How to Prompt Engineer A VC
Everybody's figuring out how to prompt an AI right now. Be specific, give it context, tell it exactly what you want and what a good answer looks like, and it gets you there faster.
Founders should be doing the same thing to VCs.
Here's what I mean: The pitches that actually move me aren't the ones where someone walks me through their TAM or tells me how much their customers love them. They're the ones where the founder hands me the test before I've even asked for it.
"Call this hospital's head of supply chain and ask how long it takes to onboard a new vendor. She'll tell you nine months. Then ask what she'd pay to cut that in half."
"Ask any board member in a building less than 100 units whether they feel like they’re getting their money’s worth out of Buildinglink."
That's prompting your VC. You're not asking me to take your word for anything. You're telling me exactly who to call and exactly what they're going to say.
When I make that call and the answer comes back the way you said it would, the whole dynamic shifts. You've just done my diligence for me and handed me the result with the receipt attached. You know your customer well enough to predict what they'll say to a stranger who calls them cold. That's not a pitch skill. That's evidence you've actually been in the field doing the work, which is the thing I'm really trying to figure out in the first place.
Think about what a normal pitch actually is. It's a contest of opinions. You think the market's ready, I'm not sure it is. You think people will pay for this, I've seen ten companies die believing exactly that. So we go back and forth, and the meeting turns into you trying to convince me your read on the world is right and mine is wrong.
Nobody wins that one in a conference room.
The minute you can say "Don't believe me? Here's who to ask," you've taken your own opinion off the table and mine along with it. Now we're both just checking your prediction against what the customer actually says. You made yourself easy to verify, and easy to verify is about the most credible thing a founder can be.
VCs don’t like doing a lot of work when the payoff is unclear, so you have to ask yourself how you can make my work easier.
This only works if the story holds up, obviously. Send me to a customer who gives me a shruggy "yeah, it's been fine" and you've done worse than not sending me at all. Now, I know you either don't really understand your customers or you were hoping I wouldn't bother to check. So, don't point me at answers you haven't already heard with your own ears. The move only works because you did the listening first. You've had the conversation enough times that you can call the response before it happens, instead of crossing your fingers that it goes your way.
Which is the actual point. Prompting your VC isn't a trick for steering diligence. It's just what naturally comes out of a founder who's done the work. The ones who know their market cold don't need me to trust them.
They point me at the truth and get out of the way while I find it.