The following is the introductory article to the 1999 Management Roundtable publication, “Product Development Metrics Handbook: What every manager needs to know about measuring product development.”
Sure, it is a bit past its freshness date, but it all still mostly holds true.
Truth In Numbers or Mathemagic?
What do metrics mean to business? Since the time that W.E. Deming introduced statistical process control to the Japanese, manufacturing metrics have led the world down a quality revolution that, while significantly matured in the last half of the 20th century, remains fundamentally incomplete. However, as programs such as TQM and Six Sigma succeed on the shop floor, similar improvement methods and philosophies have yielded comparably little progress in the product development systems that feed the production schedules of manufacturing. Many have made gains in cycle time, product cost and other elements, but few feel they have achieved the desirable level of process control.
What is the difference that is keeping engineering, marketing and affiliated functions from achieving the quantum improvements enjoyed by their manufacturing brethren? One has only to look at the metrics.
The critical component of successful quality in manufacturing is that its events and phenomena consist of controllable, finite variables. The output yield of capital equipment, such as tool and die machinery, can be optimized to a predictable level with extreme accuracy, but the output of a product development team by contrast seems intangible, unpredictable, and uncontrollable. After all, product development activities are largely non-physical, unrepeatable, and exponentially more variable. Most say this is because of product development being a “creative process,” while manufacturing is merely its “physical execution.”
Think about it. There is a giant difference between a) “how long will it take to make 100 widgets?”; and b) “how long until you finish the drawing for the control interface?” In the first instance, the parameters are relatively finite, in the second, nearly infinite. The widget producer knows his machine makes between 48-52 widgets every hour, plus he knows that this is dependent on limited variables such as inventory, machine setup time and a few other things that are predictable and controllable. By contrast, the engineer may provide an estimate for completion time, but his accuracy is much worse, plus there are many more variables that can affect him. He need only be distracted by one phone call, say, from the person who took over his last project, and also to discover that a part he specified was no longer available, to be delayed three days as a result, potentially costing the project several thousand dollars in the long run that nobody will notice until it is too late, if at all.
If you take an honest look at the way companies are structured, there is also a very critical cultural component to this problem. If you compare the demographics of shop floor staff with engineering staff, you will quickly notice a significant socio-economic gap between the two. Engineers are some of the most highly educated employees of any company, and often in highly specialized disciplines from prestigious universities. Such people are not typically open minded about corporate initiatives in general, and have even less affection for attempts to have their work “controlled” by things such as metrics.
On the contrary, manufacturing personnel live in an daily environment of control, with defined reward systems for compliance. Good shop floors maintain tight work discipline for things like cleanliness, preparation, and safety. These things are nearly non-existent in engineering and marketing departments, making this side of the house much less willing to accept changes that incorporate increased accountability. Metrics are often used to expose problems that many don’t want credit for. To ignore these critical cultural issues is an immediate death sentence for any improvement effort.
So why then pursue such a challenge? The downstream effects of this lack of control in product development mirrors what happens with poor production methods: excessive rework, abundant waste and scrap, and embarrassing and costly late deliveries to customers. Once companies initiate effective quality programs on the shop floor and get a taste of the good life of statistically controlled, predictable and accurate manufacturing, it is not at all a surprise that their efforts would then turn to the rest of their business. What is most desired is a process that makes your entire business predictable, a “stretch goal” that is perhaps unachievable.
Why shouldn’t companies simply be satisfied with efficient manufacturing? It’s simply the economics. After all, it is widely argued that 80% of a company’s profits can be directly attributed to 20% of the organization — the product development function — which includes engineering, R&D, marketing and all of the pieces that compose this cross-functional group. In theory, to press on this leverage point and streamline development would complete the picture and enable one to take full advantage of previous gains in manufacturing efficiency. The result would hopefully be a money making machine.
Additional excerpts of the Metrics Handbook will be published in the near future, keep watching this space or subscribe to this blog for updates.