Ford CEO Alan Mulally’s “Most Important Thing”

Interesting interview of Alan Mullaly here and also here.  Mulally is Ford’s CEO, and ran Boeing’s commercial airplane business until 2006 when he replaced Bill Ford in that role (Bill Ford is still Chairman).  In my mind, hiring Mulally may have been Bill Ford’s best and most important decision.  Both articles are worth a few minutes, but one thing that really stood out to me is this proclamation from Mulally:

“The most important thing that you do is size your production to real demand and not overproduce.”

In some respects this sounds like “Business 101”: supply and demand have a relationship, and over the long term it’s generally more profitable for a producer to adjust supply than inflate demand with price cuts.  BTW, This is what people in lean thinking circles mean when they say let demand pull production volume levels for goods and services. 

And yet for so many years this simply did not happen in the American auto industry.  The reason, according to Mulally, is this: 

“In the auto industry, they would keep the production up because they thought every cost was fixed.”

In particular, labor was treated as a fixed cost, thanks to the so-called Jobs Bank program that paid UAW workers 90% of their regular salary whether they were working or not.  This program was designed to provide companies with a financial incentive to focus on growing demand versus laying off workers.  By structuring the program to work in harmony with the way production volume decisions were made inside the auto companies, the program created the *appearance* that it was more profitable to produce and sell additional vehicles at lower prices than reduce production levels (and jobs) to just meet the real demand levels.   Keep in mind, the people making these decisions were some of the smartest and most dedicated people working anywhere. 

I personally believe the problem was that over-producing *appeared* to be a profitable strategy because the decision-making system was set up to focus on only a part of a picture, when people should have been seeing a series of bigger pictures over time (i.e., a “movie”, to push the analogy a little farther).  In other words, the metrics guiding the decisions did not cover enough of the total business, and did not provide sufficient insights over a sufficient time horizon.  Of course, everyone is different and some individuals surely saw things more broadly and over a longer time-horizon than others, but, overall, the normal decision processes continued to encourage over-production. 

Although Mulally neatly articulates the importance of producing to meet demand, this was not a brand new idea when he arrived at Ford.  While I was at Ford, I recall this mindset was changing even before Mulally took the reins in 2006.  I’ve been told that in the early to mid-1990’s (before I joined Ford), Eli Goldratt, author of the iconic book The Goal about the theory of constraints, was working with Ford to help them adjust the way they thought about the flow of their product development and manufacturing processes.  The Ford Production System was modeled in some ways after the lean Toyota Production System (itself based in part on Henry Ford’s own system of production 80+ years ago).   I don’t know, but I suspect Jim Womack of the Lean Enterprise Institute did his share of consulting with Ford.  And I do recall that process improvement initiatives were seemingly continuous (not exactly the same as “continuous improvement”, but i digress… ;) ).  I suspect most if not all of these drives to improve efficiency and quality helped make important improvements. 

Even so, my perception is that these improvements tended to help Ford run it’s sub-optimal system more efficiently, but did not really change the decision structures that created the sub-optimal system of over-production in the first place – and therefore allow for a more optimal system to replace it.  In my opinion it was actually the emergence of tools to understand and describe complex adaptive systems that made this possible. 

Tools like Systems Thinking and Systems Dynamics gave people a rigorous way to learn about the relationships between, for one of many possible examples, fleet residual prices and new vehicle pricing power at one point in time, and production volume and price discounting strategy three years earlier.  And just as importantly, to communicate those ideas in ways that let people have reasoned discussions about the assumptions of the “old” model of over-production and the fit of a new alternative model that was more encompassing.  This allowed for things like losses in future years due to low fleet residuals to be factored into current year production planning, and thus show a truer (if not perfect) picture of the higher costs over time that existed in the old over-production model. 

By the way, my own thumbnail sketch version of the story of how fleet residual pricing, new retail pricing, previous price discounting strategy and production volumes relate goes like this: 

Over-production leads to price discounting, and price discounting leads to more vehicles getting sold, and more vehicles getting sold leads to more used vehicles in the marketplace 2-3 years later, which itself lowers used vehicle prices, which then contributes to less pricing power for new retail vehicles and also lowers the residual value of vehicles coming off lease, which in turn raises the payment price for new leases and lowers demand, which then leads to more over production and price discounting… a spiraling, self-reinforcing vicious cycle. Worse, this whole little vicious cycle contributes to thinner (or non-existent) profit margins, which *can* lead to incremental cost cutting and delays in new product development, which leads to less competitive portfolio of vehicles to sell, which leads to even lower demand, which leads to more over production… ugh! ANOTHER vicious cycle!

Finally, if you’ve read this far you might be thinking, this is sort of interesting, but how does it apply to IT or software development?  This blog is on MSDN, after all!  :)  The operating model that Mulally has helped establish in the mainstream at Ford are based in large part on the principles of Lean Thinking and Systems Thinking.  These principles are equally relevant to IT and software development as they are to manufacturing! 

For more info, check out:

• Mary and Tom Poppendieck’s site for really, really good info on Lean Software Development. 
• Mary and Tom also have some highly recommended books on lean thinking and systems thinking in software development, including this one that just came out in October 2009: Leading Lean Software Development.   
• Alan Shalloway’s NetObjective’s site for another voice on lean software development.
• The book, The Fifth Discipline by Peter Senge – this book really helped launch systems thinking as a discipline.  NOTE: The specific version I’ve linked here is the March 2006 edition, which contains significant new material, including a whole section on how former Ford CIO, Marv Adams, used Systems Thinking within the IT organization at Ford. 
• Society of Organizational Learning (SoL)

Happy reading! 

Technorati Tags: Ford,Mulally,Systems Thinking,Lean Thinking