Everyone knows to talk to customers before building. Far fewer people run those conversations in a way that produces usable information, and the failure is systematic rather than random.
The problem is that you arrive wanting a particular answer, the person you are talking to can tell, and most people are polite. What you collect is not evidence. It is a reflection of your own hypothesis, returned to you warmly.
The question that ruins it
Contaminated question
Answerable question
Would you use something that did this
When did you last deal with this
Does this sound useful
Walk me through what you actually did
Would you pay for it
What did it cost you the time it went wrong
Here is what I am building
I am looking at how people handle this
Would you use something that did X. Or its close relatives: does this sound useful, would you pay for this, is this a problem for you.
Every one of those invites a prediction about hypothetical future behaviour, which humans are poor at and which costs nothing to answer generously. You will get yes a great deal, and the yes carries no information, because the person saying it has not been asked to give anything up.
The giveaway is that nobody ever says their idea was invalidated by these conversations. If a method never returns a negative, it is not a method.
Ask about the past instead
“Nobody ever says their idea was invalidated by these conversations. If a method never returns a negative, it is not a method.”
Replace every hypothetical with a historical question. Not would you, but when did you last.
When did you last deal with this. Walk me through what you actually did. What did you use. Who else was involved. How long did it take. What did you do with the output. What happened the time it went wrong.
These are answerable from memory, they cannot be answered generously because they are factual, and they produce the specific detail you need: the workaround they built, the spreadsheet they maintain, the person they call. A workaround is the strongest possible signal, because somebody has already paid a cost to solve this without you.
Nobody ever says their idea was invalidated by these conversations.
If nothing comes back, that is your negative result, and it is the thing the method exists to produce. A problem nobody has worked around is a problem nobody has.
Turn what you know into what you own.
Vibepreneur builds structured ventures from professional expertise, with positioning, launch assets, and growth systems included.
Join the WaitlistDo not describe your idea
This is the hardest part, and the most important. Say nothing about what you are building until the conversation is over.
The moment you describe it, two things happen. The person starts helping you, which means they begin generating reasons it could work, and they stop telling you about their actual situation in favour of reacting to yours. You have converted an interview into a pitch and you cannot convert it back.
If asked directly what you are working on, something vague and true is enough. You are looking at how people handle this. That is accurate and it keeps the conversation where it needs to be.
What a real signal looks like
Three things, in increasing order of strength. They describe a workaround in detail without prompting. They offer to introduce you to somebody else with the same problem. They ask when they can have it.
The last one is the only true positive, and when it happens you should take it literally rather than as encouragement. Ask for something: a deposit, a pilot date, a commitment to be the first user with a named start. Enthusiasm that survives a request for commitment is data. Enthusiasm that does not is politeness, which is pleasant and worth nothing.
See zero dollar validation experiment templates for the structures, kill criteria before you start for deciding in advance what a negative looks like, and the first ten customers for what comes after. The waitlist runs the structured version.