Essentially, what this means is that you have two different ways of innovating within a system. Change the full system (car) or change a single module of the system (watch). Reducing the drag on a car could require a full system overall, because you'll be changing the size of the front end, which could impact the maximum size of the engine (or shape of the engine), or could impact the maximum headroom of the vehicle. So, you could have a radically different looking vehicle from model to model. In fact we can see this if we look at the evolution of the car (below). This change is extremely expensive and requires a huge amount of work. It's not likely that a company would pursue multiple designs beyond the drawing board or initial mockups. It would simply be too expensive to build multiple prototypes that are fully functional.
|Evolution of Lamborghini|
With watches you could have the exact same watch with several different materials to ensure the watch keeps proper time. In terms of watches there have been several radical innovations, including the wristband and digital. However, if the watch is not digital, the changes in some parts of the watch are extremely easy to test and compare on the market. For instance many pocket watches use rubies to protect the metal pieces in a watch from rubbing against each other. In this case it's possible to test many different gems to protect the components, it's also extremely cheap and if something fails completely it would never move into production. However, you could test hundreds of types of gems (sizes or whatever), at a significantly lower cost than testing many different full system designs.
So what's the difference between the two? In this case we're changing a full system compared to a module within the full system. Of course changing the gear structure of a watch would require a full redesign, but there are many parts that can be changed independently. In many aspects this can happen with a car, but there are limitations as well.
This modularity allows designers to innovate on separate aspects of the product without decreasing the quality of the overall system. This same idea can be applied in other business settings in terms of rapid and continual improvement processes. Many business processes are systems that integrate many different groups and aspects. Splitting the system into modular components allows continual improvement on many different aspects of the system at the same time. This modularity decreases the cost of improving individual aspects of the system as well as allows for more improvement projects throughout the system.
Why would the costs be lower? Well, as I mentioned with the watch, it's cheaper to test different components for the gems, time keeping crystal and face glass than to test a change in drag for a car. The change in drag could require changes to the seat heights, new design for the windshields, possibly an entirely new chassis. In the case of reduced drag, if the design works you may have to redesign all these other components. In the case of the watch finding out that the new glass face doesn't work wouldn't impact which crystal works best. This reduces the costs for testing the improved system.