27 January 2012

All Models are Wrong (Part 1)

You've heard the phrase "All models are wrong: Some are useful," by George Box, a chemist who taught himself to be a statistician.  My takeaway from this is that all models - and that is everything written down in all management texts and training manuals everywhere - is that the model is a lens to look at a problem through.

When I was an undergraduate at university studying marketing we were discussing the Marketing Mix.  "When launching or developing a product..." the lecturer said, "pay attention to Price, Promotion, Product and Place..."

I sat there thinking.  I didn't have corporate experience and this didn't make sense.  It just seemed to simple and to pat. (Four Ps!)  I was trying to learn this stuff by applying it to the bar and restaurant jobs I had.

How can I apply the four Ps to the restaurant?  We can't move it, especially me as a waiter, so the distribution aspect is a moot point.  Maybe we can change the prices and the menu (product) and sure we could do something to raise the bar when it came to promotion.  But what about the other aspects - the relationships the staff had with each other, the way we interacted with the customers, the suppliers (especially the alcohol sales reps!)

So I said as much to my lecturer; "This doesn't seem to be enough.  Surely you need to think through more than this to launch and manage products."

It was then I got one of the best nuggets of information from that degree.

"It's just a checklist to help you along.  It doesn't give you all the answers.  You have to use your brain to solve complex problems."

I wonder how he remembers the conversation, if at all.


  1. I'd be a fool to disagree. Of course models are wrong. Heck, we cannot use 1 language to accurately communicate between 2 people! There is a chance we can document a static model of what goes on in a business and hope it's correct. Certainly not past the end of this week. :-)

    But I love models because we are so much richer for them. To not have models is to miss so very much information. We (most of us) cannot keep enough in our heads to understand the basics without models.

    So, if you had to choose a model you like more, which would it be?

  2. I use them all. The more the merrier!

    I am often surprised at how some of the most basic and simplistic work well for me. the SWOT is a good example. So is the Urgent/Important grid.