“Make something people want” is the most important guiding principle in design.
The goal of any product is to delight, or help get something done, or deliver some type of utility to the user.
So how do you figure out what people want?
My entire career from Stanford d.school, to Product Manager at YouTube, to founding and scaling Ketone-IQ, I've been in a position of needing to understand the needs of users.
A common, well-meaning instinct is, “let’s do a survey.”
Unfortunately, many surveys are poorly executed, and even a well-written survey can be counterproductive simply because respondents are not that good at self-reporting what’s going on in their head when they make decisions. Volumes have been written about how stated preference rarely aligns perfectly with revealed preference.
What makes surveys especially insidious is that often times, the fact that “a survey was done” gives a false sense of confidence, something very easy for committees to nod their heads to, and can lead to high confidence bad decisions.
So what’s a builder to do? How do I learn about my customer?
Here are a few strategies I find useful.
The best survey is the credit card survey, aka — did someone take out their credit card and buy your product or not.
https://x.com/visualizevalue/status/2076715189131460802?s=20
A person’s wallet will tell you more than what comes out of their mouth.
If at all possible, try to get your product in front of paying customers.
Prototype whatever it is that you are thinking, and try selling it.
Most surveys are just a proxy for “will this sell?”
The best way to answer that question is to actually sell it. Don’t let surveys be a tool for procrastinating on what you really want to find out.