In his groundbreaking book, Crossing the Chasm, Geoffrey Moore describes a market as being composed of sub-markets representing people with different purchasing propensity (different adoption rates). Each of these sub-markets, when represented as a probability distribution describing the likelihood that they will purchase at a given time, are summed together to represent the larger market.
In the image below, I’ve created a probability distribution. We can imagine that 100 people were shown a new product at time 0; the curve describes sales over time. They start slowly, gain speed, and then slow down.
It is hardest to sell to the first and last customer in a market, and, in the figure below, I’ve drawn the cost of a sale as an inverted curve. There is no vertical scale, here. It’s really the shape I want to focus on.
So now we see that sales get easier (cheaper) and then get harder (more expensive) (red curve) as we capture an increasing amount of the market (blue curve).
As before, focus on the shapes, and not specific units of measure. Let’s start with the purple “experience” curve. Note how the experience curve lags behind the red sales cost curve (i.e., it doesn’t start dipping down till after the red curve does). This is the assimilation rate; a company can not assimilate experience faster than it actually experiences an experience. As we gain more experience, then the improvements we glean start to pile up; the slope of the curve increases. Experience curves are the result of both process improvements and economies of scale. So, up to a point, both the sale itself and the servicing of that sale become less costly. I’ve shown the experience curve as being symmetrical, but that isn’t necessarily the case; I’ve just taken artistic license to make it look nice. But let’s switch over to the green curve. The green curve is representing the company’s growth. The left side of the curve sort of resembles a hockey stick; this is what folks usually focus on (and what is usually shown to the VCs). But notice the inflection point right in the middle; growth slows when you’ve reached the midpoint of market penetration. But there’s more…
The graph above plots the rate of change in sales. Note how the rate peaks at our 38th customer, i.e., when we’ve penetrated 38% of our market. Sales are still growing, but at an ever decreasing rate. At 50%, the rate of growth inflects and actually starts to slow (rate of growth turns negative). Now, it is a fair point to say that, had I changed the characteristics of the curve, such as making it more peaked in the middle, then this change might not always happen at 38%. That is true, but it will always happen BEFORE 50%. After we’ve reached the inflection point, sales start to slow; just a bit at first, but it rapidly picks up steam. If you look at the blue curve in figure 4, you’ll see this phenomenon right in the middle, where the blue line turns sharply and then descends at a fairly consistent rate until a similarly sharp curve reverses the trend.
So we now have slowing sales, increasing sales costs, and decreasing efficiencies elsewhere all coming into play at the same time, and by the time it’s recognized it’s too late for a company to handle gracefully. This is because it starts gradually, and isn’t recognized for what it is; a turning point in the company’s relationship with this market. If the normal set of temporary, short-term fixes are tried, various marketing programs, sales initiatives and such, then management is distracted until the evidence becomes irrefutable, but by then it’s too late for anything resembling an orderly retrenchment. The experience curve can actually contribute to the problem, because it could tend to paper over the sales problem if things aren’t looked at carefully enough. In addition, after a point, the experience curve reinforces the company’s ties to a particular market, drawing its focus more into that market when it should be doing the opposite.
Investment in a market, such as product refinement, capital upgrades, etc., need to be driven by an expected rate of return. But, as illustrated above, this rate is not static, it depends on where along the path of market saturation we have reached. Investing too long, too late in a market that has passed the inflection point can have dramatically negative consequences when sales behave as shown in this post. Realizing that markets don’t look exactly like a normal curve, and that data is going to be quite noisy, understanding where you are in a market could prove a useful tool in making investment decisions.