Thursday, May 27, 2010

Travel Insurance, Real Options, Cloud and VFM from IT

With the recent focus on cutting IT waste in government - a thought experiment. Before going on vacation you purchase a one week travel insurance policy for £15. You go for a weekend in Paris and return without incident. Was the money spent on the policy value for money at the point of purchase ? Would your answer be different looking back if you had a wallet picked on the Paris metro and ended up claiming £50 ? Or if you were hospitalised with some exotic virus, also caught on the metro, and had your medical expenses covered ? Did you get Value for money ? Would an annual multi-trip policy that cost £100 provide more value for money if you normally make 10 trips abroad ? Would your answer be different if you ended up making just 4 last year ? When you go up to buy the policy how do you assess if you're getting a 'good deal' ? Is it the cheapest policy on the market for the same features ? Is it a better deal not to buy insurance at all ?

Bizarre as it may sound, almost all large capital expenditure decisions exhibit similar characteristics and a similar difficulty in getting non IT related decision makers to understand the real value from significant IT investments. This non-obvious value derives from the risk reduction and future flexibility that an organisation gets when they put in significant investment into IT capabilities.

Almost every IT investment proposition of reasonable size gets put through a "Value for Money" ( VFM) assessment. This is due to the combination of the increasingly large proportion of IT capital expenditure in a typical organisational budget and a recent legacy of IT silver bullets whose effects were closer to those of the projectile than the precious metal.

One major danger in making these assessments though is the over-simplification of the measures employed in objective assessment. For example, how do you choose the best option when confronted with multiple options of varying degrees of complexity and benefits. The traditional and the most common way of doing this is to evaluate IT capital expenditure as a upfront investment that will give a stream of of quantifiable benefits in the
future. This is very much an NPV/IRR approach and by far the most common in industry. The basic method involves estimating future benefit streams and adjusting them by a risk factor to take into account the uncertainty associated with the benefits as well as the time value, earlier benefits being weighted higher.

IT investment, however, is not so easily categorised in this simple model because it has value in two crucial areas that these dimensions do not measure :
- Risk Reduction.
- Flexibility in dealing with future circumstances

In both these cases, the value of the investment largely derives from investing now for either avoiding situations that MAY arise in the future or reacting to such situations. The situations themselves, of course, may not materialise at all. This however does not mean that the investment does not have any value.
I started encountering this issue about 5 years where a significant number of enterprises where considering, or at least were being seduced by SOA based investment opportunities. A typical SOA initiative proposed by an IT department involved a signficant investment in platform infrastructure in Enterprise Service Bus software, Canonical Data model and schema development, the modelling of Business Process orchestrations etc. This was expensive and proved hard to defend on anything other than a "visionary" basis since VFM assessments forced people down the hard quantifiable benefits road. The value of the investment that derives from the SOA investment making the organisation more flexible was typically not assessed
since the flexibility is only measurable in relation to a range of future possibilities that may or may not arise. For this reason, a number of Enterprises opted for small SOA-lite initiatives, nto because the lack of "vision" but mainly because they lacked the tools to assess the longer
term value of flexibility. I see similar issues these days in clients dealing in particular with investment with "Green" themes or significant investment in creating private-cloud models of infrastructure provision where a one off upfront investment buys future flexibility which is hard to value with existing tools.

How can a decision maker take this into account whilst evaluating competing proposals ? Tools such as Real-Options analysis are out there but get into advanced mathematics too quickly and lose their intuitive appeal and get too hard for the average organisation to use effectively. The other technique I've seen organisations use is to make the risk reduction and flexibility features a standard part of the requirement against which they get competiting compliant proposals and then choose the cheapest. The results, though leave no mechanism for value-add to be accurately assessed.

What else are people using to assess value from IT ?

Labels: ,