Dave Dixon wrote a nice post a little while ago – Why You Should Care About Shareholder Value – and I have been mulling writing a response. Dave works at Provisdom and has recently started a blog there (despite being a fairly new blog, he has already written some interesting posts so check it out and subscribe). Anyway, Dave talks about the value of using shareholder value to “explicitly evaluate the effect of choices and future uncertainties on shareholder value”. This is an often under-valued aspect of adopting a decision management mindset. Too often operational decisions – those with high volume but low individual value – are not managed or controlled explicitly. Front-line staff make the decision as best they can or programmers embed their thinking in code. Enterprise decision management, in contrast, takes control of these decisions to ensure that corporate objectives are applied consistently. As Dave notes, using shareholder value as a driver for this is particularly effective. Doing so extends the maximization of shareholder value from what Dave calls “goal … common to all users” to a goal common to all users and systems so that all operations “reflect the core goal of the entire corporation”. A powerful concept indeed because, as Dave notes, “EDM is typically applied to high-volume decisions” so that “even a small improvement in the value of each decision quickly adds up”.
Dave talks about Decision Yield, a concept Fair Isaac has been developing (see this section on the Fair Isaac blog) and that we discuss in the book in an appendix. This is better matched to shareholder value analysis than Dave gives it credit for as the purpose of decision yield is to consider how to improve a decision in one of the five dimensions but the definition of “improve” would almost always be to boost shareholder value overall. Decision Yield simply let’s you consider different aspects of a decision – Precision, Consistency, Agility, Speed, Cost – to see which will have the most impact on shareholder value.
Another aspect of EDM that is relevant here is that of adaptive control (covered on Fair Isaac’s blog here and in a whole book chapter). This allows you to work through various scenarios to see which will have the most positive impact on shareholder value. Some transactions are run through each of several different decision models to see what works best. When these models include predictive models built with different assumptions, you essentially include “future choices, uncertainties and payoffs” in your decision making.
Dave also talks about risk and discount rates and I don’t have much to add except to say that building risk into a decision – credit risk, retention risk, long term profitability risk – is one of the main reasons why decisions take both business rules and predictive analytics. Predictive analytics take uncertainty (about the future) and turn it into a probability so you can act on it.
While Dave does a good job of summarizing how to use EDM in systems development I would also recommend these two posts on decision services (1,2)
