All of my portfolio companies are using some form of net promoter score. And the centrality of the metric, especially at startup board level, has bugged me for a while. But I couldn't quite put my finger on why. So I was thankful that late last year, Jared Spool tackled the fairly obvious deficiencies of NPS that I had only intuited.
Especially in the early stages of a startup, customer satisfaction and loyalty is an incredibly qualitative endeavour. It's another example of rampant reductionist materialism that we think we can condense this complexity into one metric.
A startup is an organization in search of a scalable, repeatable business model. It's like an engine starting. In a modern engine, the ECU or ECM is where all of the sensor data comes together and actuators are controlled. Which is fine once the engine turns over, but dramatically less useful while building one.
Ken Wilber, the founder of integral theory, calls this flatland - following the dissociation of science, art and morals (or nature, self, culture), science is morphing into scientism. Beware of data reductionism in building companies. Sometimes focusing on the good and the beautiful will turn out truth on its own.