Logic can be a double-edged sword, especially when it comes to predicting customer behaviour. Let's consider the case of Gary, who's been in his line of work for several years. He's confident in his understanding of his product and customers, and this confidence has led him to discontinue user testing. On the surface, this decision appears logical – he knows his product and understands his customers, so why expend unnecessary effort on testing and discovery?
However, this approach overlooks that there are typically three customer types. Firstly, there are those who love the product and make their satisfaction known. Then there are those who dislike the product and express their dissatisfaction. Lastly, and most significantly, the silent majority neither voices their approval nor disapproval. While Gary might engage with the first two groups, he's potentially missing out on insights from this vast majority who come and go quietly.
It reminded me of a situation in San Francisco's Noe Valley neighbourhood when I was living there. The bakery there was a popular hub, serving throngs of customers every morning, seemingly attuned to their preferences. One day, however, half of their customers vanished without explanation. Product quality at the bakery and service levels hadn't changed – if anything, they'd improved. The bakery owner sought answers from his regulars, those who loved or disliked the establishment but found no satisfying answers. They still liked or disliked the bakery for the same reasons.
A month later, the owner stumbled upon an unexpected reason. A new coffee shop had opened near a bus stop, just a few blocks from his bakery. With lines stretching out the door, he noticed many of his former customers. Although his bakery offered superior products, the new coffee shop's location, directly in front of the bus stop, made it more convenient for downtown commuters to grab their morning coffee before boarding the bus. Another mom-and-pop-sized San Francisco coffee shop had passed entirely under the bakery's radar. However, it had nothing to do with the quality of the product or the knowledge the bakery had about their customers. Something, very fundamental, had changed.
The lessons here are clear. First, understanding your customers does not guarantee prediction accuracy unless all other variables remain constant. Second, a superior product can sometimes be overshadowed by environmental changes or new offerings that may not be as good but are more convenient or practical.