The common advice to conduct unbiased discovery interviews sounds logical but often fails. The truest way to validate an idea and understand customer needs is through the act of selling. This forces a concrete value exchange and reveals genuine demand in a way that hypothetical conversations cannot.

Structure discovery into two distinct conversations for maximum effect. The first meeting should focus exclusively on uncovering the customer's blocked goals (demand), without mentioning a product. Use the second meeting to validate if a high-level solution sketch (supply) gets ripped out of your hands.

Standard discovery questions about 'pain points' are too broad. Instead, focus on concrete 'projects on their to-do list.' This reveals their immediate priorities, existing attempts, and the specific 'pull' that will drive a purchase, allowing you to align your solution perfectly.

A potential customer can logically agree with your framing of their problem yet have no intent to buy. True demand isn't intellectual agreement; it's a palpable force. You must sense the pressure of them actively pushing against a wall. A customer leaning back and nodding is a red flag.

Treat customer conversations like coded messages. Create a "translation guide" by categorizing every statement into three buckets: their core goal (demand), the tools they use (supply), and random noise (irrelevant). This structure reveals what they'll actually pay for.

To test for genuine 'pull,' a founder should challenge a prospect's need rather than sell to it. By asking questions like, 'Why can't you just do this with your existing options?' you force the prospect to articulate why their problem is urgent and unsolvable. This 'leaning back' approach makes high-intent buyers sell you on their problem.

Most problems customers describe are "pain points" they won't act on. You can't distinguish these from real, actionable demand ("pull") through interviews alone. The only true test is presenting a viable solution and attempting to sell it. Their reaction—whether they try to pull it from you—is the only reliable signal.

The core of a Jobs-to-be-Done (JTBD) interview is to uncover the full story of what causes a customer to switch from one product to another. This line of inquiry reveals their true motivations and the underlying 'struggle' that triggers their search for a new solution.

When customers talk, trust their articulation of what they're trying to accomplish (demand) and why their current tools fail (supply problems). However, completely disregard their suggestions for what product or feature you should build (supply they want). That is your job to design, not theirs.