David Weinberger
KMWorld Archive
This column is part of an archive of David Weinberger's columns for KMWorld. Used with permission. Thanks, KMWorld!


Link to Original at KMWorld  Index

David's home page | Bio | Speaking | Everyday Chaos

The MVP process then and now

29 September 2014

The term “minimum viable product” was coined by Frank Robinson in 2001 to describe a process that reverses the usual order of “design, build, sell.” By putting the earliest usable version of a product into the hands of a relatively small group of early adopters, the company can see what features are truly desired by its customers before it has invested heavily in features that its customers don’t actually want.

It makes sense. So why is the MVP process only now becoming “hot”? Why not, say, a hundred years ago?

In his book Any Colour - So Long as It’s Black, John Duncan describes the process Henry Ford went through in 1906 when designing the Model T. The Ford Model N was already out and doing pretty well. But Ford saw a clear field for a new design because the industry was young enough that “there was no consensus on a standard automobile” (Kindle page 38). Ford began with a few design criteria based on the realities of the market. Driving was so novel that most drivers were bad at it, so the new car should be easy to drive, especially when it came to shifting. Because paved roads were infrequent, it had to ride high enough for all sorts of conditions. It should be lightweight but strong. And cheap enough to sell to the masses.

Innovative assemblages

So Ford gathered a handful of his best engineers and for months met with them in a room 12 feet by 25 feet equipped with blackboards and a few tools. Because Ford preferred to see objects rather than designs, the engineers brought in models for the pieces they were proposing. Assemblage by assemblage, they came up with innovative solutions—a firmer way to connect the axle and frame, the use of new steel alloys, separating the cylinder head from the cylinder block, casting the cylinders as part of the crankcase, a two-speed gearbox, and a transmission housing and oil pan stamped out of a single piece of metal, about which Duncan says: “No man will ever design a structure as wonderful as our skeleton, but among manmade artefacts, this stamping has to rank highly.”

In fact, sheet metal stamping became such an important part of the Model T that Ford bought the company that did the initial pressings even though it was far more expensive to set up the castings than the process used by car manufacturers up to that point. “Mr. Ford was one of the first to see that even if a die cost $10,000, it was cheap if it made a million parts.”

In fact, manufacturing costs guided the design process all the way through. The gas tank was positioned high under the front seat so gravity would do the work of a fuel pump. The frame was designed to reduce the number of bolts that would have to be screwed in place.

Rapid repetition

So, why didn’t Henry Ford, a genius of innovation, come up with the concept of the minimum viable product?

In one sense, he did. The Model T was designed as a minimum viable vehicle. It wasn’t particularly fast or comfortable, and it sure wasn’t as pretty as some of its competitors. But it achieved Ford’s minimal design goals.

But the aim of a modern MVP is not just to be minimal. It’s to enable rapid iteration. But rapid iteration is hard when you’re dealing with atoms, not bits. And these atoms required that metal be cast and entire factories tooled. There was a high price for not getting things right on the first try.

And if you did, you could become very reluctant to change. In his book, Duncan writes:

“It is reported that in the early 1920s when many felt that the Model T should be drastically redesigned, a group of men built a prototype that had all the improvements they wanted; his response when they proudly showed it to him was to pick up a sledgehammer, smash the car to pieces and walk away without saying a thing.”

The MVP process strikes us as attractive not only because bits make it feasible, but also because we’ve come to believe that a technology that isn’t changing every six months is failing. Yet, in the almost 20 years it took Ford to introduce a new model, 15 million Model T’s had been sold. And during that entire stretch, never once did Henry Ford put on a black turtleneck and tease an audience with what would be new next month.