An 86-minute guide to finding PMF | Rob Snyder (Author, The Power of Pull)

Rob Snyder introduces “pull” as the condition in which a person would find it weird not to buy a product. He contrasts this with the usual startup advice about finding “hair-on-fire” pain, offering differentiated value, or working with design partners. Snyder had previously checked all of those boxes, yet spent two years pushing a product that nobody bought. He eventually realized that pain and value are not enough: customers must have a priority they are trying to complete now and cannot complete with their existing tools, methods, or options. PULL is an acronym. P stands for a Project on the person’s to-do list. U means the project is Unavoidable now—there is a reason it cannot be postponed. The first L is the List of options the person has considered, including doing the work internally, hiring someone, or using an existing product. The final L is the Limitation that makes those options inadequate. A prospect with all four conditions is like a “coiled spring”: the founder does not need to persuade them that the problem matters; the prospect is already trying to get unstuck. Snyder stresses that pull belongs to a person, not a company. A startup is not selling to “SpaceX”; it is selling to the head of legal at SpaceX who has a specific project she cannot complete herself.

Snyder uses Jump, an AI note-taking product for financial advisors, to make the framework concrete. The advisors’ Project is minimizing the time they spend taking notes during client meetings and entering those notes into their CRMs. The work is Unavoidable because failing to record and store the notes could lead to regulatory trouble or even jail. Their List of options included generic AI note-takers and other ways of recording meetings. Those options had Limitations: they did not integrate with the advisors’ CRMs, and it was unclear whether they complied with SEC regulations. When Jump offered a product that addressed those specific limitations, prospects pulled it out of the founders’ hands. One person reportedly said, “This must be how Alexander Graham Bell felt when he invented the telephone.” Snyder contrasts this with insurance agents, who also meet with clients and might want to reduce note-taking. But insurance agents are not required to put those notes into their CRM, so the project is not truly Unavoidable. Their existing behavior—simply not taking the notes—can be good enough. That group may express interest, but it lacks the force that produces an immediate purchase. The framework therefore requires all four elements at once. If people have a project but their existing options work, they may complain or praise the product without buying. If the work is not urgent, they will defer it.

Snyder argues that founders should use PULL upstream of the product idea. Instead of brainstorming products, they should state a hypothesis about a specific person who is stuck and would be weird not to buy. To find that person and problem, he recommends taking customer discovery unusually seriously: if your life depended on finding a real problem, you would not conduct a few casual Zoom interviews. You would go in person, shadow people, perform the work yourself, and observe the moment when someone is blocked. He describes a friend who wanted to serve small businesses but had no product idea. The friend listed industries, shadowed workers, and rode along with solar-panel installers, whom he ultimately disliked. He then spent days with independent physical therapists, found that he and they worked well together, and even considered buying a physical-therapy franchise so he could understand its economics. Through that immersion, he noticed a specific project the therapists were trying to complete and where existing options were too expensive or slow. Snyder recommends documenting the hypothesis on a slide: “There’s a project on your to-do list” that is happening because of a particular reason, and the options already considered are inadequate for a stated reason. A second slide explains the proposed solution. The founder shows the first slide and asks, “Do I have that roughly right?” Then they show the second and ask whether the proposed help fits. The first customer should say, in effect, “Oh my God, yes,” and attempt to buy; the founder should repeat the test with several other people.

Snyder says the strongest validation is not enthusiasm but a “pull-shaped action.” A prospect with pull says, “How do we get started?” or, before a product exists, “If you can do that, I will pay you right now.” They lean in, ask how the product works, and try to move forward immediately. By contrast, “That’s amazing,” “Tell me more,” “Can you send information for our team?” or “We could use this in Q4” are signs of the prospect pushing the decision into the future. When a conversation is lukewarm, Snyder recommends the Hail Mary question: “Is there anything on your to-do list that you would pay me to do right now?” The question reframes sales as looking for a problem the prospect already needs help with, rather than convincing them to want the founder’s product. He gives the example of a startup selling an applicant-tracking system to governments. A prospect was mildly interested, but the Hail Mary revealed an urgent need around job descriptions. That became the startup’s first contract and eventually its growth path. He allows “solve your own problem” as a valid route, but only when the founder honestly would have paid someone to solve the problem. It is not enough to think a product should exist. He says the distinction applies to consumer products as well: Spanx is an example of a founder discovering and solving her own immediate problem through direct observation of how people reacted.

To find initial conversations, Snyder recommends short, interesting outreach rather than a conventional pain-point pitch. Founders should not ask whether someone has a list of problems and then present a “game-changing solution.” They should give the recipient a reason to talk. One example was, “Snoop Dogg invested in us. I thought you’d find it interesting too. Would you be open to chat?” Other examples mention dropping out of a PhD program or having worked at a family plumbing company. The message need not contain a value proposition; its job is simply to earn a conversation. Snyder starts on LinkedIn, where a prospect’s profile can provide credibility, and suggests adding up to 150 people per week—about 20 per day. He gives rough benchmarks: 30–40% should accept a connection request, and 30–50% of those should reply. If people do not accept, the founder’s profile may look like an impending pitch; if they accept but do not respond, the message may not be interesting enough. He has not found AI personalization especially useful because it often tells recipients things they already know about themselves. For a 30-minute sales call, he begins with, “Why did you take the call with me today?” Ideally, the prospect explains the PULL framework unprompted. If not, the founder proposes an agenda based on what the prospect said, asks permission to ask discovery questions, and then investigates the project, urgency, alternatives, and limitations. The founder should feel relaxed and curious, not as though they must convince the prospect.

Once a prospect has demonstrated pull, Snyder recommends returning quickly with a concrete concept and plan. One founder selling into pharma had no prepared slides when an executive described a slow, ineffective market-research process. The founder summarized the need as faster research with the useful characteristics of the current process, then proposed returning in a few days with a rough mock-up. The founder could build an early version manually or with lightweight tools, get feedback, and ask what else the executive needed to decide. Snyder uses the PULL conditions to frame pricing. The value of completing the project and the cost of the current alternatives set intuitive boundaries. If a company currently spends $50,000 per year on market-research consultants, charging around that amount can make the price easy to understand. When the founder wants to discuss a large number without making it feel like an arbitrary demand, Snyder suggests presenting a case-study range: other customers have paid anywhere from $80,000 for a small project to $1 million for a larger one, with many landing in the middle, and then asking the buyer how they are thinking about it. He strongly distrusts proactive design partnerships, pilots, and free proofs of concept. People with real pull do not want to “help build” a product; they want the blocked project completed. A service-shaped initial offer can be appropriate, however. The team can deliver the work manually while learning what should eventually be automated, provided it does not get trapped permanently as a consulting business.

Snyder says founders should conduct roughly five sales conversations and then review what they learned. If prospects are consistently leaning back, the founder should determine whether any of them actually had pull, identify the mistaken assumption, and create a new hypothesis. A friend initially spoke with banking prospects and found no urgency, then received an introduction to a defense company where the buyer immediately responded, “Are you serious?” The team changed its hypothesis. Snyder also recommends “goblin mode” when progress is stalled: go to a conference, get on a plane, visit customers in person, or schedule 20 meetings and ask people what they would buy now. One founder’s parents went door to door at dental offices to find potential customers. After a sale, Snyder says the customer is effectively guaranteed to churn until reaching a specific retention milestone. The company must discover the moment at which it would be weird for the customer to leave and design onboarding to move every customer there. For one restaurant-retention product, that moment was receiving an alert that an employee was about to quit and successfully preventing the departure. For a horizontal AI-workflows company, most customers churned, but freight forwarders repeatedly used one workflow. The company narrowed its focus to that segment. Snyder’s overall operating model is a “case study factory.” The startup’s job is to repeatedly recreate one case study in which a customer says “hell yes” before buying and behaves as though the product is indispensable afterward. Pipeline, sales, and delivery are factory stages; the current bottleneck determines what the company should work on next. He applies the same process to new products inside established companies.