Cloud Computing – The next big thing or the Emperor’s New Clothes?

27 February 2012

Cloud services have emerged as a genuine threat to established hardware and software providers as well as to the many thousands of IT professionals who have made their living selling and supporting these products over the last twenty or more years.

This threat makes ICT investment decisions by businesses particularly challenging as we are encouraged by one group to stick with the tried and true and by another to throw away the old and bring in the new.

How can the average business customer judge the importance of a new technology or the continued relevance of existing technology choices to their business?

While the answer to this question can often be complex, at its heart is a simple challenge “What is the purpose of this technology?”

This three part series assesses the importance of cloud services to business by looking at whether a particular service reduces costs and/or improves businesses performance when compared to the existing way of doing things.

Part One of this series addresses Infrastructure as a Service – where traditional ICT services such as storage and processing power are provided.

Parts Two and Three will address:

Platform as a Service – where organisations can purchase the toolkit necessary to build and operate their own software

Software as a Service – where organisations can access software such as email or office productivity tools.

Part One

Infrastructure as a Service

So what is the purpose of this service? In essence this service aims to replace existing infrastructure and instead deliver it via the internet.

In my view this service usually fails the “so what?” test simply on a cost effectiveness basis.

Let me give you a simple example.

Companies offer you storage on line at a price – say $1 per annum per unit of storage. Often this price is at a discount to what a business might pay if it bought and ran the storage on its own behalf.

The trouble with this offer is that the amount of electronic data your company will store is growing every year by between 50% and 100%. Thus if you buy one unit of storage in year one and your data needs grow at 50% then you will be paying $8 per year for data storage in just over five years.  Scarily this amount will become $16 eighteen months after that, then $32, then $64 and so on. At a time when ICT budgets are being carefully examined and even reduced this is not an offer your Financial Controller should relish.

In the meantime the price of actually buying and running your own storage is halving every eighteen months. Thus if you paid the same $1 for your own storage in year one that same unit of storage would cost around 12.5 cents in year five. Thus your cost to own and store after five years, even when compounding data at 50%, is still only $1 per year and will remain so by and large.

Admittedly the complexity, and therefore cost, of supporting and maintaining data storage also rises with the amount we store but this is the same for both providers – except that the company outsourcing its data storage is now paying eight times the price per unit stored – at least until it renegotiates its contract.

This situation makes it hard for both the supplier and the buyer. The supplier wants to sell a longer contract in order to pay back their investment while the buyer wants a shorter contract to take advantage of the fall in data storage prices.

So why would business buy storage via the internet?

Let’s go back to our basic premise. What problem are we trying to solve?

One big business problem is protecting valuable data for legislative, company policy, records management and disaster recovery purposes. Many businesses struggle with the complexity, cost and reliability of backing up this data. It is not uncommon to hear of businesses who try to restore lost data and find that there has been a failure of technology or process that makes it very difficult or even impossible to get their data back.

Backing up data via an internet service can provide businesses with sophisticated technology and processes as well as contractual guarantees that would normally be well outside their normal financial and personnel capabilities. Thus the risks that are mitigated could well justify the extra expense incurred.

Infrastructure as a service also makes sense when it fits a temporary infrastructure need. This might occur, for example, in a general election when large amounts of processing and storage may be required for a short, finite period to handle electoral roll and vote counting.

Outsourced storage costs can also be reduced by insisting on regular contract reviews and ensuring that technology processes are in place that reduces the problem of replication – unnecessarily backing up the same data over and over again.

If a business is cash flow rich but capital poor Ieasing may offer opportunities to grow by creating ICT advantage without having to sink a fortune into depreciating assets – but that is a whole other article!

In summary

Mark: FAIL – must do considerably better

Careful analysis based on the simple question “what  is the purpose of this technology?” will reveal whether your business should go further in investigating Infrastructure as a Service. Right now, if I were the Financial Controller or CEO I would be asking some very searching questions and running the numbers very, very carefully.

We may see significant price cuts in the years ahead as multinationals and even local corporates offer New Zealand businesses storage based on their marginal cost of supply. Until then local New Zealand suppliers of Infrastructure as a Service may struggle to come up with a convincing offer that encourages businesses to change from the old to the new.


David Spratt is an associate with TUMG and has held senior ICT strategy and technology selection roles in business for over thirty years – he can be contacted on 027 574 9141.