A client involved in high-level government decision-making made an observation the other day. We were talking about how decisions actually get made out there in the real world. He suggested that senior decision-makers should either immerse themselves in enough detail to be competent to make the decision themselves; or they should check that a robust decision process is being followed, and then accept the recommendation that emerges. However what usually happens is that they don’t know what to look for in a decision process, and instead engage with the decision itself, but typically only getting into some aspects of the detail.
Now we know that humans can’t do multi-criteria analysis in their heads, which is why software models are necessary (the humans remain important too, otherwise the model would be empty!) What humans can do, and naturally do, is evaluate one criterion at a time. And if two criteria give the same answer, it might be tempting to assume that’s the job done. But if those two criteria lead to different answers, or there are other important criteria as well (which is generally the case), then help is required.
Multi-criteria decision modelling doesn’t have to be onerous or time-consuming. Admittedly it takes some experience to do it well, that’s why specialists exist. I have seen a model built from scratch and pointing to a clear best decision within a few hours, all before lunchtime. And this was a situation where the client had been agonising for months, with individual stakeholders and technical experts all pulling in different directions. Of course it’s not always that easy, but in general the elapsed time taken by the decision process invariably comes down to the availability of the decision-makers and stakeholders, and the need to get the right people face-to-face in a room. A good process will then ensure that the time spent in that room is efficient and effective, with outcomes that are robust to any subsequent scrutiny.