Now, who should you test your prototype with? As a rule of thumb, try to cover a cross-section of the target population to increase the validity and generalizability of your collected data, but also include potential extreme users (which you might have initially met when reading about ‘customer discovery’). An extreme user will likely not be a person from your main target group. However, they often provide more detailed feedback and additional inspiration for further iterations of your prototypes. This sampling approach is accurately dubbed “extremes and mainstreams” in the IDEO.org Design Kit.
A word of caution: be aware of your own sampling biases. For example, if you are a tech-savvy engineer and are building a highly complex and expensive remote controller for speed racing drones which you have to program yourself, many of your engineering buddies might think that this is an awesome product idea. You might even get them to pre-order the controller during your conversation. However, the reaction from the average drone user, who is mainly interested in taking a few nice photos from a different angle, would probably be very different.