I recently received an email inviting me to fill out a customer satisfaction survey from one of my favorite online outdoor equipment retailers. I freely admit that I only opened it out of competitive curiosity. When I opened the email, and it promised to take only two minutes and ask two simple questions, I exclaimed, “An NPS survey! ROCK ON!”
But that was me speaking as the survey taker. As a data scientist and survey analyst, I recognized the shortcomings of an NPS-style survey. There’s no way to control the data points in the sample, so this equipment retailer wouldn’t be getting the data required to slice and dice my results in interesting and meaningful ways.