Skip to main content
San Francisco, 6th July 2025 As I am currently in the pre-PMF stage of Chonkie, I’ve been spending a good amount of time thinking about what makes a product reach PMF or if it’s a dead end. The classic advice, and the one taught at YC as well, is you’ll know when you have PMF. You’ll be struggling to “keep the product in stock,” users will be snatching it out of your hand, and the demand will feel overwhelming. Of course, “you’ll know when it comes” is solid advice, but it’s not great from a planning perspective. This kind of advice is not something I can detect in a small focus group or in a small amount of time. Neither is it something that makes PMF a strong detectable signal. “You’ll know when you’re there” is advice that kicks in when you’re deep into PMF. Ideally, I realize it much sooner, both so I can feel better about myself and plan for what to do next. So, how would I know when I’m close to PMF? I suspect it’s something to do with the kind of questions users ask. Usage is a fickle metric. Hype, free tiers, and promos can all make it look like you have extreme usage when you actually don’t. However, if a product is truly being used and giving people a valuable experience, they’ll start asking interesting questions. What makes a question “interesting”? It’s a question that forces you, the founder, to think of the product in a light you hadn’t considered before.
  • “Could I use this for X?” where X is something you hadn’t considered but doesn’t require new features. Users are seeing possibilities you didn’t design for.
  • “What happens when I have 10,000 of these?” or “How does this work with my existing workflow?” Users are imagining scale and integration.
  • “What if I wanted to do Y with this?” where Y stretches the product in new directions but within its core capabilities.
The key distinction here is between reactive questions and generative questions. Reactive questions include feedback like reporting bugs and requesting better performance. These are questions which you may find in the early days of a product that may solve a genuine problem, but is not big or ambitious enough to be a product that flies off the shelves. Generative questions are the interesting ones because they generate new possibilities. They’re almost always not feature requests that require more engineering efforts. Rather, they take the core premise and capabilities of the product you already have and extend it in new directions. It should not be hard to 1) get this signal, and 2) tell reactive questions apart from generative questions. Humans are inherently curious and very vocal. If you’re truly exciting users, they will wonder and ask you what is next. I know we are not there because our user requests hover mainly around performance and bug fixes. A constant stream of user feedback signals to me that we are solving a problem. But the fact these are mostly about performance and bug fixes signals to me that the specific service we offer is not large enough to reach PMF in the way that would support an ambitious startup. At minimum, our current offerings are not exciting users in a way that turns them into evangelists for Chonkie. The combination of constant feedback that isn’t inquisitive signals to me that we need to expand our offerings vertically, by building a more complete and lower touch service. On the flip side, I’ve talked to startups that I do believe have PMF, and their inbound looks like the generative questions I would expect them to have. For example, a startup I know initially built their platform to support the They built a fantastic product, and won a lot of customers. Soon enough though, companies that had internal software teams also started to reach out. Because if you think about it, an internal software team is just “implementing complex software” for the rest of the company. The core thesis of their product stayed the same, but it was now being used in a way that the founders did not initially imagine. A key thing to note here is that these ways do not require the founders to rethink their product. At most, they had to add a security feature or some new UI elements to start supporting the new use case. Of course this is not a perfect signal. Your early users might be enthusiastic, but you may find that there is a limited amount of them and the market you are in is simply not big enough. Worse, you could have a skewed pool of early users, maybe too many of them are your friends and family, consequently leading to biased feedback and forced usage. The worst thing that could happen is you get shut down by regulators. But assuming no bias in the sampling of your early users, I think having people asking interesting questions is a strong signal of PMF. It demonstrates that users see themselves continuing to use the product, are excited about its future, and are imagining new ways to use it. Excitement is the best thing you can hope for, and wonder is the precursor to it. This brings us to the practical question: how do you systematically listen for and categorize these questions? This framework works best when you’re getting questions from multiple users independently arriving at similar imaginative use cases. A single enthusiastic user might be an outlier, but when several users start asking about the same unexpected application, that’s signal. The key is to track not just the volume of questions, but their nature. Are users asking about bugs and missing features, or are they asking about possibilities and connections? Are they thinking tactically about the product as it exists, or strategically about how it fits into their broader world? This means structuring user conversations to go beyond “what’s broken” to “what’s possible.” Ask users not just about their current pain points, but about their workflows, their growth plans, their adjacent problems. The most interesting questions often come from these broader conversations. Currently, I’m tracking user questions informally in an Excel sheet, noting what seems curious to me. It’s not a systematic approach yet, but even this loose tracking has helped me see patterns. The nature of questions reveals to me whether users see Chonkie as a complete solution to their current problem or as a foundation for bigger ambitions. Successful evergreen businesses don’t run out of things to do. For a startup to become such a business, you need users who are constantly pushing you to think bigger. They should be the ones demanding that you grow, not just asking you to fix what’s already there. When users start asking questions that reveal they’re imagining your product as part of something larger, that’s when you know you might be onto something big.