Ask founders the right way

Most founder advice fails before the advice even starts. A first-time founder sends a vague message to someone 10 years ahead, asks for “15 minutes to pick your brain,” and gets either no reply or a generic answer. Abhilash Chowdhary offered a tighter script: ask people only 2 to 3 years ahead of you, show them your exact stage, do the work they suggest, and come back with results. (x.com) (ycombinator.com) Chowdhary is the co-founder of Crustdata, a Y Combinator-backed company in the Fall 2024 batch, so the advice comes from someone actively building, not from a generic networking playbook. His post framed founder outreach as a repeatable operating system instead of a one-off cold message. (ycombinator.com) (x.com) The first move in his method is choosing the right person. He argues that founders who are 2 to 3 years ahead are close enough to remember your exact problems and far enough ahead to know what actually worked. A seed-stage founder asking a public-company chief executive officer for tactical help is like a middle-school runner asking an Olympic sprinter how to shave 0.2 seconds off a local race. (x.com) That distance matters because startup problems change fast by stage. A founder trying to get the first 10 customers is dealing with different bottlenecks than a founder managing 200 employees, and advice from the wrong stage often turns into abstract slogans instead of usable steps. Chowdhary’s script starts by stating your current stage so the other person can answer the problem you actually have. (x.com) The second move is making the ask concrete. Instead of sending a broad note like “Would love your thoughts on growth,” his format pushes you to describe where you are now, name the specific problem, and ask one focused question. That gives the other founder a narrow target, the same way a good product brief gives an engineer one job instead of five half-formed ideas. (x.com) The third move is the part most people skip: do the work before you ask for more. Chowdhary says to take the advice, apply it, and return with proof of what happened. If someone suggested a pricing test, a new outbound message, or a hiring change, your next note should include the exact experiment and the result, not another open-ended request. (x.com) That follow-up changes the relationship. A founder who sees that you executed their suggestion now has evidence that their time produced an outcome, so the next reply is easier. In practice, you stop looking like a stranger collecting opinions and start looking like an operator building a feedback loop. (x.com) This is why his post centers on proof of work. Proof of work means visible evidence that you tried something in the real world, such as a revised landing page, 20 customer calls, a new pricing page, or conversion numbers from a test. It is the startup version of showing your math instead of saying you “thought about it a lot.” (x.com) The script also fixes a common cold-outreach mistake: asking for trust before earning it. When you lead with action, you reduce the social risk for the other person because they do not need to guess whether you are serious. They can see a timeline, a problem, an experiment, and a result. (x.com) There is a second benefit hiding inside the method. Even if the founder never replies, the structure still forces you to define your stage, isolate one problem, and run one test. That turns outreach from a passive search for wisdom into an active way to sharpen your own thinking. (x.com) The reason this advice is spreading is that it is short enough to use immediately. Chowdhary packaged it as a simple, repeatable outreach script rather than a long theory about mentorship. In a startup world full of broad advice about “networking” and “building relationships,” his version is closer to a sales sequence: right target, clear message, measurable follow-up. (x.com) The practical takeaway is blunt. Do not ask the most famous founder you know for general advice. Ask someone 2 to 3 years ahead, tell them exactly where you are, act on what they say, and come back with receipts. (x.com)