Autumn saw a cloud computing conference in London every month - October, November and December. I again see the risk of a lot of over-enthusiastic hype endangering an emerging transformational development of great potential. Little wonder that, as an industry, we fail the ‘Radio 4 Today Programme Test' so frequently - our reputation is of endlessly over-promising, combined with a repeated failure to deliver.

So here are two real-life developments that relate to the cloud which I will use to test aspects of hype versus actual practice.

CSC has announced that it has signed an agreement with the UK's Royal Mail Group to provide "cloud computing IT services". The 23 November press release reports that "the new agreement expands the company's current contract signed with Royal Mail Group in 2003 to maintain its desktop computers and manage and develop its servers, mainframes and IT processes. CSC will provide Royal Mail Group's 30,000 employees with access to new IT services using Microsoft's Business Productivity Online Suite (BPOS), part of Microsoft Online Services".

Is it, I wonder, genuinely ‘cloud' or a classic facilities management outsourcing deal where CSC will apply effective virtualisation to the Royal Mail Group's outsourced computing resources, thus sharply reducing costs through much improved asset utilisation?

Gartner's Paul Bittman identifies four fundamental attributes that he posits differentiate a genuine on-premise private cloud from a well-run in-house IT organisation or outsourced FM deal:

1. Services-oriented. The in-house IT or outsourced FM teams sell services, and the customer only deals with IT through self-service sourcing of the service offerings.

2. Variable pricing, based on usage. The in-house IT organisation or outsourced FM team maximizes technology sharing and automation for efficiency.

3. Elasticity. Customers deploy/grow/shrink/retire services quickly. IT ensures fluid scaling ‘on-demand'.

4. Independence. The in-house IT or outsourced FM team could move a service to the public cloud by flipping a switch.



I would add a fifth fundamental attribute - a genuine cloud-sourcing arrangement does not require a term contract between customer and supplier. Sourcing data processing and storage power from the Amazon Elastic Compute Cloud is on-demand: you source as you require and pay for what you use, and developing and launching new apps on Salesforce.com's service does not require a term contract with Salesforce.com.

At November's Cloud Computing World Europe conference in London, Kerny Ustrup, who is group vendor manager at Danish shipping and oil giant AP Moller-Maersk, gave a detailed presentation on ‘Outlook As A Service - OAAS'. He recalled the migration from a trio of tailor-made legacy mail solutions to a new cloud-based standard solution delivered by Microsoft Online as a Service. This legacy setup required a two-year transformation to prepare for the migration.

Kerny's contract is directly with Microsoft and was signed two years back for five years. In his presentation he admitted that the contract timeframe could have been shorter - but balanced this with a very open opinion, shared with the conference audience, that Microsoft as a cloud services provider is still developing and maturing - and learning a lot from AP Moller-Maersk as the client.

In contrast, the Royal Mail Group is partnering with CSC in its sourcing of the Microsoft Outlook OAAS/BEPOS. While the press release says nothing about the length of the renewed deal, there clearly is a classic term contract in place.
So, back to three observations on the fundamentals beneath the hype.

Firstly, sourcing a genuine cloud-sourced service such as OAAS should not require the intermediation of a third party. Unless, possibly, transformational work is required at the customer in order to be able to directly source the service from the cloud. AP Moller-Maersk is handling that transformation itself. At the Royal Mail Group, it appears that CSC is the transformational partner.

Secondly, a genuinely cloud-sourced service does not require a term contract, and the precedent set by the likes of Google and Amazon.com underpins that reality. In practice, established IT services players are deeply wedded to business models requiring term contracts so there is transformational work required at the supplier in order to reposition as a genuine longer-term purveyor of cloud services.

Thirdly, if the real potential of the cloud model is to be delivered in practice, I would expect these OAAS services to be managed and optimised in a blended fashion on Microsoft's rapidly growing global service infrastructure, with neither customer needing to know ‘on what piece of tin (or where) is my stuff'. In the cloud-sourcing world; that is solely Microsoft's responsibility as the service manufacturer.

A brave new world? Indeed. But let us use the ‘cloud' word with precision, please.