Computers are simple

Computers remember, calculate and communicate. They are valuable because they are quicker, more accurate or cheaper than people. Sometimes they even make the infeasible possible. But people often claim computers can do things they can't. We need a simple way of analysing claimed benefits to see if they are real.

To start getting better value from IT, we need to go back to basics. A computer is a machine that can:

• Remember any information that can be encoded digitally: numbers, words, pictures, sound, and video.
• Perform calculations.
• Communicate to and from its users, and to and from other computers.

Computers add value because they are quicker, more accurate or cheaper than people at remembering, calculating and communicating. For example, an accountancy package remembers the current state of accounts and calculates the new state based on transactions communicated to it, which removes the need for armies of clerks.

Computers can also provide second-order benefits when they make the infeasible possible. For example, an ecommerce web site allows a company to communicate its products to customers all over the world, which would not be feasible without computers.

Second-order benefits are still based on the underlying capabilities of the computers. The ecommerce site is beneficial because it makes the infeasible possible by communicating so much quicker, better or cheaper than people could.

This type of analysis can break down sales messages. One IT product claimed it would "transform your corporate culture". To see if this claim is valid, we need to ask, "What does this system remember, calculate or communicate, and how does that transform corporate culture?" If we can't answer this, the claim is false.

This analysis also helps to separate your ability from the capability of the computer system.

A CAD program advertises that it will "let you design in 2D and 3D". Really? A CAD program can remember my design and the graphics it should use to represent it. It can calculate what my 3D design looks like from any 2D viewpoint. It can communicate my design back as drawings, and manage changes to the design. But I can't use it to design a gearbox, because I'm not an engineer. The computer system may be a great help, but it shouldn't claim its users abilities as its benefits.

Lastly, this analysis helps separate out user benefits (reasons for having the system), from supply benefits (why this system is better than another). Often systems are described as flexible, standards compliant, or cheap. These may be attractive, but they are not of themselves reasons to use the system.

To recap, all benefits of using IT boil down to being quicker, better or cheaper than people, or making the infeasible possible. IT does this by automating remembering, calculation and communication. Use this analysis to check that claimed benefits:

• Could possibly be related to a computer system.
• Come from the computers' capability, and not your own ability.
• Are user benefits, not supply benefits.

Anything else is hogwash.

Cut your IT costs by 90%

Though it sounds unlikely, it is possible to cut IT costs by up to 90%. The key to this is managing IT demand. This isn't simply a matter of saying "no", but requires a reappraisal of the nature and value of IT.

What would you say if I claimed you could cut your IT costs by 90%?

You would probably dismiss it immediately. It can't be true. It doesn't make sense. If it was possible, everybody would be doing it.

But it is true.

How? By managing IT demand.

This doesn't mean just saying "no" to every request for IT. Managing demand involves:

• Understanding what IT is and how it adds value.
• Delivering only the IT that adds value.
• Identifying and avoiding the pressures that lead to excessive IT costs.

How much you can save depends on you and your situation. You might be able to save 20%, 50%, 90%, or even 99%. Even a 5% saving would be worthwhile, wouldn't it?

I don't suppose you are convinced, so here is some food for thought.

Over the past few decades we have become much more efficient at supplying IT goods and services. And yet far from getting easier and less risky, IT has become more difficult and more risky.

Every day there are stories in the newspapers about major IT projects that have gone horribly wrong. Household names like Sainsbury's have been brought to their knees by IT system failure. The government is constantly embarrassed by high-profile projects like the Air Traffic Control systems. Studies [1] estimate that 84% of IT projects fail to meet their objectives, 91% in large organisations.

Surely IT shouldn't have these failures, and their associated costs. IT is based on the promise of efficiency. Every decade the cost of IT capacity decreases by orders of magnitude. Millions of talented people are employed in IT. IT has mature disciplines of software engineering, business analysis and project management. IT should, by now, be cheap, well understood, and low risk.

The main reason for all this waste is that we have only addressed part of the efficiency equation. We have learnt to supply IT very efficiently. But we have ignored demand efficiency. We never limit what we want. Gains in supply efficiency are eaten up by ever increasing demand.

Most of this demand is unnecessary. It does not benefit the people and organisations that it is meant to serve. It gets in the way. It adds time, complexity, cost, and risk. Global IT spend is now in excess of one trillion dollars. Doing away with excess demand could save a large part of this.

Over the coming weeks, this newsletter will probe deeply into the problems of excessive IT demand, and excessive IT solutions. It will show how to understand the true nature and value of IT, how to craft minimal IT solutions that meet the true demand, and avoid the pressures that lead to excessive IT costs.

Next week's newsletter will start by expanding on the true nature of IT, how it delivers value, and a simple way of checking whether you will really benefit from an IT system.

[1] Standish, quoted in Does IT Matter, by Nicholas G Carr, published by Harvard Business School Press.