How to make the most of cloud computing

NB: A shorter version of this article ran as part of the NBR's special report on cloud computing.

Michael Snowden is the owner of IT services company OneNet.

Cloud computing offers many benefits, including lower IT costs, stronger security and a more agile business.  How should you build a business case to reap these rewards?

Firstly, ask the question: what is the IT job we are trying to do?  Do we want IT to help run the business or to change the business?  The latter usually requires new software and processes for a new or improved business model and is outside the scope of this article. 

Otherwise, the IT “job-to-be-done” is the delivery of the firm’s software to its users, accessible from wherever they may be, securely, and quickly.

Is running IT assets a core competency that you must have to succeed, or is up to 98% of your IT budget consumed just to run your business, leaving little scope for innovation?  Can you generate any competitive advantage from owning and running your own IT?  Is there a better way?

Cloud computing may help, but how should you start?  The first stage is to better understand the nuances of the words “cloud computing”.

In essence, cloud computing is provided by a third party who owns the IT assets and delivers their functionality over the Internet for a fee. 

If you have been persuaded to buy or lease some IT assets for cloud computing, you are not doing cloud computing at all. Traditional IT vendors, their resellers and many in-house IT staff like to describe internal IT a “private cloud”, and hence the confusion.  This just puts a “cloud-washed” vendor label on business–as-usual IT. 

Surveys on the adoption of cloud computing reflect a higher rate of adoption for “private” or internal clouds.  But this simply reflects the ambiguity and obfuscation generated by traditional IT vendors stretching and breaking the concept of cloud computing.

Why is this?  Cloud computing is inherently a business model innovation that disrupts and threatens traditional IT distribution business models. The promulgation of fear, uncertainty and doubt about external clouds will extend the life of the status quo and its beneficiaries.

The word “cloud” is a metaphor for the Internet.  Ask yourself: how can you take computing out of the “cloud” and still call it cloud computing?

An internal cloud is likely to be “sold” to management on the basis that it is new technology, it is more secure than an external cloud and/or that firms should firstly virtualise their servers and applications before going to an external cloud.

This is wrong on at least two counts.  Firstly, cloud computing is not a “new” technology, but an alternative economic consumption model.  Secondly, virtualising your current servers, or making them more efficient, will invariably require substantial new investment in server hardware and virtualization software, together with professional installation fees.  Not to mention high on-going support costs.  Doesn’t this sound like traditional IT with a different name tag?

Conversely, an external cloud provider will typically install the very same applications on their own servers and virtualization software and deliver the software to the firm’s users, usually for a fee per user.

When you compare the costs of an external cloud to an internal cloud, it is very important to dig deep and ask the right questions.  The questions will differ, according to the size of your business and how IT is currently provided. 

An external cloud is a clear threat to most internal IT staff members who support the firm’s IT infrastructure.  Turkeys do not vote for an early Christmas.  Accordingly, the cost analysis may be biased by a conflict of interest.  You will need to seek full transparency of costs.

For example, some IT-related costs do not get coded to IT general ledger accounts.  Few firms measure the power required to run internal IT resources.  The cost of electricity now exceeds the initial capital cost plus maintenance costs for the servers over their life cycle.

The assumption of incremental internal costs will also bias the results.  That is, an internal assessment will likely skew the analysis because it will make the assumption that you will continue with most, if not all, of the current IT infrastructure and support arrangements.  The best way to overcome this built-in bias is to take a “clean sheet of paper” approach and ask what you would do if you were starting from scratch.

In addition, opportunities to quickly reposition the firm may be lost if the firm’s infrastructure is ossified and inflexible.  Lost revenue resulting from unreliable systems will not be captured as an IT expense in the general ledger.

An internal cloud will perpetuate the problem of over and under-provisioning of IT investments and skills.  The costs of always having too much, too little or the wrong mix of IT assets and skills do not have a general ledger code.

Internal IT maintains unnecessary complexity in your business.  The cost of complexity, in something you cannot get any competitive advantage from, together with the costs of reduced business agility, do not have a general ledger code.

While these IT-related costs are rarely measured explicitly, few CFOs or CEOs will not understand the very real costs.

Vendors of internal or “private clouds” tout the benefit of IT “chargeback” whereby the costs of IT are apportioned according to how much is used by each operating unit.  Good luck with that.  No matter what you do, your internal costs will be fixed for a reasonable period of time.  Try returning leased equipment or selling surplus gear and licenses.  Try to rapidly scale your IT infrastructure team up or down.  Any internal cloud charge-back system simply allocates a fixed cost over some slightly more meaningful criteria.  No matter what you do, you are still stuck with either too much or not enough IT resource for a fixed cost in the short to medium term.

In contrast, an external cloud will provide the most transparent measurement of cost, because it is based on a consumption model.  More importantly, there are multiple vendors competing for your business, all driving costs down and quality up in a highly competitive market.  How could you possibly get such market rigour and discipline from your internal cloud?

An external cloud will eliminate the challenge of scalability, or having just the right mix of resource at the right time.  An external cloud will remove complexity and deliver simplicity.  Remember, the job-to-be-done is to deliver your software to your users, quickly and securely.  By paying for this on a variable cost basis means that you will only pay for what is actually consumed, not what previous IT investment decisions have dictated for the future.

The security argument against an external cloud will likely arise.  While security is always of the utmost importance, it is the threatened IT employee’s best weapon of defence against the prospect of lost employment resulting from a move to an external cloud. 

Few CFOs have more than a rudimentary knowledge of computer security and the IT industry’s legendary penchant for unintelligible acronyms and confusing, fast-changing terminology provides the incumbent with a quiver full of arrows to seed uncertainty and doubt in the decision makers’ minds.

The reality is that enterprise-class cloud computing providers view IT security as a core competency and will more than match a firm’s internal capabilities for all but the largest firms.  Due diligence and an external security audit of both the current internal and preferred external provider’s security will likely allay any concerns about security and IT governance.  Indeed, an independent IT security audit on the current internal system may uncover some surprising and alarming issues.

Unless a thorough independent analysis of “apples-for-apples” quantitative and qualitative factors is done, it is highly probable that your cloud computing business case will come to the wrong conclusion.

Even if you decide to stay with internal IT, at the very least you will gain a great deal by carefully measuring the opportunity cost of doing so and you will better understand the costs and benefits that you will have foregone.