A Better Approach To Project Selection

LinkedIn and Facebook are very similar, but also very different. LinkedIn is how we act at the Christmas party; Facebook is how we act at the bar down the street after the Christmas party.  One of the things I respect about LinkedIn members, for the most part, is that we seem to “self-police”.  LinkedIn members typically meet political propaganda with something like “there’s no place for that nonsense here!”

There is an interesting point to make about politics without selling anything but a good practice. I’m only raising it because of its application to the private sector.  In a politically charged argument my best friend once said (about a decision by a past vice-president), “how can you know what the right thing to do is? You weren’t there; you don’t know all the details so you don’t know what the right decision is!”

My friend was implying that by the mere fact that I was not personally involved in the particular situation that I am disqualified from analyzing the decision because I don’t have all the information. That is a mathematical claim that happens to be false. Let me illustrate with a more concrete example that is less emotionally charged than politics.

I don’t gamble, but I watch it on television sometimes.  I remember watching the World Series of Poker where players compete in the game of Texas hold’em. If you are familiar with this game or have seen it on television you can imagine the following fictitious scenario:

Daniel Negreanu and Phil Hellmuth are the remaining players at the final table. Although the match is being televised, viewers can only see Daniel’s hold cards. Daniel is dealt an ace and king of spades.  On the flop, a queen, jack and ten of spades are drawn.

At this point you don’t have to look at Phil’s hold cards: no matter what he has, no matter what the turn or river card is, Daniel wins. It’s just a matter of how much Phil is going to lose and how long it’s going to take.

This is an extreme example but it illustrates the value of information. There is no value in analyzing anything beyond Daniel’s hold cards given the flop, or vice versa, because there is no chance of being wrong.  When there is no chance of being wrong or no cost of being wrong there is no value of additional information.

Why is this applicable to the private sector?

If you work for an organization that gets together at some frequency, perhaps annually, to allocate capital and expense dollars to investments or projects, this is particularly relevant.  If you are like most organizations you might evaluate the investments based on business cases presented by a variety of business units in your organization. 

How do you know these represent the best opportunities? How can you be sure the costs and benefits are presented accurately? If you can’t answer this question, it doesn’t matter what project methodology you use for delivery; be it Six Sigma, Waterfall, or Agile to name a few popular approaches, it doesn’t matter if you chose the wrong opportunities to begin with.  But let’s hold that point for a later discussion.

Depending on the strength of the business case you might narrow the number down to some arbitrary set of finalists. Then, a management team or governing body might deliberate on some kind of score card ranking or even unaided intuition to get a prioritized list.  Sound familiar?

This approach is a waste of time. In fact, you are better off writing the names of the projects or investments on pieces of paper and randomly drawing them from a hat – at least you saved time foregoing the time and energy going through the balanced scorecard analysis.  We know this to be true and yet organizations large and small go through this process every year.

Any balanced scorecard method is easy to debunk; in fact, there is a lot of research on this. You might feel better about the outcome based on the process, but there is no reason to infer that these scorecard methods improve decision making over time – in fact, the exact opposite is often the case.

What’s interesting to me is the value we often place on unaided intuition. I like to sum up this approach as follows, “When you hear someone start a sentence with, ‘it’s been my experience that…’” The best thing you can do is stop listening; you’re better off daydreaming about your next vacation or what you plan to do that evening.

This sounds harsh, but it is none-the-less true.  Unaided intuition, as a decision mechanism, shows no improvement over time; again, the exact opposite is often true.  This claim is also backed up by a body of research that is undisputed.

Think of it this way:  Imagine you have a room full of CEOs, each flipping a coin thousands of times.  You would expect to see 8 or even 10 heads in a roll.  Would you describe the CEO who flipped 10 heads in a roll as a “great coin flipper?” Obviously, the answer is no. Yet, we use this logic to evaluate executives and other decision makers all the time.

The point is, there is no substitute for predictive models and analyzing the value of additional information.  Even if you don’t have all the information, the best approach to decision making under uncertainty is stochastic modeling. In fact this is precisely the reason for using predictive models.

Further, even if you don’t know all the goals of the stakeholders, you can still improve their situation (or the organization’s stock price) by modeling the current state of uncertainty, determining the value of additional information, and then taking a few secondary measurements to optimize a decision.

In the poker example above, I have no idea what Phil’s hold cards are, I haven’t seen the turn or the river card – most importantly I have no idea what Phil’s goals are – I can only assume they are to optimize his next decision. His best option is to fold.  

Think about this the next time you find yourself basing a major decision on unaided intuition or the results of some scorecard method – you are using an inferior method.  Your single biggest risk is that your risk analysis mechanism doesn’t work…and you wouldn’t even know it!

 There is a better way.