“And then a miracle occurs…”

A famous cartoon by Sydney Harris has two men standing in front of a chalkboard with equations on it. In the middle of the equations are the words “And then a miracle occurs,” with one of the men telling the other (in the caption) “I think you should be more explicit here in step two.” Of course, being the geek that I am, I thought this hilarious. It also came to mind repeatedly during a real-life experience, which I will share.
At a conference on the topic of machine learning and predictive analytics I attended, there were some paper presentations worth the price of admission, but others seemed oriented towards a specific product. Intrigued, I visited the exhibitors’ hall, and interacted with many of the vendors who had set up booths. Without exception, they boasted of their abilities to pull data from multiple sources and multiple formats, and also showed a variety of histograms, pie charts, area diagrams, etc., etc., as their available outputs (unfortunately, none of them demonstrated the capacity to deliver information via Chernoff Faces, but that’s another day’s blog). I approached several of them, and politely engaged them thus.
“It’s my opinion that all Management Information Systems have the same basic architecture, in three sequential steps. Step One: data is collected based on a certain discipline, or binning structure. Step Two: this data is processed via some sort of methodology into information. Step Three: this information is delivered to decision-makers in such a way that they can use it to make, well, decisions. You have demonstrated your product’s ability to collect data from different sources, and your booth is festooned with samples of its output. Here’s my question: what methodology do you use to convert the data into usable information?”
I was somewhat disappointed to receive the same basic answer (or a derivative) from every single vendor: It’s whatever the customer wants to use. One vendor in particular started a discussion so overwrought with jargon that I had to interrupt, and ask point-blank “Your marketing material claims the ability to produce ‘predictive analytics.’ Suppose I came to you representing a mid-sized company, and I was ready to purchase this product. Exactly what would you be delivering?” The answer, padded as it was with the aforementioned excessive jargon, basically came down to “it depends.” At least they didn’t go straight to “And then a miracle occurs.”
 I thought then, and believe now, that this was something of a disingenuous answer, in that on the one hand, these vendors were selling a capability in the vein of predictive analytics, and on the other were simply deferring to “customer preferences” when it came to how such an output was to be achieved. I could have told them then and there, the customer wants you to deliver an information stream that could be legitimately considered to be in the realm of predictive analytics, with the data they currently have available. If these potential customers knew how to do that at the present time, they really wouldn’t have a need for your product, now would they?
But besides the pursuit of a “predictive analytics” package that could deliver the relevant, accurate, and timely information stream needed to obviate most (or even all) of the decisions for whatever level of management at which it’s aimed, we have the layered dynamic of macro-organizational decision oversight versus the latitude of movement that managers would require in order to bring their projects in on-time, on-budget. Put another way, does the owning organization reward managers who make all of their decisions consistent with policy, procedure, and even the unstated aspects of organizational culture, with mixed ultimate results; or, does it value those managers who bring their projects in on-time, on-budget, but bend an occasional unstated/undocumented rule from time to time? I would maintain that this is not a trivial dichotomy. For another mental exercise, imagine that one of these vendors did have the methodology to convert available data into a reliably predictive analytical information stream, and were simply reluctant to disclose it to some wise-guy inquisitive booth-visitor. And suppose that, once this system came on-line to its buyer’s home organization, that it informed said manager that the best course of action was contrary to previously-communicated organizational goals or agendas. What happens then?
It's long been my considered opinion that organizational culture is downstream from individual Project Teams’ success. Are you looking for a significant organizational culture change? Don’t waste your time kvetching to the staff about how they should behave better. Instead, get the Projects in the portfolio to come in on-time, on-budget a majority of the time, using a specific set of PM strategies.
That’s when the real culture-change miracle occurs.