Forecasting the flip

Another way to assess the likelihood of scaling an innovation eluded my searching yesterday. This morning I found it in Chapter Four of the book: Disrupting Class - How disruptive innovation will change the way the world learns by Clayton Christensen, Michael  B. Horn and Curtis W. Johnson. In the words of the authors:

It turns out there is a way to forecast the flip. ... one must plot on the vertical axis the ratio of market shares held by the new, divided by the old (if each has 50 percent, this ratio will be 1.0). Second, the vertical axis needs to be arrayed on a logarithmic scale—so that .0001, .001, .01, .1, 1.0, and 10.0 are all equidistant. When plotted in this way, the data always fall on a straight line. If the first four or five points do not lie in a line, it is a signal that there is no compelling driver for substitution. But the line is always straight if a disruption is occurring. Sometimes the line slopes upward steeply, and sometimes it is more gradual. The reason the line is straight is that the mathematics "linearizes" the S‑curve. When the substitution pace is plotted in this way, one can tell what the slope of the line is even when the new approach accounts for only 2 to 3 percent of the total. That makes it easy to extend the line into the future to get a sense of when the innovation will account for 25 percent,50 percent, and 90 percent of the total. We call this line a "substitution curve." (pages 97-98)

The compelling driver for "the innovation replacing the incumbent" is evident in the early market share data. Scaling of the innovation is simply the accumulation of the ongoing logarithmic increases. The S curve flips from "curving up" to "bending over". That turning point has been compared to "taking swings at beach balls where you can't miss", "getting caught up in a tornado" and "hitting the first bowling pin that knocks over all the others".