaS would liberate them from the never-ending task of application upgrades. IT executives believed that these benefits would lead to cost savings.

However, the same survey also highlighted concerns that local IT executives had about going down this path. In particular, there were strong concerns about security and recognition that management of the corporate data was too important a task to leave to others. This no doubt reflects the fact that corporate data is probably a company’s greatest asset as well as the increasing privacy obligations being placed on businesses to safeguard the data details of their clients. Besides security the other issue the survey highlighted was a concern among IT executives that if there were no Internet service available many users would be left with an inability to access the applications through which they do their jobs.

Interestingly, these security concerns are supported two years later by the Ernst and Young 2010 Global Information Security Survey. One aspect of this study was devoted to issues around Cloud Computing and, in particular, respondents were asked to nominate what they saw as the security challenges of Cloud Computing. The findings revealed that the chief concern of IT executives with Cloud Computing was data leakages where confidential or sensitive data was forwarded to outside parties. There were also worries about an inability to track what happens to company data and fears that unauthorised parties would be able to access corporate information.

Nevertheless, despite these worries many ICT executives also realise that Cloud Computing is a phenomena that they cannot just ignore. Business demands greater transparency in where it spends its money. As such, all executives have a responsibility to be measured and creative in how they spend money. The days where large capital IT projects were the norm are probably numbered, particularly when the evidence is that many of the resulting IT assets are under-utilised. Moreover, the ability to access resources on an as needs basis and better attribute these expenses to the business unit accountable clearly has an appeal for CIOs subject to the relentless scrutiny of their peers as to why ICT budgets are so high.

Therefore, it seems that the decision for many IT executives is not whether they will deploy Cloud Computing but rather where they will do so. In this regard the thoughts of the US consultant Geoffrey Moore may be helpful. Moore believes that as a starting point in their Cloud Computing deliberations CIOs need to distinguish between whether an application is a core system or a context one. He classifies those as core as the ones that offer competitive differentiation and which are, typically, mission-critical. He regards context applications as usually internal ones that support business activities like Human Resources and Payroll.

In Moore’s mind the unavailability of a core mission critical application can threaten the viability of the organisation. As such, he advocates that these should always be run internally behind the organisation’s firewall. On the other hand he believes that anything context and non mission-critical should always be placed in the Cloud. However, he also recognises that there are occasions when a context application can be mission critical, (e.g. Payroll systems on pay day), and there are times when a core application is not mission-critical (e.g. many Business Intelligence systems). Moore believes that the job for CIOs debating the applicability of Cloud Computing is to focus their energies on determining if it makes sense to put core non mission-critical applications in the Cloud and whether it is better to run a context mission-critical application in-house.

In the end, though, perhaps the biggest question mark hanging over Cloud Computing is a philosophical one. It is human nature to trust what we can control. Cloud Computing entails delegating some of that control to someone else. Letting go is never easy but holding on is not always productive. Yet if organisations are going to be competitive CIOs clearly must help them by being more productive themselves. As such, ICT executives are probably likely to find, at least for the foreseeable future, that the IT industry is going to continue to outdo the Bureau of Meteorology when it comes to talking about the potential of clouds.