As an architect, every so often you have to make a business case for something where you hit some unknown factors that can’t really be quantified, at least not without a significant margin for error. Two typical examples for me have been firstly “risk avoidance” type investments in process compliance tools etc. where there’s an upfront investment but no real ongoing business benefit (or even a minor ongoing loss due to lower process efficiency). The upside of an investment of that sort is typically to avoid audit failures, avoid submitting incorrect tax/duty reports etc. which may not happen, but will be extremely costly if they do.
The second part is a benefit where there is a profound disagreement among stakeholders on the value provided. One example of this scenario: Many consumer-goods companies have direct distribution to end-customers in many of their markets. A typical discussion is whether this distribution operation should be outsourced or kept in-house. Most factors in this discussion can be quantified; capacity requirements, fleet utilisation, running costs, driver qualifications, training and so on.
The one factor that definitely can’t be quantified (well, one of them, but a big one) is the value of having your own people visit the customer on a daily basis. Everyone agrees there is probably some value in this, but how much exactly and how much it is actually utilised is generally impossible to pin down. So, one suggestion to break the deadlock is to “turn the business case on its head”, meaning instead of asking “how much is this worth to me?” you ask “is this worth what I am paying for it?”.
In the example above, that means for example taking the difference between the two cost models and dividing by the number of customer visits made per year. That gives a cost per customer visit which is then the premium that we pay for the privilege/opportunity/hassle (depending on your viewpoint…) of distributing to the customers ourselves. Having that figure then allows for subsequent discussions along the lines of “what else could we get for that money?”, which then in turn allows you to zero in on the exact value for the business.
Using this approach doesn’t completely solve the problem of trying to quantify the unquantifiable, but it at least gives some options for asking questions, mostly along the lines of “does this seem like a fair price for what we are getting?” or “could we do more with the same money if deployed elsewhere?” etc.
These questions and answers in turn give a much greater clarity and a much more solid base for discussion than each stakeholder picking (and defending) their own random savings figure – which seems to be a fairly typical alternative.