Cloud Computing: The answer or another problem? Reply

Cloud computing has been hailed as the answer to a company’s ICT needs.  In brief, the financial benefits are:

  1. Reduced cost of infrastructure, and
  2. reduced labour.

Fulfilling The Dream

At first glance this is brilliant.  By taking advantage of someone else’s expertise and economies of scale a company has the ability to reduce their infrastructure costs significantly.  Not only do they pay less for the same equipment, they pay less for maintenance and also regain valuable office space.  Now, potentially, a firm can actually capitalise on the age-old promise of ICT by not only substituting technology for labour but actually reducing both.

So What Is The Problem?

The problem is that cloud services can require higher integration costs and longer setup times between transactions.  Although the answer is better data integration between applications to improve workflow, I don’t see open source code changing this situation any time soon.  Firms will still need programmers or technical integration specialists to assist.  One only need look at the plethora of Microsoft specialists to see that integration is a bourgeoning cloud industry.

SaaS is rigid by its very nature.  For reduced cost it offers certain parameters on integration.  ERP solutions where integration costs can be astronomical are a testament to this.  The increased costs don’t just stop with integration, however.  Transactional costs (which cannot be attributed to the cost of good sold) also increase as setup times between workflow tasks rise.  The management burden is also greater as additional decision making is required to make sense of isolated data (this may come in the form of downstream consulting or upstream meetings etc). Companies which are unable to take advantage of service/platform/infra outsourcing will notice an overall increase in both their overhead costs and total employment of information workers.  This will not happen in the ICT cost centre where the costs will largely decline.  However, it is likely that ICT will largely pass on a greater and more expensive problem to the business.

The solution is not greater vertical integration of shared services.  It may reduce the overhead in the short term but in the long term the increase in the cost of goods sold will price products out of the market.  One part of the solution is a greater focus on data for decision making through the expert use of case workers (information workers covering multiple transactions).

A key factor is that along with outsourcing our platforms we also outsource our decision making/support.  Don’t forget to retain the information management function and your company won’t succumb to the siren song of cloud applications.

The Cost of Information Reply

In simplest form the cost of information is the aggregate of all information systems (less depreciation), management salary costs and management expenses.  This is roughly equal to the net cost of information in a company.  Unremarkably, the number is huge.


Information is expensive.  It is not merely an amorphous mass of noughts and ones floating about in the commercial ether.  Indeed, that might be the nature of information but costing it is different.  Information is the lifeblood of an enterprise.  In fact, information is the lifeblood of management.  More specifically, the role of management (those not directly involved in the manufacture of the company’s core products or services) is to optimise corporate output by using information to make the best decisions on allocations of capital or resources.  In order to achieve this management requires information for decision making.  This information comes at a cost; the cost of capital and the cost of resources namely:  The cost of capital systems such as information systems (or machines at least with a diagnostic function) and services to provide the information.  This includes overhead costs of shared peripherals (printers, routers etc) and organisational costs such as support staff and training.  In addition to capital costs are the resource costs (the people/knowledge workers) to process the information.


Direct and indirect labour costs are always the most expensive aspect in the cost of information.  The cost of computers has dropped significantly in the last 40 years and yet the cost of labour has risen exponentially.  In accounting terms, per unit costs of computing are down but per capita costs are up.  Consequently, for cost control, with every reduction in the cost of workstations there is generally a corresponding rise in the use of services.  A single computer workstation is cheaper but the cost of running it is much higher. The cost of labour, therefore, will always exceed the cost of capital and so the organisational costs of information will always be the highest.


Significantly, organisational costs increase the further away from the customer the information is pushed.  Just look at any back office and to think that all that activity is triggered from single customer requests.  Information is eventually decomposed and pushed off into various departments.  In this way, the further away from the customer the volume of information increases.  Processing systems and resources proliferate in order to deal with this thoughput.  As transactions increase, so to do set up costs, further systems (network devices, security etc) as well as additional management for governance and oversight.

Transactions cost money.  The higher the number of transactions the higher the setup costs between unit workers, the higher the management for oversight, governance and workflow and, more pertinently, the lower the overall productivity.  Just think about each time a manager turns their computer on (setup costs) to check a spreadsheet, a project plan or approve a budget expense or purchase they increase the costs for the business and slow down the eventual output.

Information is produced through the direct application of labour to capital (people using machines).  Information is a cost that must be apportioned and controlled like any other capital cost.  The high costs of information should not mean that information should be undervalued.  On the contrary, the low capital costs mean that capital purchases are always attractive so long as a business can control the associated labour rates.  After all, information may be expensive but properly developed and properly used it is also extremely valuable.


The moral of the story is that organisations should reduce the cost of people and not the cost of machines.  The market is a good arbiter of the cost of information systems.  They are cheap and haggling over price in the current market is wasted effort.  Try and reduce overall labour costs in departments but more importantly reduce transactional volume by increasing organisational learning.  In this way, single workers will be able to perform more types of transactions and therefore increase process efficiency.  In this way, per capita costs may rise but units of work will also increase.  In the end, the idea is to have fewer but more expensive workers who will ultimately get more done in less time.