16 January 2007

## Justifying IT investment 2: calculations

We can model the financial benefit of any IT investment by using a handful of simple calculations.

We need to calculate financial benefit for all IT investments. Business-IT investments are relatively easy, because we can base the benefit on predicted business changes. It is much harder to estimate benefit for purely IT investments, for IT infrastructure. Without ongoing investment systems decline. They become out-of-date, badly maintained, unsupportable, inflexible, and non-compliant. They end up costing more. To avoid this, we have to calculate financial benefit so that we can make the case for ongoing investment.

Last week I covered how to use system governance techniques to score systems against a basket of criteria. This shows where improvements can be made. I also covered calculating the overall volume of IT. This week I cover calculations that translate these measures into financial benefit. This shows which investments are truly worthwhile.

The financial estimate uses a simple equation for IT costs (excluding the cost of new business changes). It assumes that cost is related to the volume of IT, multiplied by a factor that represents the difficulty of IT, multiplied by unit cost.

cost = volume x difficulty x unit cost

Difficulty can be split into two.

• Base difficulty. This is the difficulty of IT under perfect conditions, when everything is up to date, perfectly maintained, compliant, and so on.
• Distance from perfection. Assuming the scoring system goes up to 100%, the distance from perfection is 100 - score.

Putting this together gives:

cost = volume x (base difficulty + ( 100 - score )) x unit cost

This equation can translate improvements in score to a reduction in cost. (We also need to calibrate the model with figures for unit cost and base difficulty, which I will cover next week.) If s1 is the current score, and s2 is the predicted score if improvements are made, the current and predicted costs are:

current cost = volume x ( base difficulty + ( 100 - s1 )) x unit cost

predicted cost = volume x ( base difficulty + ( 100 - s2 )) x unit cost

Annual benefit of the improvements is:

benefit = current cost - predicted cost

Which resolves to:

benefit = volume x (s2 - s1) x unit cost

You can use this benefit figure as an input to whatever financial analysis your organisation uses.

For example, if volume of IT 350 and unit cost is £960, an IT investment that increases the score from 75.0 to 76.2 would be valued at:

benefit = 350 x ( 76.2 - 75.0 ) x £960 = £403,200 per year

If this investment cost £600,000, then on a simple payback basis, the investment would pay for itself in a little under 18 months.

You can use the same calculation to value improvements to a single system by using volume and score figures for just that system.

This method is not perfect. But it is rational, understandable, and directionally correct. It can be applied to purely IT investment, IT infrastructure, which is hard to justify using traditional business benefit measures. It makes a financial case for ongoing investment, which helps us move to low risk and low cost IT.

Next week I will cover how to calibrate the model with figures for unit cost and base difficulty.